phd thesis - region hovedstadens psykiatri · ken, b: kendall’s taub soi: second order...

PhD thesis

Jesper Bak

Mechanical Restraint Preventive Factors in Theory and Practice

Academic advisor: Mette Brandt-Christensen

Submitted: 29/01/15

Title: Mechanical Restraint

Subtitle: Preventive Factors in Theory and Practice

Subject description: Identifying mechanical restraint factors to generate knowledge of non-medical

MR-prevention.

Author: Jesper Bak

The studies were completed at Mental Health Centre Sct. Hans, Research Institute, Boserupvej 2,

4000 Roskilde, Denmark and St. Olavs University Hospital, Forensic Department Bröset, Centre for

Research & Education in Forensic Psychiatry, Norway, while the author was a PhD student at the

University of Copenhagen, Faculty of Health, and Medical Science, University of Copenhagen,

Denmark.

Supervisors:

Mette Brandt-Christensen, MD, PhD, Head of Department, Department Q, Mental Health Centre

Glostrup, Glostrup, Denmark.

Dorte Maria Sestoft, MD, PhD, Head of Clinic, Clinic of Forensic Psychiatry, Ministry of Justice,

Copenhagen, Denmark.

Vibeke Zoffmann, RN, MPH, PhD, Head of Research, Women and Children’s Health, Juliane

Marie Centre, Rigshospitalet, Copenhagen, Denmark.

Assessment committee:

Ingrid Egerod, Clinical Professor, Department of Clinical Medicine, University of Copenhagen and

Trauma Centre, Rigshospitalet, Copenhagen, Denmark (Chair).

Lise Hounsgaard, Professor, Odense Patient data Explorative Network, Institute of Clinical

Research, University of Southern Denmark, Odense, Denmark.

Frans Fluttert, Associate Professor, Centre for Research and Education in Forensic Psychiatry,

Ullevål University Hospital, Oslo, Norway, and senior researcher, FPC Dr. S. Van Mesdag,

Groningen, The Netherlands.

This thesis has been submitted to the Graduate School of the Faculty of Health and Medical

Sciences, University of Copenhagen.

2015 © All rights reserved. No part of this publication may be reproduced in any form or by any

means without clear indication of the source. DOI: 10.13140/RG.2.1.2681.5528

MECHANICAL RESTRAINT – PREVENTIVE FACTORS IN THEORY AND PRACTICE

1

1 Table of Contents

1 TABLE OF CONTENTS........................................................................................................... 1

2 PREFACE................................................................................................................................... 2

2.1 Acknowledgements ..................................................................................................................................... 2 2.2 List of Papers ................................................................................................................................................ 3 2.3 Table of Contents: Figures, Tables, and Appendixes ......................................................................... 3 2.4 List of Symbols and Abbreviations ......................................................................................................... 4 2.5 Key Definitions .............................................................................................................................................. 4

3 SUMMARIES IN DANISH AND ENGLISH ........................................................................... 5

3.1 Summary in Danish (Dansk Resumé) ..................................................................................................... 5 3.2 Summary in English .................................................................................................................................... 7

4 INTRODUCTION ....................................................................................................................... 8

5 OBJECTIVE ............................................................................................................................. 13

6 DESIGN AND METHODS ..................................................................................................... 14

7 RESULTS ................................................................................................................................. 24

7.1 Summary of Results, Study 1 .................................................................................................................. 24 7.2 Summary of Results, Study 2 .................................................................................................................. 27 7.3 Summary of Results, Study 3 .................................................................................................................. 29 7.4 Summary of Results, Study 4 .................................................................................................................. 31

8 APPROVAL AND ETHICS .................................................................................................... 35

9 DISCUSSION ........................................................................................................................... 35

9.1 Discussion of Individual Studies ........................................................................................................... 35 9.2 Design Issues .............................................................................................................................................. 45 9.3 Cause and Effect ........................................................................................................................................ 47

9.3.1 Chance ..................................................................................................................................................... 47 9.3.2 Bias........................................................................................................................................................... 48 9.3.3 Confounding ............................................................................................................................................ 50 9.3.4 Strength of the Association ................................................................................................................... 51 9.3.5 Biological Credibility ............................................................................................................................... 51 9.3.6 Generalisability and Consistency ......................................................................................................... 52

9.4 Discussion of International High Quality Research .......................................................................... 52

10 CONCLUSIONS ...................................................................................................................... 54

10.1 Conclusions from the Individual Studies ............................................................................................. 54 10.2 General Conclusions ................................................................................................................................. 55

11 PERSPECTIVES ..................................................................................................................... 56

11.1 Clinical Implications .................................................................................................................................. 56 11.2 Future Research Recommendations ..................................................................................................... 57

12 REFERENCE LIST ................................................................................................................. 58

13 APPENDIX ............................................................................................................................... 68

13.1 Appendix A................................................................................................................................................... 68 Paper 1

Paper 2

Paper 3

Paper 4


2

2 Preface

This thesis addresses the mechanical restraint (MR) of adult psychiatric inpatients. I have dealt with

this phenomenon for many years, first as a nurse in acute psychiatry later as a clinical nurse

manager of closed units, and finally as a worker in the education and development field. I have

always tried to minimise the use of MR by delivering empowering and humanistic nursing care and

simultaneously introducing safe and secure nursing procedures. Although some reduction in the use

of MR has been observed locally, much more progress is needed. The question thus arises of which

procedures, interventions or conditions we should develop and implement. When interviewing

patients in 2004 to collect their suggestions on how to reduce the number of MR episodes, they

gave many suggestions, convincing me that my research should generate knowledge of non-medical

MR-prevention. I hereby present this thesis.

2.1 Acknowledgements

I thank The Mental Health Centre Sct. Hans, especially Margit Asser, for supporting my PhD

education.

I would like to thank Mette Brandt-Christensen, Dorte Maria Sestoft and Vibeke Zoffmann for their

continual advice and supervision.

I would like to thank Thomas Werge and my co-workers at the research institute for helping me

with countless questions and for being my colleagues.

I would like to thank Roger Almvik and my former colleagues in Norway at the Bröset Centre for

Research and Education in Forensic Psychiatry for the interesting discussions and guidance and for

“having me around” in autumn 2010.

I would like to thank Helle Aggernæs, Mette Dons, Torben Hærslev and Louise Gjørup for their

collaboration on the European study.

I would like to thank Volkert Siersma for supervising the statistical analyses and enduring my

endless questions.

I would like to thank the clinical nurse managers of the psychiatric units in Norway and Denmark:

this study would not have been possible without you spending time on the questionnaires.


3

Special thanks are extended to some of the clinical nurse managers of the units at Mental Health

Department Østmarka, Mental Health Centre Copenhagen and Mental Health Centre Sct. Hans for

helping to pilot the questionnaire.

Finally, on a personal level, I would like to thank my family for their patience when the study

demanded late hours and months of separation that kept my attention away from them.

I would like to thank the following organisations for funding this project: the Mental Health Centre

Sct. Hans, the Health Science Research Foundation in the Capital Region of Denmark, the Mental

Health Services in the Capital Region of Denmark, Copenhagen University Hospital and the

National Board of Health.

2.2 List of Papers

1. Bak, J., Brandt-Christensen, M., Sestoft, D. M., and Zoffmann, V. Mechanical Restraint -

Which Interventions Prevent Episodes of Mechanical Restraint? - A Systematic Review.

Perspectives in Psychiatric Care 48, 83-94. 2011.

2. Bak, J. and Aggernæs, H. Coercion within Danish psychiatry compared with 10 other European

countries. Nordic Journal of Psychiatry 66(5), 297-302. 2011.

3. Bak, J., Zoffmann, V., Sestoft, D. M., Almvik, R., and Brandt-Christensen, M. Mechanical

Restraint in Psychiatry: Preventive Factors in Theory and Practice. A Danish-Norwegian

Association Study. Perspectives in Psychiatric Care 50(3), 155-166. 2014.

4. Bak, J., Zoffmann, V., Sestoft, D. M., Almvik, R., Siersma, V. D., and Brandt-Christensen, M.

Comparing the effect of non-medical mechanical restraint preventive factors between

psychiatric units in Denmark and Norway. Nordic Journal of Psychiatry 23(1), 1-11. 2015.

2.3 Table of Contents: Figures, Tables, and Appendixes

Figure 1: Mechanical Restraints in Denmark, 2004-2013

Figure 2: Effect and Recommendation Grade Matrix

Table 1: Hierarchy of Mechanical Restraint Preventive Factors

Table 2: Background Data

Table 3: Applied Methods of Coercion

Table 4: Physical Coercion per 100,000 Inhabitants in Denmark and Norway

Table 5: Ranking of Coercive Measures With Respect to Perceived Intrusion

Table 6: Effects of Potential Mechanical Restraint Preventive Factors on the Annual


4

Number of Mechanical Restraint Episodes

Table 7: Differences in the Effects of Potential Mechanical Restraint Preventive

Factors on the Annual Number of Mechanical Restraint Episodes in Denmark

and Norway

Appendix A: Appraisal, Grading, and Combining the Recommendations of the Mechanical

Restraint Preventive Factors

2.4 List of Symbols and Abbreviations

B: Estimate of regression coefficient n/No: Number

χ2: Chi square N.a.: No available information

CI: Confidence interval NICE: National Institute for Health and

Clinical Excellence

DRPS: Danish Reference Programme

Secretariat

Pearson’s r: Pearson´s correlation coefficient

Δ: Difference p: Probability value

EPSO: European Platform of Supervisory

Organizations

P.n.: Pro necessitate (as needed)

ERM: Early Recognition Method QIC: Quasi Likelihood under Independence

Model Criterion

Exp: Exponentiated QICC: Corrected Quasi Likelihood under

Independence Model Criterion

GEE: Generalized Estimating Equations RCT: Randomised controlled trial

IM: Intramuscular injection SIGN: The Scottish Intercollegiate

Guidelines Network

Ken, b: Kendall’s Taub SOI: Second Order Interaction

m2: Square meter UEMS: Union Européenne des médicins

Spécialistes (Psychiatric Section of the

European Union of Medical Specialists)

MERGE: Method for Evaluating Research

and Guideline Evidence

WHO, ICD-10: World Health Organization,

International Classification of Diseases, 10th

version

MR: Mechanical restraint

2.5 Key Definitions

Mechanical restraint (MR): The use of restraining straps, belts or other equipment to restrict

movement (1).

Coercion: The action of making somebody do something that they do not want to do (2).


5

Restraint: The use of physical force to control somebody who is behaving in a violent way (2).

Physical restraint/Holding: Physically holding the patient, preventing movement (1).

Seclusion/Isolation: Isolated in a locked room (1).

Time out: Patient asked to stay in room or area for period of time, without the door being locked

(1).

Forced admission: An involuntary admitted patient (3).

Forced compulsory retention: Forced compulsory retention of a voluntary admitted patient (4).

Forced treatment: An involuntarily treated patient, normally by antipsychotic medications (3);

sometimes called forced medication/long period (4).

Forced acute tranquillising medication/IM medication: Intramuscular injection of sedating drugs

given without consent (1); sometimes called forced medication/short period (4).

Constant observation/Person-to-person-/face-to-face observation: An increased level of

observation, of greater intensity than that which any patient generally receives, coupled with

allocation of responsibility to an individual nurse or other worker. Constant: within eyesight or arms

reach of the observing worker at all times (1). This type of measure is registered as a coercive

measure, in Denmark, if it is involuntary and prolonged for more than 24 hours (5).

Unit: The smallest organisational collection of beds, led by a clinical nurse manager.

3 Summaries in Danish and English

3.1 Summary in Danish (Dansk Resumé)

Tvangsfiksering af psykiatriske patienter indlagt på psykiatrisk afdeling er et voldsomt indgreb i

den enkelte patients autonomi og kan medføre både fysiske og psykiske skader. Det kan dog i

særlige situationer være nødvendigt blandt andet for, at beskytte patienten selv, medpatienter eller

andre. En sådan situation kan for eksempel opstå hvis en indlagt patient er meget aggressiv,

voldelig, selvskadende eller suicidal truet.

Selvom også længerevarende tvangsfikseringer udføres i henhold til den danske lovgivning,

understregede Den Europæiske Komité for Forebyggelse af Tortur (The European Committee for

the Prevention of Torture and Inhuman or Degrading Treatment or Punishment) i rapporter fra


6

2002, 2008 og 2014 at, tvangsfiksering af psykiatriske patienter flere dage i træk, ikke har nogen

behandlingsmæssig berettigelse og må beskrives som mishandling/fejlbehandling.

Der er bred enighed om, at anvendelsen af fysisk magtanvendelse herunder tvangsfiksering i

behandlingen af psykiatriske patienter bør mindskes mest muligt, men hvordan dette skal gøres i

praksis, hersker der usikkerhed om. Evidensen på området er stigende, men der mangler fortsat

sikker evidens. Ingen studier har undersøgt hvilke faktorer, der mest effektivt forebygger

anvendelse af tvangsfiksering i praksis i Danmark og Norge.

Det overordnede formål med denne afhandling er at generere viden om ikke-medicinsk

forebyggelse af tvangsfikseringer. Afhandlingen består af fire individuelle studier: Først

gennemførtes et systematisk review af den internationale forskningslitteratur for at identificere

evidente og effektive faktorer (forhold og interventioner), der kunne reducere antallet af

tvangsfikseringer. Formålet var at skabe grundlag for at udvikle et spørgeskema med henblik på at

belyse, i hvilken grad formodede forebyggende faktorer var til stede i de psykiatriske afsnit, som

indgik i afhandlingens studie 3 og 4. Dernæst blev psykiatrilovgivning og anvendelse af tvang i 11

europæiske lande beskrevet og sammenlignet i samarbejde med Sundhedsstyrelsen. Resultaterne

dannede grundlag for udvælgelse af et egnet andet europæisk land med sammenlignelige forhold

som kunne indgå i afhandlingens 3. og 4. studie. Afhandlingens 3. og 4. studie var baseret på en

multicenterspørgeskemaundersøgelse, hvor data blev indsamlet fra lukkede psykiatriske afsnit i

Danmark og Norge for at belyse i hvilken grad de enkelte tvangsfikseringsforbyggende faktorer var

til stede i afsnittene og antallet af tvangsfikseringer det forgange år. Formålet for det 3. studie var, at

undersøge hvorvidt tilstedeværelse af formodede tvangsfikseringsforebyggende faktorer i et afsnit

var associeret med mindre anvendelse af tvangsfiksering. I afhandlingens 4. studie blev data

analyseret med henblik på at identificere, om nogle af de formodede forebyggende faktorer

muligvis kunne forklare forskellen i antallet af tvangsfikseringer mellem Danmark og Norge.

Der blev identificeret tre forebyggende faktorer, som var signifikant associeret til et lavt antal

tvangsfikseringer i Danmark og Norge: Obligatorisk review af tvangsfikseringsepisoder (64 %

færre tvangsfikseringer), Patient involvering (58 % færre tvangsfikseringer) og Crowding (46 %

færre tvangsfikseringer). Derudover blev identificeret fem faktorer, der havde konfunderende effekt

(reducerede forskellen mellem landene), disse kunne være en del af forklaringen på, hvorfor

Danmark anvendte 92 % mere tvangsfiksering end Norge. De fem var: Personalets

grunduddannelses niveau (51 % af effekten), Anvendelse af afløsnings personale (17 % af


7

effekten), Arbejdsmiljø (15 % af effekten), Andelen af personaler per patient (11 % af effekten) og

Identifikation af patientens krise-triggere (10 % af effekten).

Resultaterne har uddybet vores viden om hvordan specifikke faktorer reducerer antallet af

tvangsfikseringer. De er ikke identificeret ved gennemførelse af randomiserede kliniske studier, og

selv om nogle svagheder findes i studierne, blev spørgeskemaet udviklet på en grundig måde, der

blev justeret for flere mulige konfunderende faktorer i analyserne, og lignende resultater er fundet i

tidligere forskning fra andre lande. Derudover er der ikke noget, der tyder på, at de fundne faktorer

kan have negative effekter, snarere kunne det tænkes, at den mulige positive effekt, faktorerne har

på forebyggelse af tvangsfiksering, kunne være et udtryk for en generel styrkelse af pleje- og

behandlingsmiljøet. Det anbefales derfor, yderligere at undersøge effekten af at implementere:

obligatorisk review, patient involvering, mindre crowding, højere grunduddannelsesniveau blandt

personalet, anvendelsen af færre vikarer, et bedre arbejdsmiljø, højere antal personaler per patient

og identificering af patienters krise triggere, i praksis i Danmark og Norge.

3.2 Summary in English

Mechanical restraint (MR) is a major infringement on the psychiatric patient’s autonomy. MR can

cause physical and mental harm but may be necessary, e.g. to avoid putting an individual’s health at

risk. The nursing staff is tasked with protecting the life and health of not only the individual patient

but also other patients and relatives. A situation can occur in which staff is obligated to use force

and occasionally MR, e.g. if a patient is very aggressive, violent, self-destructive or suicidal.

Although MR is legal, the European Committee for the Prevention of Torture and Inhuman or

Degrading Treatment or Punishment concluded, in two reports from 2002, 2008, and 2014 that no

medical justification exists for applying instruments of physical restraint to psychiatric patients for

days and that doing so amounts to ill treatment.

Although the number of MR episodes should be reduced as much as possible, how this goal should

be accomplished is quite unclear. Despite the growth in available research in the area decisive

evidence is still lacking. No studies have investigated which of the many MR-preventive factors are

the most effective in practice, in Denmark and Norway.

Therefore, the overall objective of this thesis was to generate knowledge of non-medical MR-

prevention. The four studies that contributed to this objective had separate purposes. First, a

systematic review of international research papers was conducted to identify evident and effective

MR-preventive factors; this review served as a basis for developing a questionnaire examining the

degree to which MR-preventive factors have been implemented in psychiatric units and the


8

numbers of MR episodes in those units. Second, a comparative investigation of European countries

was conducted to identify a country comparable to Denmark to include in the cross-country

questionnaire survey. Third, the collected questionnaire data were analysed to identify the

associations between the MR-preventive factors and the number of MR episodes. Finally, the data

were analysed to identify if the MR-preventive factors could explain the difference in the number of

MR episodes between Denmark and Norway.

Three MR-preventive factors were significantly associated with a low frequency of MR episodes in

Denmark and Norway: a mandatory review of MR episodes (64% fewer MR episodes), patient

involvement (58% fewer MR episodes), and crowding (46% fewer MR episodes). Further, we

identified five MR-preventive factors with confounding effects (reducing the difference between

countries), which may explain in part why Denmark used 92% more MR compared with Norway.

These factors included: staff education (51% of the effect), substitute staff (17% of the effect), work

environment (15% of the effect), patient-staff ratio (11% of the effect), and identification of

patients’ crisis triggers (10% of the effect).

These results have increased our understanding of the ability of specific MR-preventive factors to

reduce the number of MR episodes in Denmark and Norway, thereby generating knowledge in the

field of MR-prevention. These findings have not been identified via randomised controlled trials

(RCTs), and although some biases could be present, the questionnaire was thoroughly developed to

include several potential confounders. Furthermore, similar results have been demonstrated in

previous international studies. These factors are not likely to have adverse effects on the patients or

staff. Rather, the potential positive effects of these factors on the prevention of MR episodes may

reflect a general strengthening of the care and treatment environments. Therefore, further

investigation into the effects of implementing the following within Danish and Norwegian practices

is recommended: mandatory review, patient involvement, less crowding, higher staff education, less

substitute staff use, better work environment, increased number of staff per patient, and the

identification of the patient’s crisis triggers.

4 Introduction

In Denmark, MR can legally be performed, e.g. when psychiatric inpatients pose a risk to

themselves or to others (6). Other countries do not allow the use of MR: forexample, only the use of

seclusion and holding (physical restraint) is allowed (except in exceptional circumstances in Special

Hospital environments) in the United Kingdom. Thus, the use of MR differs greatly across Europe


9

and the rest of the world. The practical procedure is primarily prescribed by a psychiatrist, is

performed by nursing staff and involves using leather or fibre belts to secure the patient to a bed,

occasionally with additional straps used around the wrists and ankles.

MR using leather belts has been used in Denmark since, at least 1870, when the use of a

“straitjacket” was abolished (7). Reiter described the use of MR in 1927 (translated by the author):

Of restraints, the belt (ed. MR) is especially used; this is an upholstered girth, that can

be locked around the patient … and secured to the bed … It is only used when other

measures fail and only after a doctor’s prescription. (7)

Unlike most other countries, the use of seclusion is illegal in Denmark. In 1901, Professor Knud

Pontopidan argued (translated by the author):

These (ed. chains and straitjackets) were abandoned from the asylums

armamentarium long ago. However, the medical profession has first achieved its goal

in this area, by abandoning the remaining mechanical restraints, which represents

themselves in the closed cell. The future solution must involve consistently combating

cell treatment. (8)

In 1985, the first legislation (9) on the use of specific types of restraint was published, and seclusion

was finally abandoned, except in a special high security department (Sikringen).

In Denmark other types of coercion can legally be used for psychiatric patients, including forced

admissions/compulsory forced retention, forced treatment, forced acute tranquillising medication,

holding, person-to-person observation, follow-up after discharge, mail examination, patient rooms,

personalised alarm systems, and protective MR (6). Legislation stipulates that the least intrusive

type of coercion must be used only after trying everything possible to avoid using coercion. The

most intrusive coercive measures used in Denmark are: forced admission/compulsory forced

retention, then MR, forced treatment, forced acute tranquillising medication, and holding, and

finally person-to-person observation (10). Danish legislation follows the United Nations Principles

for the protection of persons with Mental Illness and the improvement of Mental Health Care (11).

Thus, as most European Countries also follows these principles, which do not describe the

intrusiveness of the different types of coercion, individual countries likely defined them based on

their own interpretation of human rights, as influenced by their historic– and cultural development.

In Denmark, MR is considered the most intrusive intervention for controlling aggressive or self-

mutilating/suicidal inpatients in psychiatric units.


10

According to the Danish Mental Health Act, the following conditions must be present to legally

initiate MR (translated by the authors):

Mechanical restraint may be used only when necessary to prevent a patient from

1. exposing their body or health or the body or health of others to danger,

2. pursuing or grossly molesting fellow patients in any way, or

3. committing significant acts of vandalism. (6)

The same conditions must be present to initiate holding. If a patient is very agitated, forced acute

tranquillising medications can be used. The same conditions must be present in many other

countries to initiate MR, seclusion, holding, or forced acute tranquillising medication (10). The

central concept is danger towards self or others, which is the most common reason for using

coercive interventions towards psychiatric patients behaving in an aggressive or self-

mutilating/suicidal maner. Thus, although the use of MR is legal in many countries, the discussion

is ongoing regarding how often this intervention should be used, given the patients’ right to self-

determination and human rights in general. Following visits to Denmark, the European Committee

for the Prevention of Torture and Inhumane or Degrading Treatment or Punishment stated that no

medical justification exists for applying instruments of physical restraint to psychiatric patients for

days and that doing so amounts to ill treatment (12-14).

MR exposes the patient to potential physical injuries such as deep vein thrombosis, pulmonary

embolisms, pressure sores, and bone fractures (15-19). For most patients, MR also involves

negative psychological effects: feeling helpless, angry, trapped, sad, fearsome, unpleasant, or

offended. Individual patients often regard the use of MR as unjust or even as a punishment (20-24),

in the worst case developing long-term trauma and possibly features of post-traumatic stress

disorder (22-24). In addition to affecting the individual patient, the use of MR has a negative effect

on the atmosphere, care, and treatment of the whole unit (20;21). These negative consequences

appear inconsistent with health care providers’ basic attitudes regarding “doing good” for patients.

However, strong counter-arguments regarding the use of MR suggest that health care professionals

are obliged to secure and care for the health and life of the patients and other individuals, while

avoiding violence towards fellow patients, relatives, and staff. Some studies indicate that up to 10%

of patients request the use of MR to prevent acting out (4). Overall, enhanced efforts are required to

reduce the frequency (10).

In a textbook on mental health nursing (translated by the author), Jacobsen wrote the following in

1920:


11

“Straitjackets”, where the arms were tied to the chest, as used in earlier times, are

not currently used in any mental health hospital, and even the afore mentioned milder

restraints (ed. MR) are used less frequently and will most likely be rendered

superfluous by numerous and careful nursing staff. (25)

This goal has not been reached; instead, the numbers of MR episodes in Denmark have remained

constant over the previous decade (26) (Figure 1).

The relevant legislation, which was thoroughly

revised in 1989 has been continually updated, with a

special focus on efforts to reduce the use of MR.

These efforts have shown no significant effect,

which is consistent with reports in other countries

(27). This finding is most likely because many

factors influence the events that lead to the use of

MR. The factors that influence an individual

patient’s aggressive or self-destructive behaviour

and hence the frequency of MR use in a psychiatric

unit could be related to the patient, unit, staff or

“situational/interactional conditions” (28). As

Bowers (29) describes in the very recently

developed “Safewards Model” on conflict and

containment on psychiatric wards, six domains of

originating factors exist: the staff team, the physical

environment, outside of the hospital, the patient

community, patient characteristics, and the

regulatory framework. All of these domains give rise

to potential flashpoints, with the capacity to trigger conflict and/or containment, representing

another way to categorise the originating conditions while underpinning the fact that many factors

influence how often coercive measures are used in psychiatric units.

In Denmark the 6,165 MR episodes recorded in 2013 involved 2,084 individual patients with an

average of 3.0 episodes per patient; 14% of these episodes lasted longer than 48 hours. The

distribution of MR episodes across the five regions in Denmark was as follows: Capital, 115 per

100,000 inhabitants; Zealand, 67 per 100,000 inhabitants; South Denmark, 78 per 100,000

Figure 1. Mechanical Restraints in Denmark, 2004–2013

Number of Mechanical Restraint Episodes

Number of Individuals who Experienced Mechanical

Restraint

Note. The numbers does not include the special high security

department (Sikringen).

0

1000

2000

3000

4000

5000

6000

7000

2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

0

500

1000

1500

2000

2500

2004 2005 2006 2007 2008 2009 2010 2011 2012 2013


12

inhabitants; Mitt Jutland, 165 per 100,000 inhabitants; and North Jutland, 101 per 100,000

inhabitants. The highest numbers of MR episodes were found in regions with the largest cities;

however, despite of including only the second largest city, Mitt Jutland had the highest numbers of

MR episodes. A similar tendency was found upon investigating the MR episodes in relation to the

number of psychiatric beds, or the number of psychiatric patients across the regions (26). Other

conditions in addition to the often-seen urban/rural differences (30) must also influence the regional

distribution of MR episodes.

In Denmark in 2008, the average length of a MR episode was 27 hours. The average age of the

patients subjected to MR was 40 years, and the average age of all inpatients was 42 years,

indicating that patients sumitted to MR was younger. More male (58%) than female patients

experienced MR, although men represented 51% of all inpatients. The diagnostic distribution of

patients experiencing MR were (WHO, ICD-10, principal codes): F00-09, 8.5%; F10-19, 15.3%;

F20-29, 46.6%; F30-39, 16.0%; F40-49, 3.9%; F50-59, 0.3%; F60-69, 6.1%; F70-79, 1.6%; F80-89,

0.6%; and F90-99, 0.9%. The diagnostic distribution of all inpatients was as follows: F00-09, 2.4%;

F10-19, 16.1%; F20-29, 31.5%; F30-39, 24.8%; F40-49, 15.2%; F50-59, 0.7%; F60-69, 6.4%; F70-

79, 0.5%; F80-89, 0.4%; and F90-99, 1.2%. Thus, the patients experiencing MR were diagnosed

more often with F20-29 and less often with F30-39 or F40-49. Different types of departments had

different diagnostic distributions. For example, in special forensic departments, 46% of the patients

were diagnosed as F20-29 and 12% as F30-39, whereas these frequencies were 31%, and 25% in

acute departments, respectively. In 2008, 22 adult inpatients experienced MR more than 25 times,

comprising 1,150 total MR episodes. Nineteen of these inpatients were woman and 17 were

diagnosed F20-29. Their average age was 30 years, they were admitted in all regions, and they

spend an average of 280 days as inpatients. Thus, few patients experience numerous MR episodes,

and whether these patients respond to the same MR-preventive initiatives as other patients is

unknown. These 2008 data originate from the Danish Health and Medicines Authorities Register on

Coercion and the Danish Psychiatric Central Research Register (31).

Between 1999 and 2003, immigrants and refugees experienced more than twice the amount of MR

episodes as Danish-born population did (32) perhaps because the rates of psychosis are higher

among migrants than non-migrants, because family support systems are lacking, or because

language barriers existed between the staff and migrants in the psychiatric units. Although

differences in MR rates are known to exist, the reason for the difference remains unknown. The

same differences regarding regions, diagnoses, genders, departments, the low number of patients

experiencing multiple MR episodes, and ethnicities are found in other countries (33-37).


13

More than 30 reviews of the management of agitation, aggression, violence, and coercion (such as

MR, seclusion, physical restraint, and forced acute tranquillising medication) were identified. Ten

of the reviews explicitly included MR, several of them quite recent (38-42), and all of the reviews

described the absence of high-quality research1. No meta-analyses or meta-syntheses

2 were

identified. Thus, which preventive interventions are effective and which interventions only appear

to be effective due to favourable study conditions remain unknown.

Physical restraint/holding, constitutes a coercive intervention in itself but is also employed as part

of several other procedures e.g. MR, seclusion, and forced acute tranquillising medication. Thus if

preventive interventions reduces the number of seclusions, forced acute tranquillising medications

or physical restraints, they are likely also effective in reducing MR. After all, each of these

modalities represents a staff management strategy towards aggressive or self-mutilating/suicidal

patient behaviours.

Numerous factors have been recommended to reduce the use of MR, including risk management,

physical environment, and alarm systems, staff attitudes and care alliance. In risk prediction, the

often highlighted issues include antecedents, warning signs, and risk assessments. The training,

education, skills, competencies, and attitudes of the staff who work with psychiatric patients, as

well as the de-escalation methods and techniques, post-incident review, staff environmental factors

(both mental and physical working environment), medication, effective illness and abuse treatment,

effective observation, and physical training have also been considered (43;44).

Although some of these actions have been locally proven to reduce the number of times MR is used

(45;46), national figures have failed to significantly decrease in some countries, e.g., Denmark and

Norway (26;47).

5 Objective

As described in the introduction, MR is presumed to be the most intrusive intervention towards

aggressive or self-mutilating/suicidal psychiatric inpatients in Denmark. MR is considered a high-

risk intervention for both patients and staff, and many preventive interventions, mostly non-medical

1 High quality research may comprise the following types: qualitative such as at least one high-quality meta-synthesis,

systematic review, meta-summary, or a body of high-quality evidence; or quantitative such as at least one meta-

analysis, systematic review or body of high quality RCTs. 2 A meta-synthesis is similar to a quantitative meta-analysis. Definition: Meta-synthesis “is the bringing together and

breaking down of findings, examining them, discovering essential features, and, in some way, combining phenomena

into a transformed whole” (Schreiber, R., Crooks, D., and Stern, P. N. Qualitative Meta-analysis. Morse, J. M.

Completing a Qualitative Project: Details and Dialogue. 311-326. 1997. Thousand Oaks, Sage).


14

in nature, have been suggested in the international literature. Despite political/legislative initiatives,

the number of MR episodes does not seem to be decreasing in Denmark; thus, the overall objective

of this thesis was to generate knowledge of non-medical MR-prevention.

The four studies that contributed to this objective had separate purposes:

Study 1: Which conditions in nursing and which nursing interventions have been shown

to reduce the frequency of mechanical restraint episodes?

Study 2: To compare the use of coercive measures on psychiatric inpatients between

Denmark and Sweden, Norway, Finland, Iceland, Belgium, The Netherlands,

United Kingdom, Ireland, France, and Italy, in order to share knowledge and

learn about what other countries do to minimize coercion.

Study 3: How are presumed MR-preventive factors of non-medical origin associated with

the frequency of MR episodes in Denmark and Norway?

Study 4: To examine how presumed MR-preventive factors of non-medical origin may

explain the differing number of MR episodes between Denmark and Norway.

6 Design and Methods

To generate knowledge of non-medical MR-prevention, the research was designed in the following

manner:

1. A systematic review of international research papers, combining quantitative and qualitative

research, was conducted to identify evident and effective MR-preventive factors (Study 1).

2. A descriptive comparative investigation, regarding the use of coercive measures in European

countries, was performed to gain knowledge about the use of coercion in other European

countries and to identify a comparable country that uses less MR to be included in the

subsequent questionnaire survey (Study 2).

3. A questionnaire survey based on the results of the review (Study 1) was conducted to

examine the relationship between the number of MR episodes and the degree to which the

MR-preventive factors were implemented in “closed” psychiatric units in Denmark and

Norway (Studies 3 and 4).


15

4. Finally, two analytic investigations were conducted to identify the potential associations

between the number of MR episodes and MR-preventive factors (Study 3) and to determine

to what degree the identified MR-preventive factors could explain the different number of

MR episodes between Denmark and Norway (Study 4).

Regarding the systematic review (Study 1), this particular design was chosen because very few

RCTs and no meta-analyses were found. Therefore, searches were not limited to these study

designs. Additionally, some of the areas under investigation could not be covered by quantitative

papers alone. Therefore, a combination of quantitative and qualitative papers into ranked

recommendations was developed to deduce maximal information from the available papers.

The systematic review process was based on the principles outlined by, first, the Danish Reference

Programme Secretariat (DRPS) (48), and the National Institute for Health and Clinical Excellence

(NICE) (43) for the evaluation of the quantitative research, and second, by Sandelowski and

Barroso (49) and Kearney (50) for the evaluation of the qualitative research. The Scottish

Intercollegiate Guidelines Network (SIGN) (51) inspired the process of the combined evaluation of

both qualitative and quantitative research. The combination of principles from the above mentioned

sources was necessary because no one source in itself covered both quantitative and qualitative

research quantification.

The stages of the review process were as follows: developing the scope of the review, drafting

review questions with a clinical focus, and inclusion and exclusion criteria, implementing a search

strategy and search phrases, identifying sources of evidence, carrying out searches, sifting through

the evidence, extracting relevant data from studies meeting both the methodological and clinical

criteria, evaluating the methodological quality using relevant checklists, compiling papers into

relevant evidence tables, interpreting and grading each paper, abstracting interventions,

summarising intervention levels, separately grading the recommendations for the quantitative and

qualitative evidence, and summarising the whole body of evidence.

The scope of the review was to answer the following question: which conditions in nursing and

which nursing interventions have been shown to reduce the frequency of MR episodes? Next,

review questions were developed, e.g. “Do communicative de-escalation techniques prevent MR?”

or “Do specific types of limit-setting prevent MR?” The searched papers included original peer-

reviewed papers on the care of adult psychiatric inpatients that were physically restrained. Physical

restraints were included in several procedures e.g. MR, seclusion, and forced acute tranquillising

medications; as such, if the preventive factors described in the literature reduces the number of


16

seclusions, forced acute tranquillising medications, or holdings, they are likely to be effective

towards MR reductions as well. Papers were excluded if they were reviews, or if they described

interventions and conditions to prevent physical restraint in non-psychiatric inpatient settings,

learning disability settings, nursing homes, or prisons. The papers were obtained from 1998 to April

2009. The searches included English, Danish, Norwegian, and Swedish language citations, all with

English abstracts. A total of 84 key words were used in different combinations and inflections in the

search, e.g. affect, aggression, alliance, coping, debriefing, intervention, and restraint.

The structure of the search strategy was as follows: an overarching strategy for conditions and

interventions across selected databases, a search to identify guidance and reports (for example,

governmental reports, not indexed in the major databases), and topic-specific search strategy on

PubMed. For each condition or intervention, evidence supporting effectiveness or harm was sought.

A manual search was performed in key areas that were not electronically indexed (e.g., PhD thesis),

and the reference lists of the most relevant papers were checked for articles not otherwise found.

Sifting through the evidence was completed in three steps. First, we sifted through the material that

potentially met the eligibility criteria based on the title/abstract. Second, full papers were reviewed

for relevance/eligibility or for vague relevance/eligibility. Third, the papers were critically

appraised checked for quality.

Checklists for evaluating the methodological quality of the papers, and evidence tables for

compiling the studies were gathered (48-50;52-54). Once the individual papers had been assessed

by the author for methodological quality and relevance in terms of the clinical questions, they were

graded according to the levels of evidence. Grading of the quantitative research papers was

conducted as outlined by the DRPS (48). The levels of evidence and grades of recommendation

were developed from the DRPS (48), the Method for Evaluating Research and Guideline Evidence

(MERGE) (54), and SIGN (51). Grading of the qualitative research was performed according to the

levels of complexity in the evidence described by Kearney (50), and the process was inspired by

Sandelowski and Barroso (49), Kearney (50), the MERGE (54), the DRPS (48), and the SIGN (51).

The evidence for each intervention/condition was then compiled into tables. The quantitative and

qualitative grading systems can be found at the end of Appendix A, the different checklists are

available upon request from the author.

The final synthesis was based on consensus decisions by the four authors of Paper 1 and was guided

by the following three questions. Which interventions match the patients’ views and experiences?


17

Which interventions have yet to be evaluated in clinical settings? Does the effect of the intervention

seem plausible from a clinical point of view?

The results of the systematic review are displayed in a hierarchy, combining recommendation

grades and effects. The recommendation grade describes the ability of an intervention to exert an

effect by reducing the number of mechanical restraints in the clinical setting—that is not the degree

that such an intervention would reduce the number of mechanical restraints but rather whether the

intervention would reduce the number. This procedure is described in five levels from 1 to 5, with 1

being the best. Accordingly, 1 indicates an intervention that most certainly would have an effect on

the phenomenon (MR), and 5 indicates a high degree of uncertainty as to whether the intervention

would have an impact on or have any relation at all to the phenomenon. Combining the quantitative

and qualitative recommendations was achieved by taking the better of the two. The effect is

considered as effect of the intervention on MR (MR, seclusion, physical restraint, etc.) reductions,

whereas the reduction in the number of MR was divided into five levels from 1 to 5, with 1 being

the best and ranging from 80% to 100% reduction, 2 from 60% to 80%, and so on. The effect of the

intervention was extrapolated from the quantitative papers and combined in the following manner:

with no quantitative papers, the level was set to 5 (0–20% reduction); with one, the level was

transferred directly; with two or more the level was first influenced by the paper with the best

recommendation grade, second by the paper best describing the intervention, and third by a simple

average.

Regarding the comparative investigation into the use of coercive measures in European countries

(Study 2), this particular design was chosen to screen selected European countries based on their

use of coercive measures. This aim was achieved in two ways: first, by reviewing international

reports, research literature, and national registers, and second, by performing a survey that asked

government officials and medical specialists to complete a questionnaire.

Primarily to examine “numbers” on background features and coercive measures, the literature

review covered the legislation and coercive measures used in psychiatry across Europe follows:

forced medication/long period, forced medication/short period, MR, seclusion, holding, time out,

and constant observation. The review was conducted between April and June 2009, with an update

in March 2010, and covered the years from 1999 to 2010. The searches included literature in the

following languages: English, Danish, Norwegian, Dutch, Finnish, and Swedish. The search

strategy was structured as: an overall search for legislation and coercion across selected databases, a


18

search to identify registers and reports (e.g. governmental reports not indexed in the major

databases), a topic-specific search on PubMed, and lastly, the reference lists of the most relevant

papers to identify articles not otherwise found. Once the references were retrieved, sifting process

was initiated including a first sift for material that potentially met the eligibility criteria based on

title/abstract and a second sift for the full references that appeared relevant and eligible or in which

the relevance/eligibility was not clear from the abstract. Finally, a third sift allowed for the critical

assessment of the references. The focused search strategy is laid out in Paper 2, Supplementary,

Appendix 1.

The questionnaire survey was primarily developed to gather knowledge on the legislation across the

selected European countries by the two authors and two representatives from the mental health

authorities in Denmark. The questions were based on Danish legislation, and on knowledge gained

from nationwide reports and papers (e.g. Policies and practices for mental health in Europe (55),

including the compulsory admission and involuntary treatment of mentally ill patients - legislation

and practice in EU-member states (56), among other (57-62)). The questionnaire covered legislative

practices in the same areas mentioned earlier in the literature review.

The construction of the questionnaire followed a guide developed by Olsen (63) comprising six

phases: guiding the formulation of the problem, understanding the questions, recalling information,

evaluating contextual effects, editing answers, and conducting tests. The test-retest validation was

primarily internally conducted among the members of the constructed group, but two responders

were asked if they understood and were able to answer the questionnaire.

The respondents included governmental health staff working with mental health issues/legislation,

because they are the most knowledgeable of the mental health legislation in the selected countries.

The questionnaires were sent to the members of the European Platform of Supervisory

Organisations (EPSO), a body that was formed in 1996 to create a network of supervisory

organisations in health care. Additionally, the questionnaire was sent to the members of the

Psychiatric Section of the European Union of Medical Specialists (Union Européenne des médicins

Spécialistes, UEMS), to gain knowledge regarding possible differences between the current

legislation and its translation into practice. These respondents represent medical specialists in

psychiatry from each of the different countries.

An example from the questionnaire can be found in Paper 2, Supplementary, Appendix 2. The

questionnaire survey was administered between May and July 2009. The results are shown as

purely descriptive “raw” data.


19

Studies 3 and 4 were designed as a cluster cross-sectional questionnaire survey based on the results

of the review (Study 1). The survey was conducted to examine the relationship between the number

of MR episodes and the degree to which the MR-preventive factors were implemented in “closed”

psychiatric units in Denmark and Norway (Studies 3 and 4).

Norway, which has very similar legislation, was included in the survey to increase the power of the

latter analysis and to strengthen its overall transferability. The following conditions must be present

to legally initiate MR according to the Danish Mental Health Act (translated by the authors):


1) exposing their body or health or the body or health of others to danger,

2) pursuing or grossly molesting fellow patients in any other way, or

3) committing significant acts of vandalism. (64)

These conditions are consistent with Norwegian legislation (translated by the authors):

Restraints may be used on a patient only when absolutely necessary to prevent him or

her from damaging him/herself or others or to prevent damage to buildings, clothing,

inventory or other things. (65)

Additionally, Norway had a lower rate of physical coercion (MR, seclusion, and physical

restraint/holding) with 29 patients per 100,000 inhabitants experiencing physical coercion compared

with 52 patients per 100,000 inhabitants in Denmark (4). Specifically regarding MR, Denmark had

122 episodes of MR per 100,000 inhabitants compared with 87 episodes of MR per 100,000

inhabitants in Norway in 2007 (66;67). The lower frequency of MR episodes in Norway might

indicate that more effective MR-preventive factors have been implemented in Norway compared

with Denmark or that MR-preventive factors exhibit differing effects between the countries.

In the questionnaire MR is defined as the use of restraining straps, belts or other equipment to

restrict movement (1). Notably, this definition refers only to the MR of adult (18-65 years)

inpatients in psychiatric hospitals and does not refer to the use of MR in prisons or nursing homes.

The MR-preventive factors identified by the literature review (Study 1) were included in an online

questionnaire and sent to the clinical nurse managers (the respondents) to quantify whether they

were implemented in the units. These managers were chosen because they had exclusive access to

the requested information.

The Danish questionnaire was administered between 23 December 2009 and 28 May 2010, and the

collected data pertained to 2008. The Norwegian questionnaire was administered between 29


20

October 2010 and 27 January 2011, and the collected data pertained to the previous 12 months,

most of which occurred in 2010.

The study included all psychiatric hospital units in Denmark (n = 87) and Norway (n = 96) that

treated adult inpatients (18-65 years) and used MR for approximately two years prior to the year of

the investigation. MR did occur in Norway in settings other than hospitals; however, given that only

1.5% of recorded MR episodes in 2007 occurred in these settings (67) and that hospitals served as

the primary focus, these units were not included. The special high-security department in Denmark

was excluded due to specific legislation on coercive measures.

If MR was used more than 25 times on the same patient during the investigation period, then the

episodes of MR > 25 were excluded. Chance determined the units to which these few patients were

admitted, but the presence of these patients could significantly affect the MR frequencies of these

units, thereby distorting the results, as described in the introduction. The number of 25 times on the

same patient during one year was chosen after visually examining a graph of MR episodes per

patient in the Danish cohort. The examination showed that the number of times the same patient

experienced MR slowly increased until the point of 25 times, were a rapid increase in times was

found, i.e. extreme cases.

In the development of the questionnaire each MR-preventive factor was “translated” into questions

to identify the degree to which the factor was present in the units. Six factors were not included.

Combined intervention programmes were also omitted because of the complexities in their

measurement and the implications of other factors, e.g. training, work environment, and early

recognition of patient triggers. The following three factors: treatment of substance abuse and

psychosis, pro necessitate (p.n.)/as-needed medication guided by individual patients, and care-

alliance, were omitted because they were not considered to be measured appropriately by a

questionnaire, leading to the emergence of eventual confounding effects created by the three

factors, was covered by other covariates such as the type of unit, type of bed, number of units

nested in the departments, or number of MR episodes experienced by patients with a forensic

provision. For instance, if care-alliance was associated to the outcome and was poorer between

patients with a forensic provision and the staff, because some are involuntarily admitted for long

periods of time, this effect would be reduced by the covariate, number of MR episodes experienced

by patients with a forensic provision. Finally, two MR-preventive factors: music used as a diversion

or for relaxation, and patients perceived as a resource in their own treatment were omitted, as these

factors were partially covered by other factors, such as activity-based nursing and patient


21

involvement. The omitted MR-preventive factors only represented a problem for the later analyses

if they acted as confounders in the association between the other MR-preventive factors and the

outcome; thus, at the same time not associated with one of the covariates making them act as proxy

covariates. The odds for this situation to occur are very small.

Cognitive milieu therapy and patient involvement as well as patient-centred care and patient

involvement were separated into three factors: cognitive milieu therapy, patient-centred care, and

patient involvement. Although patient involvement is part of cognitive milieu therapy and patient-

centred care, patient involvement might also occur by itself. These stipulations left; 22 separate

MR-preventive factors to be included in the questionnaire.

The questionnaire was developed by conceptualising and operationalising the MR-preventive

factors. To increase validity and reliability the Question Appraisal System QAS-99 (68) was utilised

to identify problems within the questionnaires, including the cognitive processes inherent in the

question-answering process. The construction and pre-testing of the questionnaire involved an

eight-step appraisal of each item (QAS-99), as follows: reading, instructions, clarity, assumption,

knowledge/memory, sensitivity/bias, response categories, and others. The pre-test was conducted by

creating a cognitive-test protocol including probes for each item. The probes covered the

aforementioned eight steps and comprised the following types: general, specific,

understanding/interpretation, safety, paraphrasing, and memory. The pre-test was performed by

cognitively interviewing (69) four clinical nurse managers individually with concurrent probing

(70), in which verbal probes were immediately asked after a question. This protocol led to revision

of seven questions that were subsequently pre-tested in a group interview.

Data collection, administration, processing, error identification, and electronic data file creation

(63;71), were conducted using the electronic web-based survey system Enalyzer Survey Solution,

Version 2010 (Enalyzer A/S, Copenhagen, Denmark).

The questionnaire was checked for test-retest reliability by asking three clinical nurse managers to

answer the questionnaire a second time after a three-month interval (the respondents were instructed

not to check their previous answers). The estimated mean correlation between all of the first and

second answers was reasonable (Kendall’s Taub [Ken, b] = .75). With regard to the Norwegian

portion of the study, the questionnaire was translated and retranslated by the author and then

checked by two healthcare professionals who were familiar with the subject. A pre-test was

conducted to test face validity through cognitive interviews of seven clinical nurse managers. This

test led to the revision of eight questions, mostly due to imprecise translations. Moreover, one


22

question was added to the questionnaire due to different staff competencies, as psychologists in

Norway are partially responsible for their patients’ treatment, which was not the case in Denmark at

that time.

Later, the outcome variable (i.e., the number of MR episodes) was re-examined. This procedure was

necessary due to confusing wording of a heading in the Danish questionnaire and the lack of precise

numbers regarding the Norwegian patients who experienced MR more than 25 times (14 answer

categories instead of the actual number were provided, causing imprecise exclusion). Re-

examination was accomplished by contacting all of the clinical nurse managers in Denmark and

four of the clinical nurse managers in Norway. The respondent rate was 94%, and the results

lowered the overall frequency of MR by 3.2%. A high correlation coefficient (Ken, b =.89) was

found between the two answers, revealing that the confusing heading and the categorisation error

did not have an important effect.

A description of how the 22 dichotomised MR-preventive factors were measured can be found in

Papers 1 and 2.

As described in the introduction, several conditions might influence how much MR is used. These

conditions include geography (region), type of unit (department), ethnicity other than

Danish/Norwegian, gender, diagnoses, and the small number of patients experiencing many MR

episodes. In the later analyses unit size, patient flow, bed occupancy, type of bed, reason for the

MR, country, and forensic provision, could also possibly have an influence on how often MR is

used. The specific background variables included as covariates in the later analyses and the

dispersion of MR episodes between the categorised background variables in the subsample (data

from the questionnaire survey) can be found within Papers 3 and 4.

Finally, two analytic investigations were conducted, the first, to identify the potential associations

between the number of MR episodes and the MR-preventive factors (Study 3), and the second, to

determine the degree to which the identified MR-preventive factors could explain the different

numbers of MR episodes between Denmark and Norway (Study 4).

The first analytic investigation was conducted by analysing the data from the questionnaire to

identify potential associations between the number of MR and the MR-preventive factors (Study 3).

Data analyses were conducted in four steps. We first examined how the factors were individually

associated with the number of MR episodes (Model 1), and then how the factors were associated

with the number of MR episodes, using a model with all factors included (Model 2). To eliminate


23

multicollinearity and the chance that important effects may hide each other from Model 2, a third

model was constructed by a forward factor-selection procedure, in which the most significant

factors were included stepwise until the corrected quasi-likelihood under the independence model

criterion (QICC) value began to rise, denoting no further improvement for the model (Model 3). All

of these models were adjusted for potentially confounding variables. Additionally, we adjusted for

the nesting of units in departments — 17 departments containing two units each and three

departments containing three units each — by a Generalised Estimating Equations (GEE)

technique. Finally, we analysed the correlations among the factors by Pearson’s correlation

coefficient, adjusted for department, to explore multicollinearity and to justify a reduction of Model

2. In Models 1–3, linear regression was used to assess the associations between the number of MR

episodes and the chosen factors. To stabilize the variance, the outcome was logarithmically

transformed. Hence, the regression coefficients are reported back-transformed (exponentiated;

exp[B]) and must be interpreted multiplicatively. Hence, because the preventive factors were

dichotomized, a regression coefficient exp(B) indicates the average number of times MR increased

for units in which this factor was present compared with units in which this factor was absent. To

account for multiple testing, a p value < .01 was chosen to indicate significance.

The second analytic investigation was conducted by analysing the data from the questionnaire to

determine the degree to which the identified MR-preventive factors could explain the different

number of MR episodes between Denmark and Norway (Study 4). The data analyses were

conducted in five steps. First, the difference in the frequency of MR episodes between the countries,

was analysed. Second, how the factors were individually associated with the number of MR

episodes in Denmark and Norway was separately investigated. Third, we assessed whether an

interaction between MR-preventive factors and the country (second order interaction, SOI) existed.

Fourth, the difference in the prevalence of MR predictive factors between the two countries was

probed using Pearson’s chi-square test (χ2). Fifth, the confounding effects of the MR-preventive

factors, i.e., the difference in the effect of the country in models with and without the corresponding

MR-preventive factor expressed as a percentage (∆ exp[B]); a percentage greater than 10% was

chosen to denote a significant difference. Linear regression was used to estimate the associations

among the frequency of MR episodes, countries, and MR-preventive factors. To stabilise the

variance, the outcome was logarithmically transformed. Hence, the regression coefficients are back-

transformed (exponentiated; exp[B]) and should be interpreted multiplicatively. Because the

preventive factors were dichotomised, the regression coefficient exp(B) indicates the average

number of times the MR increased for the units in which a factor was present compared with the


24

units missing this factor e.g., result of exp(B) = 0.42 should be interpreted as 0.42 times the average

number of MR incidents without the MR-preventive factor or 58% fewer MR episodes that occur

with the MR-preventive factor than would without it. Except for the first analysis, in which the

dichotomised variable was the “country,” the regression coefficient exp(B), indicates the difference

between the countries in the average number of MR episodes. Further, all models, except the fourth

set of analyses (χ2-test), were adjusted for background variables. Adjustments were made for the

nesting of units in the departments (17 departments contained two units each and three departments

contained three units each) using a GEE technique. To account for multiple testing, a p-value < .01

was chosen to indicate significance.

All analyses were performed using IBM SPSS Statistics for Windows, Version 19.0 (released in

2010; Armonk, NY, IBM Corp.).

7 Results

7.1 Summary of Results, Study 1

Mechanical Restraint – Which Interventions Prevent Episodes of Mechanical Restraint? – A

Systematic Review.

This systematic literature review was conducted to identify which conditions in nursing and which

nursing interventions have been shown to reduce the frequency of mechanical restraint episodes.

The search yielded a total of 2,885 papers. After three sifts, a final number of 59 papers were

included in the further review process. Of these, 48 papers were quantitative and 11 papers were

qualitative (a table presenting the appraisal, grading, and combination of the recommendations

regarding the MR-preventive factors can be found in Appendix A).

Upon hierarchically displaying the preventive factors (Table 1), the implementation of cognitive

milieu therapy e.g. through patient involvement and empowerment; combined intervention

programmes e.g. patient participation, patient education, staff education, programmatic changes,

high-level administrative endorsement, cultural changes, and data analysis; and implementation of

patient-centred care with a high degree of patients´ positive involvement in their own care reduces

the number of mechanical restraint episodes. These are among the three highest ranking MR-

preventive factors.


25

Table 1. Hierarchy of Mechanical Restraint Preventive Factors

Rank Preventive factor (reference)

1 Implementation of cognitive milieu therapy reduces the number of mechanical restraint episodes, among other things

through patient involvement and empowerment (72;73)

2 Combined intervention programmes are effective in avoiding mechanical restraints, for example, patient participation,

patient education, staff education, programmatic changes, high-level administrative endorsement, cultural changes, data

analysis, etc. (45;46;74-76)

3 Implementation of patient-centred care with a higher degree of patients´ positive involvement in their own care reduces the

number of mechanical restraint episodes (73;77;78)

4 The identification of the patients’ crisis triggers unique calming techniques and decisions on specific crisis interventions

early in the admission procedure, thereby reducing the number of mechanical restraints (79-82)

5 Introduction of validated risk assessment systems including suggestions of care interventions reduce the number of

mechanical restraints (83-85)

6 Effective treatment of substance abuse and psychosis reduces the number of mechanical restraint episodes; the treatment of

mental illnesses is also described elsewhere in the review (73;77;78;86)

7 Mandatory review from qualified clinical experts, for example, as post-incident review minimizes the number of

mechanical restraint episodes (87;88)

8 Better staff training reduces the number of mechanical restraints (longer duration of the training, more staff participating,

use of different and involving education methods, higher quality of the content and greater diversity of the topics, for

example, legislation, crisis management, physical intervention strategies, antecedents, warning signs, de-escalation

techniques etc.) (20;79;89-96)

9 Staff attitudes such as positive care for the patients, respect, collaboration, communication, humanism, empathy and less

control and involuntary procedures, reduce the number of mechanical restraints (1;81;95;97-102)

10 Higher patient-staff ratio reduces the number of mechanical restraints (103;104)

11 The higher educated staff, the fewer episodes of mechanical restraints (90;93;95;105;106)

12 The more experienced staff, the fewer episodes of mechanical restraints (90;93;95;105;107;108)

13 If nurses change their focus from considering the patient as deviant to being a resource in their own treatment it would

decrease the number of mechanical restraint (109)

14 Early and adequate administration of oral p.n. medication guided by the individual patients´ problems, medication

experiences, and preferences leads to fewer mechanical restraint episodes (80;81;97;110;111)

15 Debriefing, defusing and crisis intervention minimize the number of mechanical restraint episodes (20;79;105;112)

16 Patients´ response to rules can be positive and lead to fewer cases of mechanical restraint. For instance, if staff helps

patients to understand not only the rules, but also the rationale for the presence of such rules. If the staff manages to make

patients understand that the rules are imposed out of concern for their welfare, then patients may be less reluctant to

following these rules (81;100-102;113)

17 Activity-based nursing interventions e.g. physical activities are effective in avoiding mechanical restraints (81;114)

18 Several mental and physical working environmental factors influence positively the number of mechanical restraints, for

example, good management as a high degree of order, organization, or support; adequate staffing, psychiatric training,

rotation, education, no crowding, emergency backup, etc. Several of these factors are described elsewhere in the review

(20;115-118)

19 Less crowding, for example, less stimuli, less environmental stress and more personal space decrease the number of

mechanical restraints (81;119)

20 Higher regular staff and lower substitute ratio reduce the number of mechanical restraints (20;95;120)

21 Staffs´ opposition towards mechanical restraint, due to the perception that mechanical restraint is ethically problematic,

decrease the number of mechanical restraints (121;122)

22 Alarm systems, for example, video surveillance and radio systems and well trained response teams, reduce the number of

mechanical restraints (123)

23 Better care-alliance between patients and the nursing staff reduces the number of mechanical restraints (104)

24 Communication especially communicative de-escalation reduces the number of mechanical restraint episodes (124)

25 Well-maintained and familiar surroundings reduce the number of mechanical restraints (125)

26 Music as diversion and relaxation before and after mechanical restraint reduces the number of mechanical restraint episodes

(126)

27 Separation of acutely disturbed patients reduces the number of mechanical restraint episodes (81)

Note. Rank refers to the number in the hierarchy where 1 is highest/best. p.n. = pro necessitate (as needed).


26

The development of the combined recommendation grading system, which describes a single,

combined grade for the quantitative and qualitative evidence, allowed studied MR-preventive

factors to be displayed within a

matrix of the effect and

recommendation grades, in

which the recommendation

grade and effect were weighted

as equally important. The

recommendation grade

describes the evidence

regarding a specific

intervention level of scientific

“truth” or its potential for

clinical application on a scale

from 1 to 5, for which 1 is the

highest/best. The effect refers

to the reduction in the number

of MR episodes, from 1 to 5, e.g. with 1 representing an 80% to 100% reduction, 2 a 60% to 80%,

and so on (Figure 2).

No MR-preventive factors reached the highest degree of recommendation combined with the

highest level of effect. One MR-preventive factor reached the highest level of effect, with a

decrease in MR use of 87%; however, it received a recommendation grade of only 3. Another factor

had a recommendation grade of 2 but a decrease in MR use of only 27%.

Almost no papers described negative or harmful effects for patients or staff such as increased

numbers of staff or patient injuries or increased use of other types of more intrusive coercive

measures. Of the 59 papers included in this review, one study, by Bowers (89) identified that

attending briefer update courses in physical restraint/holding techniques, triggered small short-term

rises in the rates of physical aggression; however, these results are consistent with a threshold

effect, indicating that once adequate numbers of staff have been trained, further training keeps

incidents rates low. One study by Khadivi (74), detected an increase in assaults on patients and staff

after implementing an intervention that included staff education, early recognition, debriefing, and

post incident review. Future initiatives are suggested to be followed by safety monitoring and staff

training in the management of violent patients.

1

2,3

6,7

4

9,10

8

5

27

14,15,16,17,

18,19,20,21,

22,23,24,25,

26

11,12,13

Eff

ect

Recommendation grade

5 4 3 2 1

1

2

3

4

5

Figure 2. Effect and Recommendation Grade Matrix

Figure 2. The numbers in the matrix refer to a specific factor (the fist number in the hierarchy of the mechanical restraint

preventive factors). Recommendation grade describes the evidence regarding a specific intervention level of scientific “truth”

or its potential for clinical application, from 1 to 5, where 1 is the highest/best. Effect refers to the reduction in the number of

mechanical restraints, from 1 to 5, with 1 being the best, from 80% to 100% reduction, 2 from 60% to 80%, etc.


27

Corrigendum: Unfortunately, a few minor reference errors were identified after publication; thus, a

corrigendum was published (Paper 1, at the end).


Coercion within Danish Psychiatry Compared with 10 Other European Countries.

This investigation was conducted to compare the use of coercive measures on psychiatric inpatients

between Denmark and Sweden, Norway, Finland, Iceland, Belgium, The Netherlands, United

Kingdom, Ireland, France, and Italy, to share knowledge and learn about what other countries do to

minimise coercion.

All members of UEMS responded to the questionnaire, but only a few members of EPSO provided

answers.

Most of the background data (Table 2) and the physical coercion data (Table 4) were retrieved from

the literature review and collected from several sources, including national registers,

governmental/specialist reports, and research papers (55;57;59-61;66;67;127-133). Data on legally

applied coercive measures (Table 3) and the perception of intrusiveness (Table 5) were retrieved

from the questionnaire.

For the purpose of this thesis, Norway was included in the later described research (Studies 3 and

4), and the results from Denmark and Norway are shown.

Table 2 shows the background data from Denmark and Norway.

Table 2. Background Data

Denmark Norway

Psychiatric beds per 100,000 inhabitants (55) (2007) 61 119

Number of forced admissions per 100,000 inhabitants (56;59;131) (1999-2001) 70 135

Mean length of stay in days (55;57) (1998 and 2007) 30 29

Forensic psychiatric beds per 100,000 inhabitants (60;61) (2002-2007) 6.6 0

Psychiatric beds in jails per 100,000 inhabitants (60;61) (2002–2007) 0 0

Year of last revision of the psychiatric code (Questionnaire survey - 2009) 2006 2006

National register for coercion (Questionnaire survey - 2009) X X

Psychiatrists per 100,000 inhabitants (55) (2007) 11 16

Nurses working in Mental Health Care per 100,000 inhabitants (55;132) (2007) 64 N.a. N.a.: No available information. Year: The available data were recorded from 1998 to 2008. The numbers of forced admissions: Denmark: 2000,

Norway: 2001. Mean length of stay was in Denmark from 2007 and Norway 1998. Mean length of stay: The median number of days in mental

hospitals and community-based psychiatric inpatient units. Denmark: The number of forced admissions includes forced compulsory retention of voluntarily admitted patients; the number on forced admissions alone would be 34. “Nurses working in Mental Health Care” refers to all nurses, the

number of specialized nurses was 9 (55;132). The length of stay is only from community-based psychiatric inpatient units.

The number of psychiatric beds was almost twice as high per 100,000 inhabitants in Norway (61,

119). The number of forced admissions was also almost twice as high per 100,000 inhabitants (70,

135). The mean length of stay in days was the same (30, 29). Norway had no specialised psychiatric

forensic beds (6.0, 0), and neither Denmark nor Norway had psychiatric beds in jails. The last


28

revision of the psychiatric code was the same (2006) in 2009, later revisions has since been

published. Both Denmark and Norway had a national register on coercion, although Norway only

collected complete data some years. Denmark had fewer psychiatrists than Norway per 100,000

inhabitants (11, 16). The number of nurses working in mental health care in Norway was

unavailable.

Table 3 shows the applied methods of coercion from the two countries.

Table 3. Applied Methods of Coercion

Denmark Norway

Forced medication/long period X X

Forced medication/short period X X

Mechanical restraint X X

Seclusion/isolation X

Physical restraint/holding X X

Time out

Constant observation X X

Are other types of coercion used X X Blank: Coercive measures were not in use, allowed, or perceived as a coercive measure in the country. Denmark: If patients complained about the use

of forced medication, it would give 2-3 weeks delayed effect before medication could be initiated. During mechanical restraint, constant observation

was required. Seclusion had only been allowed since 2005 and only in a particularly high security department for extraordinarily dangerous psychiatric patients. Physical restraint was only used few minutes in connection with other coercive measures. Time out was not regarded as coercion.

Constant observation as a coercive measure was first reported if it was prolonged more than 24 hours. Norway: Seclusion was only allowed in acute

situations and was not to be used if patients were under 16 years. Time out was not regarded as coercion. Constant observation was first regarded as a coercive measure after 12 to 24 hours.

Denmark and Norway applied the same methods of coercion, except for seclusion, which was only

allowed in a special high security department in Denmark.

Table 4 shows the usage of the three most frequent physical coercive measures: MR, seclusion, and

physical restraint/holding (physical restraint/holding not as part of MR or seclusion, but as the only

coercive measure).

Table 4. Physical Coercion per 100,000 Inhabitants in Denmark and Norway

Denmark Norway

Mechanical restraint 34 21

Seclusion/isolation 0 2

Physical restraint/holding 18 6

Denmark: 2007. The numbers included: geronto-, child/adolescent, and forensic psychiatry (66). In Denmark, it was possible for the patients to be subjected to voluntary mechanical restraint, which occurred in 10% of all incidents. Norway: 2007. The numbers included: geronto-, child/adolescent,

and forensic psychiatry (67).

In 2007, Denmark used more MR and physical restraint/holding than Norway (34, 21 and 18, 6).

Norway used more seclusion/isolation than Denmark, but no numbers from the Danish high

security department were available.

Table 5 shows the ranking of the coercive measures with respect to perceived intrusion in the two

countries.


29

Table 5. Ranking of Coercive Measures With Respect to Perceived Intrusion

Denmark Norway

Forced medication/long period 5 8

Forced medication/short period 4 7

Mechanical restraint 7 6

Seclusion/isolation 6 4

Holding/physical restraint 3 3

Time out 1 1

Constant observation 2 2

Mechanically restraint ambulatory 8 5 Forced medication/short period: Was defined as medicine given against the patient’s will in an emergency (not medical treatment for a longer time

according to treatment plan). Forced medication/long period: Was defined as medicine given against the patient’s will, as treatment (medical treatment for a longer time according to treatment plan). Mechanically restraint ambulatory: Was defined as the use of restraint devices, which allow

patients to be out of bed and to walk around.

The degrees of intrusion for the coercive measures were perceived differently in the two countries,

In Denmark ambulatory MR, MR, and seclusion/isolation are perceived as the most intrusive

measures, whereas forced medications were perceived as the most intrusive in Norway.


Mechanical Restraint in Psychiatry: Preventive Factors in Theory and Practice. A Danish–

Norwegian Association Study.

In this study, a cross-sectional questionnaire was used to identify how the presumed MR-preventive

factors of non-medical origin were associated with the frequency of MR episodes in Denmark and

Norway.

The responding and non-responding clinical nursing managers in the units were equally distributed

across the countries (overall response rate = 49%). Data from 43 and 47 units were included in

Denmark and Norway receptively. The reasons for non-response were indicated by the clinical

nurse managers as “change of clinical nurse manager” (n = 9), “lack of personal resources to

accomplish the task” (n = 7), sick leave (n = 5), inability to remember/identify the number of MR

episodes (n = 5), and lack of belief that this type of research was beneficial (n = 1). The remaining

63 nurse managers did not provide a reason.

The respondents reported 2,361 MR episodes in Denmark and 1,017 MR episodes in Norway

during the examined one-year time period (2008 in Denmark and 2010 in Norway); nine units

accounted for 451 of the excluded MR episodes (> 25 per individual), leaving 1,938 episodes in

Denmark, and 989 in Norway for analysis.

Table 6 shows the results from the analyses.


30

Table 6. Effects of Potential Mechanical Restraint Preventive Factors on the Annual Number of Mechanical Restraint Episodes

Note. MR = Mechanical restraint. The exp(B) parameters were estimated using a linear regression, and all three models were adjusted for the background variables: geography, unit size, bed occupancy, type of bed, reason for the MR,

forensic provision, ethnicity, and type of unit. Additionally, the nesting of units in departments was adjusted using GEE techniques. In Model 1, the MR-preventive factors were analysed alone. In Model 2, all MR-preventive factors were

analysed together. In Model 3, a forward-selection procedure included the most significant MR-preventive factors until the Corrected Quasi Likelihood under Independence Model Criterion (QICC) value began to rise. In all three models, the outcome variable (i.e., the number of MR episodes) was log-transformed. Hence, the parameters should be interpreted multiplicatively (e.g., in Model 1, cognitive milieu therapy results in exp(B) = 0.75 times the average number of MR

incidents without cognitive milieu therapy (or 25% fewer MR episodes occur with cognitive milieu therapy compared with without therapy). p < .01. The grey marking denotes the most important values.

Model 1 Model 2 Model 3

MR-preventive factors exp(B) 95% CIs of

exp(B) p

exp(B)

95% CIs of

exp(B) p

exp(B)

95% CIs of

exp(B) p

Cognitive milieu therapy vs. no cognitive milieu therapy 0.75 [0.26, 2.07] .57 0.67 [0.37, 1.18] .17 . . .

Patient involvement vs. no patient involvement 0.51 [0.29, 0.88] .02 0.14 [0.07, 0.30] .00 0.42 [0.25, 0.73] .00

Patient-centred care vs. no patient-centred care 0.79 [0.30, 2.06] .62 0.42 [0.17, 1.01] .05 . . .

Identification of the patient’s crisis triggers vs. no identification 0.73 [0.46, 1.15] .17 0.45 [0.29, 0.72] .00 . . .

Risk assessment vs. no risk assessment 2.01 [0.84, 4.81] .12 2.91 [1.28, 6.60] .01 . . .

Mandatory review vs. no mandatory review 0.41 [0.23, 0.73] .00 0.27 [0.18, 0.39] .00

0.36 [0.21, 0.63] .00

Staff training (> 50% vs. less) 1.08 [0.69, 1.69] .74 3.60 [2.49, 5.20] .00 . . .

Positive staff attitudes vs. less positive 1.15 [0.61, 2.16] .67 1.15 [0.58, 2.26] .69 . . .

Patient-staff ratio (> 3 staff per patient vs. less) 0.72 [0.42, 1.24] .24 1.36 [0.83, 2.21] .22 . . .

Staff education (> 3 years vs. less) 1.16 [0.39, 3.41] .79 0.14 [0.06, 0.32] .00 . . .

Staff experience (> 5 years vs. less) 0.76 [0.47, 1.23] .27 0.40 [0.24, 0.67] .00 . . .

Debriefing, defusing vs. none 1.20 [0.81, 1.70] .36 0.85 [0.59, 1.22] .39 . . .

Positive patient responses to rules vs. negative responses 0.82 [0.46, 1.47] .50 0.34 [0.19, 0.62] .00 . . .

Activity-based nursing vs. no activity-based nursing 0.93 [0.51, 1.70] .82 1.65 [1.08, 2.51] .02 . . .

Acceptable work environment vs. poor work environment 0.54 [0.33, 0.90] .02 1.09 [0.76, 1.58] .64 . . .

No crowding vs. crowding 0.56 [0.34, 0.92] .02 0.50 [0.31, 0.80] .00 0.54 [0.36, 0.81] .00

Substitute staff (< 50 vs. more) 0.82 [0.51, 1.32] .42 0.57 [0.33, 1.01] .05 . . .

Staff opposition to MR (many vs. some) 0.85 [0.54, 1.34] .49 1.12 [0.77, 1.62] .56 . . .

Alarm systems (many and well-trained responses vs. less well-trained responses) 1.12 [0.42, 3.00] .82 2.31 [1.15, 4.63] .02 . . .

Communication (all staff have communicative skills vs. less than all staff) 1.08 [0.66, 1.75] .77 1.53 [0.84, 2.81] .17 . . .

Well-maintained, clean and tidy surroundings (always vs. seldom) 0.69 [0.41, 1.17] .17 1.08 [0.71, 1.66] .72 . . .

Separation of acutely disturbed patients vs. no separation 1.55 [0.95, 2.52] .08 1.14 [0.73, 1.78] .57 1.65 [1.10, 2.49] .02


31

Table 6 indicates that in Model 1, which evaluated the associations separately, only one MR-

preventive factor was significantly associated with a low frequency of MR episodes: mandatory

review (exp[B] = .41, which is interpreted as a 1-.41 = 59% decrease in MR episodes).

In Model 2, in which the associations were evaluated jointly, seven MR-preventive factors were

significantly associated with a low frequency of MR episodes: patient involvement (exp[B] = .14),

staff education (exp[B] = .14); mandatory review (exp[B] = .27); positive patient responses to rules

(exp[B] = .34); staff experience (exp[B] = .40); identification of the patient’s crisis triggers (exp[B]

= .45); and crowding (exp[B] = .50). Staff training was the only MR-preventive factor that was

significantly associated with a high frequency of MR episodes (exp[B] = 3.60).

In Model 3, in which we attempted to strike a balance between residual confounding (Model 1) and

over-fitting (Model 2), three MR-preventive factors were significantly associated with a low

frequency of MR episodes: mandatory review (exp[B] = .36); patient involvement (exp[B] = .42);

and crowding (exp[B] = .54).

Pearson’s correlation matrix identified one highly correlated (Pearson’s r > .50) pair of MR-

preventive factors: communication and positive staff attitudes (Pearson’s r = .77). Furthermore, the

following six pairs exhibited a medium correlation (Pearson’s r .30 ≤ .50): patient-centred care and

cognitive milieu therapy (Pearson’s r = .49); communication and staff opposition to MR (Pearson’s

r = .38); substitute staff and activity-based nursing (Pearson’s r = .36); communication and risk

assessment (Pearson’s r = .32); positive patient responses to rules and identification of the patient’s

crisis triggers (Pearson’s r = .31); and staff education and cognitive milieu therapy (Pearson’s r =

.31). All correlations were significant (p < .01).


Comparing the effect of non-medical mechanical restraint preventive factors between

psychiatric units in Denmark and Norway.

The data acquired from the cross-sectional questionnaire (the same as in Study 3) was analysed to

examine how presumed MR-preventive factors of non-medical origin may explain the differing

numbers of MR episodes between Denmark and Norway.

The responding and non-responding clinical nursing managers of the units were equally distributed

across the countries (overall response rate = 49%). Data from 43 and 47 units were included in

Denmark and Norway respectively. The reasons for non-response were indicated by the clinical

nurse managers as “change of clinical nurse manager” (n = 9), “lack of personal resources to


32

accomplish the task” (n = 7), sick leave (n = 5), inability to remember/identify the number of MR

episodes (n = 5), and lack of belief that this type of research was beneficial (n = 1). The remaining

63 nurse managers did not provide a reason.

The respondents reported 2,361 MR episodes in Denmark and 1,017 MR episodes in Norway

during the examined one-year period (2008 in Denmark and 2010 in Norway); nine units accounted

for 451 of the excluded MR episodes (> 25 per individual), leaving 1,938 episodes in Denmark, and

989 in Norway for analysis.

A difference in the number of MR episodes was observed between the countries in the model that

included the background variables (exp[B] = 1.92, p = .01); Denmark had 1.92 times as many (92%

more) MR episodes than Norway.

Table 7 shows the results of the analyses.


33

Table 7. Differences in the Effects of Potential Mechanical Restraint Preventive Factors on the Annual Number of Mechanical Restraint Episodes in Denmark and Norway

Note. MR = Mechanical restraint. No = the number of units where the factor was present. % = the percentage of units where the factor was present. Exp(B) = the parameter estimate. The outcome variable (i.e., the number of MR episodes) was

log-transformed. Hence, the exp(B) should be interpreted multiplicatively (e.g., in Denmark, cognitive milieu therapy results in exp[B] = 0.36 times the average number of MR incidents without cognitive milieu therapy (or 64% fewer MR

episodes occur with cognitive milieu therapy compared without therapy). 95% CI of exp(B) = 95% Wald confidence interval for exp(B). p = p-value. SOI = second order interaction, p-value of a test for the interaction between the MR-

preventive factor and the country, i.e., whether the effect of the MR-preventive factor differs between the countries. χ2 = p-value of a chi square test for the difference in the prevalence of the MR-preventive factor in the units between the countries. Δ exp(B) = Differences in the effects of the country in models with and without the corresponding MR factor (confounding effect of the MR-preventive factor) expressed as a percentage; negative Δ exp(B) values should be

interpreted as a MR-preventive factor that minimises the difference between the countries. In the estimation of exp(B), SOI, and Δ exp(B), we used linear regression and adjusted for the following background variables: unit size, bed

occupancy, type of bed, reason for the MR, forensic provision, ethnicity, and type of unit. Additionally, the nesting of units in departments was adjusted using GEE techniques. = p < .01. # = Δ exp(B) > +/- 10%. The grey marking denotes the most important values.

Denmark Norway

SOI

Interaction between

the MR-preventive

factor and the country

χ2

Difference in the

prevalence of the

MR predictive factor between countries

Δ exp(B)

Difference in the

effect of country

MR-preventive factors No % exp(B) 95% CIs

of exp(B) p No % exp(B)

95% CI

of exp(B) p p p %

Cognitive milieu therapy vs. no cognitive milieu therapy 5 13 0.36 [0.11, 1.16] .09 1 3 7.46 [2.94, 18.89] .00 .00 .09 6

Patient involvement vs. no patient involvement 34 81 0.66 [0.26, 1.71] .40 37 86 0.60 [0.27, 1.31] .20 .80 .53 -2

Patient-centred care vs. no patient-centred care 5 13 0.41 [0.12, 1.38] .15 4 10 5.01 [2.03, 12.36] .00 .00 .72 3

Identification of the patient’s crisis triggers vs. no identification 29 69 0.65 [0.38, 1.10] .11 40 93 0.75 [0.26, 2.17] .59 .67 .01 -10#

Risk assessment vs. almost none or no risk assessment 9 21 2.85 [1.28, 6.36] .01 17 37 1.39 [0.60, 3.26] .45 .13 .11 7

Mandatory review vs. no mandatory review 8 19 0.38 [0.15, 0.97] .04 7 15 0.53 [0.21, 1.37] .19 .97 .60 -1

Staff training (> 50% vs. less) 20 49 1.26 [0.65, 2.46] .49 21 48 0.88 [0.39, 2.01] .76 .22 .92 0

Positive staff attitudes vs. less positive 4 10 1.30 [0.56, 2.99] .54 13 28 1.28 [0.53, 3.06] .59 .75 .03 3

Patient-staff ratio (> 3 staff per patient vs. less) 2 5 0.96 [0.45, 2.06] .91 25 53 0.62 [0.32, 1.19] .15 .56 .00 -11#

Staff education (> 3.5 years vs. less) 3 7 0.34 [0.17, 0.66] .00 43 93 0.76 [0.28, 2.04] .59 .05 .00 -51#

Staff experience (> 5 years vs. less) 36 84 0.81 [0.45, 1.44] .47 44 98 0.56 [0.14, 2.19] .41 .84 .02 -1

Debriefing, defusing vs. none 28 72 0.79 [0.46, 1.36] .40 24 65 2.04 [1.07, 3.89] .03 .02 .52 -3

Positive patient responses to rules vs. negative responses 32 78 0.61 [0.29, 1.29] .20 34 85 1.92 [0.49, 7.44] .35 .80 .42 -1

Activity-based nursing vs. no activity-based nursing 15 38 0.78 [0.34, 1.79] .56 22 54 0.69 [0.25, 1.88] .47 .53 .14 -6

Acceptable work environment vs. poor work environment 6 14 0.52 [0.22, 1.25] .14 17 40 0.55 [0.30, 1.03] .06 .93 .01 -15#

No crowding vs. crowding 19 51 0.59 [0.31, 1.13] .11 20 59 0.55 [0.27, 1.14] .11 .22 .53 6

Substitute staff (< 50 vs. more) 8 19 0.59 [0.32, 1.10] .10 33 73 0.54 [0.25, 1.18] .12 .33 .00 -17#

Staff opposition to MR (many vs. some) 11 27 1.09 [0.47, 2.54] .84 17 39 0.70 [0.30, 1.63] .41 .65 .25 -1

Alarm systems (many and well-trained responses vs. less well-trained

responses) 3 7 0.18 [0.03, 1.17] .07

5 11 3.72 [2.16, 6.41] .00

.00 .50 4

Communication (all staff possess communicative skills vs. less than all staff) 8 20 1.56 [0.85, 2.85] .15 17 36 1.03 [0.51, 2.10] .93 .43 .08 3

Well-maintained, clean and tidy surroundings (always vs. seldom) 23 55 0.89 [0.42, 1.87] .75 29 63 1.03 [0.53, 2.04] .92 .63 .43 0

Separation of acutely disturbed patients vs. no separation 21 50 1.05 [0.46, 2.40] .90 38 83 1.94 [0.92, 4.09] .08 .28 .00 13#


34

Table 7 indicates the extent to which the MR-preventive factors explain the differences in MR

episodes between Denmark and Norway in several steps. The first column of results in Table 7, in

which the MR-preventive factors are evaluated in a similar fashion as in Table 6 (Model 1),

indicates that a single MR-preventive factor, staff education (exp[B] = 0.34, p = .00), was

significantly associated with a lower frequency of MR episodes in Denmark. In Norway, three MR-

preventive factors were significantly associated with a higher frequency of MR episodes: cognitive

milieu therapy (exp[B] = 7.46, p = .00); patient-centred care (exp[B] = 5.01, p = .00); and alarm

systems (exp[B] = 3.72, p = .00). Notably, the confidence intervals for cognitive milieu therapy and

patient-centred care were large, reducing the precision of any estimates of an increase in the MR

episodes related to these factors.

Various effects noticeably differed between the countries. In the column marked SOI in Table 7, we

identified three incidences of an interaction between the MR-preventive factor and the countries:

cognitive milieu therapy (SOI, p = .00); patient-centred care (SOI, p = .01); and alarm systems

(SOI, p = .00). The presence of these factors exhibited a positive effect in Denmark (exp[B]: 0.36,

0.41, and 0.18) and a negative effect in Norway (exp[B]: 7.46, 5.01, and 3.72).

As shown in the column marked χ2 in Table 7, the prevalence was significantly different for six

MR-preventive factors between the two countries: the identification of a patient’s crisis triggers (χ2,

p = .01); patient-staff ratio (χ2, p = .00); staff education (χ

2, p = .00); work environment (χ

2, p =

.01); substitute staff (χ2, p = .00); and separation of acutely disturbed patients (χ

2, p = .00). This

represents a criterion of the factors to represent a confounder for the association between country

and MR episodes.

In the last column of Table 7, six MR-preventive factors were identified, and confounding effects

were uncovered for the association between the country and the number of MR episodes, i.e., the

MR-preventive factor explained part of the difference in the number of MR episodes between

Denmark and Norway. A factor was identified as a confounder if the adjustments for this factor

changed the effects of the country by more than 10% (∆ exp[B] > 10%), and the identified factors

were as follows: identification of the patients’ crisis triggers (∆ exp[B] = -10%); patient-staff ratio

(∆ exp[B] = -11%); staff education (∆ exp[B] = -51%); work environment (∆ exp[B] = -15%);

substitute staff (∆ exp[B] = -17%); and separation of acutely disturbed patients (∆ exp[B] = 13%).

In both countries, the following MR-preventive factors were associated with lower numbers of MR

episodes: identification of patients’ crisis triggers (Denmark, exp[B] = 0.65; Norway, exp[B] =

0.75); the patient-staff ratio (Denmark, exp[B] = 0.96; Norway, exp[B] = 0.62); staff education


35

(Denmark, exp[B] = 0.34; Norway, exp[B] = 0.76); work environment (Denmark, exp[B] = 0.52;

Norway, exp[B] = 0.55); and substitute staff (Denmark, exp[B] = 0.59; Norway, exp[B] = 0.54). In

contrast, the separation of acutely disturbed patients was associated with increased MR episodes

(Denmark, exp[B] = 1.05; Norway, exp[B] = 1.94).

Corrigendum: Unfortunately, a minor error in the abstract of the paper were identified, the one MR-

preventive factor not supported by earlier research was, “separation of acutely disturbed patients”

and not as written “the identification of the patient’s crisis triggers”.

8 Approval and Ethics

Literature reviews such as those in Studies 1 and 2, do not require approval from the scientific

ethical committee system (134). Questionnaire surveys such as those included in Studies 2, 3, and 4

do not require approval from the scientific ethical committee systems of the included European

countries if, respondents are able to provide informed consent or if sensitive information is not

collected. The Danish Data Protection Agency (Journal number: 2007-2058-0015, PSV-2009-2013)

approved the studies and we received permission from the Danish Health and Medicines Authorities

and the Danish Psychiatric Central Research Register to use the register data on the frequency of

MR in Denmark (31) (Journal number: 7-505-29-1246/1). The studies followed the Ethical

Guidelines for Nursing Research in the Nordic Countries (135) and the White Paper Regarding a

Draft Recommendation on Legal Protection of Persons Suffering From Mental Disorder Especially

Those Placed as Involuntary Patients (136).

9 Discussion

9.1 Discussion of Individual Studies

Study 1: Cognitive milieu therapy combined with patient involvement was highest ranked MR-

preventive factor. To the best of our knowledge, this result has not been identified in prior reviews.

One possible explanation for this could be that the cognitive milieu therapy paper (72) was not

published in English; the abstract of this study only described a significant effect and did not quote

an exact figure such as a staggering 87% reduction. Later research by Chang (137) supported this

finding; this before-and-after study showed a reduction in seclusion and restraint episodes following

the implementation of a cognitive therapy training program. Cognitive milieu therapy is an active,

structured, problem-orientated, psycho-educational, and dynamic type of care. Cognitive care


36

models in this environment give the patient the opportunity to implement new alternatives to

inappropriate reaction patterns while also developing new skills (138). One of the core issues in

cognitive milieu therapy is a schema-focused dialog between the patient and the nursing staff during

situations just before potential MR use. In this moment, the patient has the opportunity to try new

actions. This type of patient involvement has also been addressed elsewhere (73) as having a

positive influence on the incidence of MR. A later paper by Soininen (139) agreed, concluding that

patients´ opinions required greater attention in treatment decisions. Several studies have examined

the effect of implementing the “six core strategies” (103;140-147), one of which touts patient

involvement. The reason for the present high recommendation grade is the combination of cognitive

milieu therapy and patient involvement, because, cognitive milieu therapy had a high effect and

patient involvement had high evidence. Synthesising these together adding evidence to cognitive

milieu therapy is possible the correct deduction, because patient involvement is part of the care

model, all though patient involvement could as well be a phenomenon on its own.

The second highest ranked MR-preventive factor was not one particular intervention but rather

programs including several interventions such as patient participation, patient education, staff

education, programmatic changes, high-level administrative endorsement, cultural changes, and

data analysis. Several papers have described different programs (45;46;74-76), indicating an

average decrease in the number of MR episodes of 76%. The evidence mostly comprises “before-

and-after” papers; because some of these papers are well founded, the recommendation grade is the

same as in the first MR-preventive factor. Later research describing combined interventions with a

higher evidence grade (145) could move this MR-preventive factor up in the hierarchy.

The third highest ranked MR-preventive factor was the implementation of patient-centred care

combined with patient involvement. Here, implementation of the Tidal Model reduced the number

of mechanical restraints by 68% (77;78). The Tidal Model is a philosophical approach to mental

health discovery that focuses on helping people who have experienced a metaphorical “breakdown”

to recover their lives as fully as possible by reclaiming their personal stories of distress and

difficulty (148). One of the core elements in the Tidal Model is patient empowerment.

Empowerment is only possible if the patients are involved in their own care. Elsewhere, patient

involvement has also been found (73) to positively influence the occurrence of MR. Thus, the

recommendation grade is the same as in the second MR-preventive factor. As in the first MR-

preventive factor, the reason for the high recommendation grade was the combination of patient-

centred care and patient involvement, where patient-centred care had a high effect and patient

involvement had high evidence. Synthesising these together adding evidence to the Tidal Model is


37

possible the correct deduction, because patient involvement is part of the care model, all though

patient involvement could as well be a phenomenon on its own.

The MR-preventive factor with the highest recommendation grade was no. 5, the introduction of

validated risk assessment systems. This has later been supported by one additional cluster RCT by

van de Sande (149), investigating implementation of crisis monitoring instruments (Kennedy Axis

V and Brøset Violence Checklist [BVC]), the result indicated an insignificant difference between

the experimental wards and the control wards of -15% in the risk ratio of seclusion incidents (15%

lower risk after the implementation). The difference in the duration of seclusion was significant,

with a risk ratio = -45%. These findings added further evidence to this MR-preventive factor,

although, it did not change its place in the hierarchy, based on the risk of moderately occurring bias

in the study (not blinded and small clusters) and the continuously low effect.

When introducing MR-preventive initiatives, only two of the 59 papers described adverse effects

(negative consequences), concluding that adverse effects very seldom arises. Never the less,

following measures of important adverse effects should always be addressed, when introducing new

MR-preventive initiatives.

The evidence supporting the different MR-preventive factors has improved since April 2009 (when

study inclusion in the systematic review ended). Some of these factors were included in the former

discussion of the individual MR-preventive factors; these factors will continue to improve over the

next few years. A superficial update in December 2014 revealed 64 new papers. Thus, few MR-

preventive factors are closing in on the highest recommendation grade, although no MR-preventive

factor has reached the highest recommendation grade.

Study 2: This study was not planned to be as an analytic study but took on a more descriptive

nature, enabling statistical analyses of the differences among the European countries. For the

purpose of this thesis, the data regarding Denmark and Norway are discussed.

The number of psychiatric beds was almost twice as high per 100,000 inhabitants in Norway (61,

119) as in Denmark. This trend could be due to the actual numbers or to way the numbers were

extracted or recorded by the WHO (55). Investigating later data from the countries central statistical

population registers and patient registers, the numbers of mental health hospital beds per 100,000

inhabitants were Denmark, 23.96 and Norway, 22.81, pertaining to the years of 2008 and 2010 (the

years of the later questionnaire survey). So the data recorded by the WHO, did not seem accurate.

The number of forced admissions was almost twice as high per 100,000 inhabitants; this finding

could indicate that patients in Norway were admitted earlier, before they became as aggressive or


38

suicidal as patients in Denmark. Alternatively, Denmark has many more forensic psychiatric

patients (in 2010, 2,638 forensic psychiatric patients (150)) compared with Norway’s maximum of

approximately 500 (the exact number is unknown); although they were unwillingly placed in a

psychiatric unit, these patients were not included in the numbers of forced admission .

In Norway, there are no specialised psychiatric forensic beds. Regional security departments admit

the most aggressive or agitated patients. In Denmark, many forensic patients are placed in ordinary

beds, and only the most severe patients are placed in the special forensic units.

Denmark had fewer psychiatrists than Norway per 100,000 inhabitants, which could indicate that

treatment in Norway is more available and of higher quality.

In 2007, Denmark used more MR and physical restraint/holding than Norway. This finding is not

easily explained due to the many similarities in legislation (64;65) and treatment culture (10)

between the two countries. Norway used more seclusion/isolation than Denmark, but no numbers

from the Danish high security department were available to further qualify this difference. Overall,

the numbers are small and do not sufficiently explain the differences in the MR and physical

restraint/holding figures. Regarding the validity of these figures, in both countries registration of

coercive measures are compulsory by law. The figures from Norway were scrutinised by Bremmes

(67), who collected complete figures from all institutions using coercive measures. Knutzen (151)

have validated 98 randomly selected reasons for restraint, finding kappa values between 0.71 and

0.94 (substantial agreement to almost perfect agreement), the scrutinised protocols were obtained

from three psychiatric wards. Although, the Danish figures have not been scientifically validated,

the nursing heads of the units answered affirmatively when asked whether all coercive measures

had been reported. Furthermore, upon comparing the individual numbers of written records from

seven randomly selected units with the national register, the highest difference found was less than

3%. Therefore, the differences between the figures are not likely due to imprecise registration

practices. A possible explanation could be different medication practises; this does not apply for

forced medication/short period, which showed the same tendencies as MR and physical

restraint/holding (in 2009, 107 and 39 episodes per 100,000 inhabitants were recorded in Denmark

and Norway, respectively (47;152)), but perhaps for forced medication/long period. In 2013, the

number of forced medications was 15.7 and 24.0 per 100,000 inhabitants in Denmark (26) and

Norway (153), respectively; some of the differences in MR and physical restraint/holding numbers

could perhaps be explained by earlier medication use and the inclusion of more patients in a longer

medication period; this does not likely explain much of the difference because Denmark more


39

frequently uses forced medication/short period. The last difference regarding medication use could

involve differences in the overall amount or the number of doses of medicine, which are indicative

of different treatment cultures. The only study found that described this condition was

“Comparisons of psychotropic drug prescribing patterns in acute psychiatric wards across Europe”

(154), which did not describe the actual relationship between coercion and medication but rather

described medication practises in general. This study did not show country differences but did show

department differences. Thus, the lower frequency of MR and physical restraint/holding in Norway

likely indicates that other factors are “responsible” for the differences.

The degrees of perceived intrusion of the coercive measures were perceived differently in the two

countries. In Denmark, ambulatory MR, MR, and seclusion/isolation are perceived to be the most

intrusive, whereas in Norway, forced medications are perceived as the most intrusive. In particular,

the difference regarding forced medication/short period is described within the Norwegian

legislation in the following order: MR, isolation, forced medication (isolated episode), and finally,

physical restraint (lasting for a short period of time) (65). This order is the same as in the Danish

legislation, except for isolation (which is only used in a special high security department) (64).

Normally, the most intrusive intervention would be listed first; thus, a misunderstanding on behalf

of the Norwegian representative of the concepts “short period” vs. “isolated episode” may have

occurred, thereby increasing the similarities between the countries.

Study 3: Mandatory review was significant in all three models; correspondingly, no significant

inter-correlations with the other factors were identified. Thus, no other factors greatly confounded

the association between mandatory review and the number of MR episodes. Mandatory review

should be interpreted as follows: the units in which the staff audited, reviewed or conducted a

similar systematic follow-up period after each MR episode (to analyse and learn from the episodes)

experienced an average of .36 times as many (or 64% fewer) MR episodes compared with units in

which a review was conducted only after some episodes or the unit did not conduct a review

process. This finding is in line with previous research, which also reported that conducting

mandatory reviews of MR episodes was associated with a lower number of MR episodes

(87;88;155-157). Previous research has suggested that an optimal process should review all MR

episodes within a few days and the review board should comprise an interdisciplinary team,

including a nurse specialist, a physician, a psychologist, and the clinical department heads with the

supervisory authority to change nursing and treatment practices. Additionally, the review process

should require a protocol that includes a description of the episode, the environmental conditions

that might have provoked the problem behaviour, possible “crisis triggers,” any unique calming


40

techniques identified, patient involvement, medications, and any behavioural treatments and milieu

therapies, among other items. Later research also supports this, several studies examining the effect

of implementing “the six core strategies” (103;140-147), wherein one of the strategies deals with

mandatory review (debriefings), a RCT found a difference between the control wards and the

intervention wards of 62%, other studies found an even higher effect. Overall, present research of

reasonable quality supports mandatory review.

Patient involvement was significant in both Models 2 and 3; the fact that patient involvement was

not significant in Model 1 could be explained by the correlations with other factors (albeit not

strongly or significantly), e.g., staff training and well-maintained, clean and tidy surroundings. The

elements of patient involvement are potentially included in some staff training, and patients who are

more involved in their own care may have a greater sense of responsibility towards their

surroundings. Thus, the non-significant findings presented in the correlation matrix with regard to

patient involvement do not precisely explain which factors influence the association between patient

involvement and the number of MR episodes. Patient involvement should be interpreted as follows:

units in which the patients participated in ward rounds/conferences, influenced the rules applied to

the unit, or influenced what was written in the nurse’s journal experienced an average of .42 times

as many (or 58% fewer) MR episodes compared with the units in which no such participation

occured. Previous research demonstrated that patient involvement might reduce MR episodes

(73;141), and more resent research also supports this idea. Soininen (139) concluded that patients´

opinions require more attention during treatment decisions. In several studies examining the effect

of implementing the “six core strategies” (103;140-147), one of the strategies represent an approach

to address patient involvement, one RCT found a difference between the control wards and the

intervention wards of 62% Subsequent studies found an even higher effect. In this study, patient

involvement was measured using three items (whether the patients participated in ward

rounds/conferences; whether they influenced the rules; and whether they affected what was written

in the nurses’ journals). Patient involvement is a very complex phenomenon and, includes the

concept that the patient is the expert on his or her life and that the staff members are experts with

regard to the patient’s care or treatment. Accordingly actions are negotiated via dialogue. However,

units with a recovery-orientated or empowering approach, that integrate patient involvement with

the delivery of care and treatment schould include one of the aforementioned items (whether the

patients participated in ward rounds/conferences; whether they influenced the rules; and whether

they affected what was written in the nurses’ journals). Overall, patient involvement has been

strongly supported as a MR-preventive factor by other research.


41

Crowding was significant in both Models 2 and 3; one minor but significant inter-correlation

between crowding and the staff-patient ratio was identified (Pearson’s r = .24, p < .05). This inter-

correlation is logical as more staff creates a perception of more crowding. Crowding should be

interpreted as follows: units in which two of three conditions were present experienced an average

of .54 times (or 46% fewer) MR episodes compared with the units where only one or none of the

conditions was present: only one bed in a patient’s room, more than 25 m2 of all-day-accessible

space per patient, and the perception of no crowding. Previous research demonstrated that no

crowding might influence the number of MR episodes (81;119;158;159), but none of the four

studies supports a causal relationship with the number of MR episodes.

One of the MR preventive factors identified in the systematic literature review, namely risk

assessment, has been supported by some of the strongest evidence, including two cluster RCTs. One

study showed a reduction of 27% in coercive measures (83), and the other showed an insignificant

difference of -15% between experimental and control wards in the risk ratio of seclusion incidents

(149). We have not found a significant relationship between risk assessment and MR reduction in

this study due to the combination of risk assessments for self-destructive/suicidal behaviours and

aggressive/violent behaviours. The difference in the risk assessments of self- destructive behaviours

is most likely impossible to measure because both Denmark and Norway have focused on this

factor over the past several years; thus, almost all institutions have implemented this assessment

system. The independent assessment of aggressive/violent behaviour has also become more

common. Fifty-five per cent of the patients in the units were systematically assessed for the risk of

developing aggressive or violent behaviours; however, this assessment system appears to have been

initially introduced in units with high numbers of MR episodes. Another explanation could be that

the effect of implementing risk assessment tools was not large enough to be identified in this study.

Study 4: Three potential MR-preventive factors displayed very different effects in the two

countries; these factors were demonstrated by significant interactions between the MR-preventive

factor (cognitive milieu therapy, patient-centred care, and alarm systems) and the country. Each

factor exhibited a positive effect in Denmark and a negative effect in Norway, but the prevalence

did not differ between the countries. Because of the different effects, the degree of confounding (∆

exp[B]) is impossible to interpret. Although this topic is potentially interesting, we cannot provide

clear reasons as to why these factors exhibited different effects in the two countries. Further

research is required to explore the different effects of these three MR-preventive factors.


42

Six MR-preventive factors exhibited a significant difference between the countries with regard to

the frequency of MR episodes (∆ exp[B]).

First, staff education: Differences were observed in units in which the staff received basic training

for at least three and a half years vs. units in which the staff had less than three and a half years of

training. This factor explains a large portion (∆ exp[B] = -51%) of the difference in MR episodes

between the countries. Staff education has been supported by earlier research. Jannsen (108)

reported that seclusion was used more often in units with less-educated staff in the long-term wards.

Gim (106) found a significant difference between psychiatric-trained and non-psychiatric-trained

staff in identifying and managing the precipitants of violence. In a large 20-year analysis of 1,565

mental health workers, Flannery (160) found that staff who were victims of patients aggression

were less formally educated and less trained. None of the studies were RCTs, but well designed

otherwise, it is reasonable to deduce, that earlier research supports the role of staff education, in the

difference in MR use between the two countries. Higher basic education among the staff probably

leaves them with higher analytic, and action competences, even in situations of crisis, thereby

reducing the number of MR episodes.

Second, substitute staff: Differences were observed in units in which substitute staff was used fewer

than 50 times (8-hour shifts) over the year vs. units in which these employees were used greater

than or equal to 50 times. The factor explains a large proportion (∆ exp[B] = -17%) of the

difference in MR episodes between the countries. The effects of substitute staff were supported by

earlier research. Bonner (20) found that both patients and staff discriminated between permanent

and regular staff. Bowers (120) reported that adverse incidents were more likely to occur in weeks

during which the regular staff was absent. In another study, Bowers (161) noted that increased

levels of aggression were associated with an increased use of bank and agency staff. None of the

studies were RCTs, but all indicated similar results supporting less use of substitute staff. The

substitute staff may lack knowledge of the structures, processes, and patients in the units thereby

creating insecurity and disturbances among the most vulnerable patients in the unit. This could be a

stressing factor for the patients leading to more aggressive or self-destructive behaviours and hence

more MR episodes.

Third, work environment: Differences were found in units in which a workplace environment

assessment was conducted no or only insignificant physical or mental work-environment problems

existed, and the overall work environment was evaluated as acceptable vs. units in which the

workplace environment was not assessed, more work-environment problems existed, and the


43

overall work environment was evaluated as poor. This factor explains a large proportion (∆ exp[B]

= -15%) of the differences in MR episodes between the countries. Assessing the work environment

is somewhat complex due to the mix of leadership, degree of order, organisation, staffing, training,

and crowding, among other factors. De Benedictis (162) reported that several organisational factors,

e.g., team climate, staff perception of aggression and interaction of team-members, were associated

with an increased use of seclusion and restraint. Hanrahan (163), observed a relationship between

psychiatric nurses’ work environments and inpatients’ environments. Van der Schaff (158),

reported a number of building design features that affect seclusion and restraint. Virtanen (159),

indicated that patient overcrowding was associated with violent assaults towards employees.

Huckshorn (143), found that leaders played a critical role in reducing seclusion and restraint.

Camarino (164), found a significant association between work-related factors and certain types of

violence. Some of the other examined factors such as staff training and the staff/patient ratio also

contribute to the overall work environment. Overall, research supporting the association between

MR use and the work environment covers many aspects of a healthy work environment supporting

the role of the work environment in accounting for the difference in MR use between the two

countries. Regardless of which piece creates a positive work environment, the wellbeing of the staff

most certainly affects the wellbeing of the patients. Accordingly, if staff members experience a

more positive work environment, then the patient likely also experiences a more positive

environment.

Fourth, patient-staff ratio: Differences were observed in units in which the staff-to-patient ratio was

higher than or equal to three-to-one on average vs. those units in which this ratio was lower than

three-to-one. In Denmark, the effect was small, possibly due to the small number of units with an

average staff-to-patient ratio higher than three-to-one. This factor explains a large proportion (∆

exp[B] = -11%) of the difference in MR episodes between the countries. The effects of the staff-to-

patient ratio were partially supported by earlier research. Smith (103) found a 66% decrease in

restraint episodes. However, because several factors contributed to the observed effect, the

proportion of the MR reduction that is attributable to the 50% increase in staffing remains unclear.

Jannsen (108) found that seclusion was more commonly used if the number of patients per staff was

greater, in this case in long-term wards. Donat (165) reported a significant association between

increased staffing levels and decreased MR and seclusion numbers. However, Owen (117) noted

that the relative risk of violence increased with more nursing staff. Additionally, Bowers (161)

reported that increased levels of aggressive incidents were associated with increased staffing.

Nonetheless, these studies with opposite outcomes did not precisely target MR or seclusion and


44

instead focused on violence and aggression; thus their results could have been influenced by staff

employment policies, which state that staffing levels are to be increased in units with more

aggressive incidents. Our study most likely eliminated this problem by introducing “type of unit”

and “type of bed” as confounders. Overall, earlier research specifically dealing with the same

conditions supports the finding that the staff-to-patient ratio plays a role in explaining the difference

in MR use between the two countries. The mechanisms involved in higher staffing levels suggest

that, in such scenarios, the staff may be more able to observe, interact, and cooperate with the

patients, likely reducing the number of stressed staff members and thereby calming the

environment. The presence of more personal resources for positive reinforcement, support, and

communication, would hopefully create a more empowering and recovery-orientated environment

for patients.

Fifth, identification of a patient’s crisis triggers: Differences were found in units in which the initial

nurse evaluation of at-risk patients always contained a written evaluation of aggressive, self-

mutilating, or suicidal behaviours (e.g., crisis triggers, unique calming techniques, and coping

strategies) vs. units in which the initial evaluation only occasionally or never contained a written

evaluation. This factor explains a large proportion (∆ exp[B] = -10%) of the difference in MR

episodes between the countries. The identification of a patient’s crisis triggers has been supported

by earlier research. In a well-designed before-and-after study, Jonikas (79) observed a significant

decrease (between 49% and 99%) in physical restraint episodes after patients were interviewed to

determine their crisis triggers and personal crisis management strategies. In a well-designed cluster

RCT, Putkonen (145) reported the same tendency, in which the difference between the intervention

and control wards was a 62% reduction in seclusion and restraint time after implementing “six core

strategies” to prevent seclusion and restraint. In both studies, the effect was attributed to variables

other than the patients’ crisis triggers. In a one-way case-crossover-designed study, Fluttert (166)

reported that implementing the Early Recognition Method (ERM) decreased the number of

seclusions by 54%. The ERM is a technique used by patients and staff to systematically work with

the patients’ triggers (early warning signs), crisis intervention, and coping strategies; this is the

same identification procedure that we asked the clinical nurse managers about in the questionnaire.

Overall, earlier research supports the finding that, identification of a patient’s crisis triggers, is a key

part of explaining the difference in MR use between the two countries. The mechanism underlying

this process is likely that patients and staff recognise earlier signs of a forthcoming crisis and act

together to counteract escalation of the crisis, thereby reducing the risk of a MR episode.


45

Sixth, the separation of acutely disturbed patients: This single factor had opposing results in that

units that were able to separate acutely disturbed patients from other patients had more MR

episodes vs. units that were unable to separate patients. The factor explains a large proportion (∆

exp[B] = 13%) of the difference in MR episodes between the countries. This negative association

has not been supported by earlier research; in fact, the opposite results have been indicated. In a

qualitative content analysis study, Meehan (81) noted that the separation of acutely disturbed

patients would reduce the number of aggressive incidents. Because this result is not supported by

earlier research, more research is needed to explore the mechanisms involved in placing one or a

few patients in a separate and isolated part of the unit.

9.2 Design Issues

The initial systematic review was conducted to identify evident and effective MR-preventive

factors. We quickly realised that no meta-analyses or meta-syntheses were available. Insufficient

high-quality studies covering individual MR-preventive factors were also not found; thus,

performing our own was not possible, although this approach would have been optimal. Through

such an approach, Studies 3 and 4 would have included powerful hypothesis testing for the effects

of the MR-preventive factors in practice. Instead a combined systematic review was developed to

identify as many evident and effective MR-preventive factors as possible. The combination of

quantitative and qualitative research results might be controversial, but this is becoming more

common in highly complex research areas involving human interaction (167). Often the qualitative

research is only “recognised” as the lowest recommendation grade, in line with the opinions of

respected authorities or reports of expert committees, and not as “real” research. To establish the

more evident power of qualitative research, Kearney’s (50) qualitative hierarchy was introduced in

the model. Still, in the effect and recommendation grade matrix, the effect side of the matrix is

“flawed” due to the basic qualitative research methodology, which does not provide an effect as part

of the results or findings. For example, if the combined research does not include quantitative

research, it is only provided with an effect of five (the lowest effect level). Regardless, this review

does provide more evident “power” for the qualitative research and thereby the combined evidence

is greater than of earlier reviews regarding complex areas involving human interaction (MR-

prevention). Unfortunately, none of the identified MR-preventive factors reached the top end of the

evidence hierarchy (here the effect size of the matrix was not very important because this would

have been the objective of later investigation) and were unable to provide Studies 3 and 4 with the

possibility of powerful hypotheses testing for the effects of the MR-preventives factors in practice.

Instead, almost all of the results from the systematic review were investigated in the later studies.


46

Regarding the comparative investigation of the use of coercive measures in Europe, this was

designed as a combination of a literature review and a questionnaire survey. This design was chosen

because it allowed for the investigation of practices in a large population in an inexpensive and

timesaving maner. A more thorough and ”scientific” design would have be required to examine all

parties involved in coercive procedures (e.g., health authorities, psychiatrists, nurses, and patients)

using large-scale questionnaire survey or personal interviews regarding the procedures, practices,

and the actual use of these measures. This approach would have produced data of a more evident

character for analytic investigation. However, this approach would have been very comprehensive,

difficult, and expensive to undertake, which was not the task at hand.

The two last studies were based on a questionnaire survey, designed to investigate the relationship

between the number of MR episodes and the degree to which the MR-preventive factors were

implemented in the “closed” units, at a single time point (cross-sectional). This design was chosen

because it made it possible to investigate large populations (Danish and Norwegian “closed”

psychiatry) and complex phenomena (many MR-preventive factors and confounders) in a

reasonable, inexpensive and not very time-consuming maner. Theoretically, the best design would

have been to conduct RCTs of all individual MR-preventive factors and to compare the results of

these trials. Apart from the excessive use of resources and the lack of psychiatric units necessary to

obtain sufficient power in the analyses, such a practice would be impossible because most patients

who experience MR are under involuntary/forced admission and are therefore not allowed to

participate in trials by law. Another approach would have been to allocate the MR-preventive

factors to the units in an efficient experimental design, removing the need for many RCTs.

However, given the number of MR-preventive factors requiring envestigation, this approach would

still require too many units. Furthermore, many units have previously implemented certain MR-

preventive factors. A third approach could be to use a case-control design asking outpatients how

and to what extent they had experienced MR-preventive factors (using those who had experienced

MR as cases, and those who had not experienced MR as controls). In this scheme, investigating

many MR-preventive factors would not be possible because the outpatients would not possess the

required information, e.g., patient-staff ratio, staff experience, staff education, and risk assessment.

The last approach that may have constituted a stronger design would have been to repeat the data

collection over time and to examine the changes in the extent to which MR-preventive factors were

implemented and in the number of MR episodes. However, this design was not chosen due to the

limited study period and problems with organisational changes, thus, some units experience changes

in functions and leaders over a short period of time. The changes in organisational factors would


47

have confounded the effects, thus minimising the validity of the study outcome. Ultimately, the

chosen design was preferred to investigate multiple MR-preventive factors and confounders in a

population of this size.

The analytic design of Study 3 was used to identify potential associations between the number of

MR episodes and the MR-preventive factors. In Study 4 similar analysis was applied to determine

the degree which the MR-preventive factors could explain the different numbers of MR episodes

between Denmark and Norway. For the main analyses linear regression was chosen because it is a

common and well-known statistical model that yiels easily interpretable results, although,

categorising the MR-preventive factors and logarithmically transforming the outcome variable (the

number of MR episodes) were nessesary. Because the outcome was a count, Generalised Linear

Models (GLMs) e.g. Poison regression could have been optimal, however, the logarithmically

transformed outcome variable in a linear regression model, provided a better fit for the statistical

model for the dataset (GEE, Quasi Likelihood under Independence Model Criterion [QIC]: 83,417),

than Poisson regression (GLM, QIC: 215,816). The resulting model investigated the outcome

variable and all MR-preventive factors and covariates/confounders (one nested) in a large

population. Using the GEE technique for nesting, we were able to include dichotomised,

categorised, and linear covariates in the models.

9.3 Cause and Effect

A causal association is one in which a change in the exposure frequency results in a corresponding

change in the outcome frequency. Making judgments about causality involves a chain of logic: first,

whether the statistical association is valid (chance); next, the strength of the association and to its

biological credibility; and finally, the generalisability and consistency of the results with those of

other investigations. Assessing the statistical validity of the association involves determining the

likelihood that alternative explanations such as chance, bias, or confounding could account for the

findings (168).

9.3.1 Chance

Chance refers to whether the included participants represent the whole population (168).

In Study 1, this objective could be translated as whether all of the papers dealing with MR-

prevention were obtained in the search strategy; the answer is likely not. However, no other

literature reviews at the time had coved other MR-preventive issues, and none included other

references. As a result, a very large portion of the available MR-preventive literature was found and

investigated.


48

In Study 2, all EPSO members did answer the questionnaire, but very few UEMS members

answered the questionnaire, thus ascertaining any possible differences between the current

legislation and how tits transformed into practice, was impossible. Although we believe that the

UEMS members represent the whole population of their country, regarding national legislation, we

did not consider questions regarding general perceptions of the intrusiveness of coercive

interventions, which could be coloured by an individual’s personal beliefs, to represent the whole

population. However, most countries have addressed this question within their legislation (using the

measure of less intrusiveness) to disconnect this decision from personal beliefs. Overall, the

answers represent the whole population regarding the legislation, if not necessarily the actual

practise.

In Studies 3 and 4, a response rate of 49% was obtained. Therefore, we can only apply our results to

the entire population with some measures of certainty due to the risk of systematic differences

between responders and non-responders. Regarding the risk of systematic flaws or differences, we

reviewed the non-responders’ reasons for not responding. Most reasons appeared to be unrelated to

the questions under investigation. If a failure to respond was caused by a failure to implement

preventive factors or connected with high numbers of MR episodes, then our results would be

strengthen our findings would be underestimated. A proper drop-out analysis has not been

conducted; consequently it is unknown whether the high drop-out rate resulted in systematic flaws.

9.3.2 Bias

Bias is an alternative explanation for an observed relationship of an exposure with the outcome in

which some aspect of the design or execution of a study introduces a systematic error into the

results (168).

Selection bias is normally described as using non-comparable criteria to enrol participants in an

investigation (168). In Study 1, this could be translated, to the use of different criteria for sifting,

grading, and synthesising the evidence. Only one person was involved in the initial review process,

up to the drafting of evidence statements. Arguably, two or more persons reading all of the material

followed by consensus evaluations, would produce more reliable results; however, this could have

introduced inter-rater differences. To reduce bias, all of the evidence was gathered and strictly rated

according to set schemes in exactly the same manner. In Study 2 the enrolment criteria for the

questionnaire survey were the same for all responders; we presume that the staff who responded

was knowledgeable in the area of metal health legislation. Regarding the literature review, this

process was conducted by several people to screen for background variables and numbers on


49

coercion mostly due to the limited data available at the time. In Studies 3 and 4, a type of selection

bias sometimes called non-respondent bias might have emerged between the respondent and the

non-respondent groups. The non-responders’ reasons for failure to reply were reviewed, and no

systematic differences were observed between the responders and non-responders, Still, a proper

dropout analysis was not conducted. Reviewing the Danish National Register on coercion, we found

that 49% of the responding units were “responsible” for 52% of the MR episodes in the same

population, suggesting minimal differences in MR episodes between responders and non-

responders.

Recall bias may arise because individuals with particular exposure or adverse health outcome are

likely to remember their experiences differently from those who are not similarly affected (168). In

Studies 1 and 2, recall bias was not an issue. In Studies 3 and 4, recall bias could have occurred,

especially for the Danish data; the clinical nurse manager had to recall all MR-preventive factors as

well as the conditions and numbers of MR episodes from two years prior. Although, an earlier study

determined that nurse recall is reliable for 4 weeks (169), the clinical nurse managers in these

studies were able to support their memories by referring to charts and protocols to answer many of

the questions. The outcome variable (i.e., the number of MR episodes) was validated or re-

examined revealing that no significant recall bias existed with regard to the outcome variable.

Additionally, recall bias could have affected the results if the MR-preventive factors had only been

implemented for a portion of the year. Regardless, possible recall bias would likely only increase

the effect. Still, we must consider the possibility of recall bias with regard to certain variables.

Measurement-/insensitive measures of bias arise if the items in a questionnaire survey are not

sensitive enough to detect potential important factors (170). In Study 2, the questionnaire survey

only addressed interpretative legislative areas, not revealing any non-detection of important matters.

In Studies 3 and 4, such significant non-detection could be of concern. However, the findings were

probably quite conservative, underestimating the actual effect, because the questionnaire only

roughly covered the MR-preventive factors with questions that could measure all aspects of such a

MR-preventive factor. A MR-preventive factor such as the patient-staff ratio is simple and straight-

forward, but a more complex MR-preventive factor such as cognitive milieu therapy, could lack the

sensitivity to identify the most important MR-preventive responses due to the many relevant aspects

such as a clear objective, a described philosophy, management support and knowledge, the

economy, qualified supervisors, cultural changes, and patient education and involvement, among

others. Still, complex MR-preventive factors may have a greater impact on the reduction of MR

episodes than those found in Studies 3 and 4.


50

Expectation bias occurs when participants err in measuring data towards the expected outcome

(170). In Studies 3 and 4, the number of MRs was an important outcome, and most of the data

regarding outcome were taken from existing electronic or paper charts with no obvious

manipulation. Whether the clinical nurse managers err toward the expected outcome when

measuring certain preventive factors is unknown. However, clinical nurse managers were urged to

describe the conditions of the unit in a neutral manner.

Analysis or transfer bias occurs upon transferring the data into organised structures for analysis

(170). In Studies 3 and 4, this problem was likely avoided in the design phase, during which

definitions, hypotheses, and analyses were prepared and scrutinised.

Cognitive dissonance bias occurs if the investigator is convinced that his or her beliefs about the

findings are valid, regardless of findings to the contrary (170). Because Studies 3 and 4, re-

examines the preventive factors that others have previously described, the risk of cognitive

dissonance bias is minimal.

Biases in questionnaires, is an important issue in this thesis. A couple has been addressed earlier,

but most of the many types of bias must be eliminated in the construct-phase, as follows: in

designing questions, in designing the questionnaire, and in administration of the questionnaire. Choi

(171) describes in all 48 common types of bias in questionnaires, nearly all eliminated by strictly

following the guide by Olsen (63) in Study 2, and following QAS-99 by Willis (68), concurrent

probing by Watt (70), and cognitive interviews by Willis (69) in Studies 3 and 4.

9.3.3 Confounding

Confounding can be thought of as a mixing of the effect of the exposure on the outcome with a third

factor. This third factor must be associated with the exposure and the outcome, independently of

that exposure (168). In Studies 1 and 2, confounding was important in the assessment of the

individual papers included in the literature reviews, but was otherwise not an issue because the

studies were not analytic. In Studies 3 and 4 confounding was a very important issue. In Study 3,

certain confounding covariates e.g. patient flow, diagnostic distribution, different population bases

of the units and age of the patients who experienced MR, might be missing from the analyses.

However, the previously described covariates did not act as confounders in Denmark because the

geographic regions and types of the units functioned as proxy covariates. Geographic regions

eliminated the confounding effect of the different population base of the units, and the type of unit

eliminated the confounding effect of the diagnostic distribution, patient flow, and age of the

patients. These effects were determined by analysing data from the Danish National Register on


51

Coercion and the Danish Psychiatric Research Register. However, these data could not be accessed

in the Norwegian sample; therefore, whether these conclusions apply to the entire dataset remains

unknown, although it is reasonable to deduce that they do. In Study 4, the same considerations

regarding confounding covariates were made, except that, patients who experienced MR were

thought to be more aggressive or self-mutilating in one of the countries, which represented a

confounder. However, an earlier study that assessed aggressive behavior (172) did not report large

differences between Denmark and Norway. Additionally, some of the covariates (number of

forensic patients, type of bed, and the reason for MR) could eliminate portions of the difference.

Thus, the potential differences in aggressive or self-mutilating patients did not likely influence the

results. Overall, the most important covariates either were included or covered the effects of others

by proxy (3).

9.3.4 Strength of the Association

The magnitude of the observed association is useful to judge the likelihood that the exposure itself

affects the outcome and therefore the likelihood of a cause-effect relationship (168). The magnitude

of the observed associations only refers to the outcomes of Studies 3 and 4. In Study 3, the three

MR-preventive factors that were significantly associated with a reduction in MR episodes varied

from exp(B) = 0.36 to 0.54, both of which represents quite large effects, supporting the plausibility

of a cause-effect relationship. In Study 4, the six MR-preventive factors that confounded the

differing effects on MR use between Denmark and Norway varied between ∆exp(B) = 10% and

51% . Although these data are not as convincing as the MR-preventive factors in Study 3, they do

address some reasonable effects. A larger sample would nonetheless have been preferred, as

stronger associations and more significantly associated MR-preventive factors would likely have

followed.

9.3.5 Biological Credibility

The existence of a cause-and-effect relationship is supported if there is a known or postulated

biological mechanism by which the exposure might reasonably alter the outcome (168). Although

this definition is more relevant when the study concerns the risk of disease, some neuro-

physiological explanations concerning factors such as anxiety, fear, and aggression have accounted

for these emotions as a part of the MR-sequence (173). However, not enough data are available to

establish a cause-effect relationship based on biological credibility.


52

9.3.6 Generalisability and Consistency

Generalisability refers to whether the results are applicable to other populations (168). The study

covered half of the eligible populations in Norway and Denmark: however, because we found no

systematic differences between the respondent and non-respondent groups, the results can

reasonably be assumed to cover these two populations. If compared to other populations (e.g., other

countries), differences in legislation, culture, practise, and other factors might cause overwhelming

obstacles to application. As the differences between Norway and Denmark were small, the present

results could perhaps be generalised by considering the differences between these countries and

other similar countries (e.g. Sweden or Finland). However, applying the results to other countries

would require a thorough examination of their cultures, legislation, and other relevant factors.

Consistency with other studies calls for comparisons with studies conducted in other populations,

by different investigators, at various times, and using alternative methodologies in a variety of

geographic and cultural settings (168). The findings of these studies are supported by many

international studies using different methodologies and by different investigators. However, if we

examine individual MR-preventive factors, the support and variation have not reached the level

required for “scientific proof” or consistency, possibly due to the complexity of the MR sequence.

9.4 Discussion of International High Quality Research

Five RCTs dealing with MR (coercion, MR, restraint, or seclusion) prevention were identified, as

follows:

Georgieva (174), “Reducing seclusion through involuntary medication: A randomised clinical

trial”: Although the use of involuntary medication was able to successfully reduce the number of

seclusions, alternative interventions are required to reduce the overall number and duration of

coercive incidents. The intervention involved the use of either seclusion or involuntary medication

as the first choice of intervention. This paper has been published after the initial literature review. In

Denmark, these guidelines have been described in the Danish Mental Health Act — the use of the

least intrusive intervention (6) — and generally medication, preferably voluntary, if not possible,

then involuntary, are considered less intrusive and hence the first choice before initiating MR. In

Norway, the principle of — the use of the least intrusive intervention — is also described in the

Norwegian Mental Healt Act (65), whether involuntary medication in practice are the first choice

before initiating MR is unknown, but the conclusion, that alternative interventions are needed to

reduce the overall number and duration of coercive incidents, complies to Norway as well.


53

Berg (175), “A Randomised Controlled Comparison of Seclusion and Mechanical Restraint in

Inpatient Settings”: The result did not provide evidence for the use of one intervention over the

other. The conclusion was that, clinical decisions should account for patients preferences. The study

compared the restrictiveness of seclusion and MR from the patient´s point of view. This paper has

been published after the initial literature review. In Denmark, seclusion cannot be used; thus, taking

patient´s preferences into account, is not possible at the time being. In Norway patient´s

preferences, in this respect, may be taken into account, perhaps reducing the numbers of MR

episodes and increase the numbers of seclusion, thereby reducing the patients experiences of

restrictiveness.

Abderhalden (83), “Structured risk assessment and violence in acute psychiatric wards: randomised

controlled trial”: Structured risk assessment during the first days of treatment may contribute to

reduced violence and coercion (ed. forced acute tranquillising medication, seclusion and MR) in

acute psychiatric wards. The study investigated the implementation of BVC, which has been

implemented in more than 80% of “closed” units in Denmark. The amount of units that has

implemented the BVC in Norway, is unknown, but it is likely the same.

van de Sande (149), “Aggression and seclusion on acute psychiatric wards: effect of short-term risk

assessment”: The routine application of structured risk assessment measures was found to possibly

help reduce the incidents of aggression and the use of restraint and seclusion in psychiatric wards.

The study investigated the implementation of BVC in combination with Kenndy–Axic V. This

paper has been published after the initial literature review. As indicated above, more and more units

have implemented BVC in Denmark and Norway. Many units, also use the Global Assessment of

Functioning Scale, which is rather similar to Kenndy–Axic V, as both are brief clinician-scored

global assessment instruments used to measure the functioning of patients (176).

Putkonen (145), “Cluster-Randomised Controlled Trial of Reducing Seclusion and Restraint in

Secured Care of Men With Schizophrenia”: Seclusion and restraint were prevented without any

increase in violence on the wards for men with schizophrenia and violent behaviour. The study

investigated the implementation of the “Six Core Strategies”, as follows: leadership toward

organisational change, workforce support and development, the use of primary prevention tools, the

use of data to inform practice, debriefings, and meaningful consumer roles. This paper has been

published after the initial literature review. Parts of the “six core strategies” have been implemented

in Denmark and Norway, e.g. the use of risk assessment tools (use of primary prevention tools) and

some debriefing techniques, but most parts have not been implemented.


54

These studies represent the highest international evidence available at this time. Comparing these

with the Danish and Norwegian MR-preventive initiatives in practise, as previously described, the

“Six Core Strategies” provide the best potential for developing MR-prevention practices in

Denmark and Norway. Comparing these strategies with the findings in Studies 3 and 4 of this

thesis, nearly all eight of the MR-preventive factors are components of the “Six Core Strategies”,

supporting the suggestion that the high evidence supporting implementation of the “Six Core

Strategies” could increase the overall causality of the eight identified MR-preventive factors or that

implementing of the “Six Core Strategies” might reduce the MR episodes in Denmark and Norway

as in a special setting in Finland (consistency). Although these five RCTs represent the highest

evidence available, none reached the highest grade of evidence. Furthermore, the settings in which

they were conducted only partially comply with the conditions of Danish and Norwegian

psychiatry.

10 Conclusions

10.1 Conclusions from the Individual Studies

Study 1: The objective of the first study was to answer the question: Which conditions in nursing

and which nursing interventions have been shown to reduce the frequency of mechanical restraint

episodes? No interventions reached the highest degree of recommendation combined with the

highest effect. Because no interventions reached the highest degree of recommendation (evidence),

the question cannot be answered with certainty. Therefore, the studies (Studies 3 and 4) conducted

later must include as many MR-preventive factors as possible, testing not only a few factors to be

effective in research settings (obtaining the highest recommendation grad), in practice (by

hypothesis tests, increasing the power of the later studies).

Study 2: The objective of the second study was: To compare the use of coercive measures on

psychiatric inpatients between Denmark and Sweden, Norway, Finland, Iceland, Belgium, The

Netherlands, United Kingdom, Ireland, France, and Italy, in order to share knowledge and learn

about what other countries do to minimize coercion. The study primarily revealed, that the many

differences in legislation between the countries and that very few countries, at the time of the study,

produced nationwide public data on coercive measures and background data, making the

comparison very difficult. Thereby, not reaching the objective on what to learn about what other

countries do to minimize coercion. Targeting the comparison between Denmark and Norway,

demonstrated a tendency for Norway to use less MR and physical restraint, than Denmark. Most


55

other indicators were similar between the two countries. Thus, Norway was included in the later

studies (Studies 3 and 4). Overall, caution must be applied in the interpretation and comparison of

coercive measures and factors related to coercion between countries.

Study 3: The objective of the third study was to answer the question: How are presumed MR-

preventive factors of nonmedical origin associated with the frequency of MR episodes in Denmark

and Norway? The results suggest that mandatory review, patient involvement, and crowding were

associated to a reduced frequency of MR episodes in Denmark and Norway. However, we were

unable to establish a causal effect between the reduction in MR episodes and these preventive

measures. Given that we included important potential confounders and that MR-preventive factors

have not been shown to present any adverse effects on the patients or staff in the literature, the

identified MR-preventive factors might be investigated further because they may contribute to a

better understanding of the phenomenon of MR.

Study 4: The objective of the fourth study was: To examine how presumed MR-preventive factors

of non-medical origin may explain the differing number of MR episodes between Denmark and

Norway. The results suggest that: the identification of patient’s crisis triggers; an increased number

of staff per patient; increased staff education; better work environment; and reduced substitute staff

use might explain why MR was initiated more often in Demark than in Norway. We cannot claim

that the implementation of these potential preventive measures, primarily in Denmark and in a few

units in Norway, would reduce the number of MR episodes. Given that we considered several

important potential confounders and that the MR-preventive factors demonstrated no adverse

effects on the patients or staff in the literature, the identified MR-preventive factors should be

investigated further because they may contribute to a better understanding of the phenomenon of

MR.

10.2 General Conclusions

The overall objective of this thesis was to generate knowledge of non-medical MR-prevention.

This thesis and the four included studies complement each other by using different means to

investigate specific preventive factors with the potential to minimise the number of MR episodes.

This thesis began with a systematic review of the literature to identify the most evident and

effective preventive MR factors. The practices in other countries were then examined, and Norway

was identified to be included in the later studies (Studies 3 and 4). Finally, specific MR-preventive

factors that could reduce the number of MR episodes in practice were investigated.


56

These results have increased our understanding of the ability of specific MR-preventive factors to

reduce the number of MR episodes in Denmark and Norway, thereby generating knowledge in the

field of MR-prevention. These findings have not been identified via randomised controlled trials

(RCTs), and although some biases could be present, the questionnaire was thoroughly developed to

include several potential confounders. Furthermore, similar results have been demonstrated in

previous international studies. These factors are not likely to have adverse effects on the patients or

staff. Rather, the potential positive effects of these factors on the prevention of MR episodes may

reflect a general strengthening of the care and treatment environments. Therefore, further

investigation into the effects of implementing the following within Danish and Norwegian practices

is recommended: mandatory review, patient involvement, less crowding, higher staff education, less

substitute staff use, better work environment, increased number of staff per patient, and the

identification of the patient’s crisis triggers.

11 Perspectives

11.1 Clinical Implications

In Denmark the need to implement MR-preventive strategies is more urgent than ever; the number

of persons experiencing MR has never been higher, and the number of MR episodes is among the

highest over the 24 years of registration. Although, Norway uses less MR than Denmark, the

numbers has not decreased over the last decade; perhaps it even has a tendency to increase, making

the need for implementing MR-preventives strategies equally important.

First, though these studies addressed MR of inpatients in “closed” units, it would have been

preferred, that the patients had never become ill enough to require forceful/involuntary admission.

Thus, outpatient prevention, e.g., as joint crisis planning (177) or the use of the early recognition

method (166), should also be addressed in cooperation with the primary and secondary mental

health sectors. Another means of decreasing forced admission rates into “closed” units might be to

make it easier for patients to be voluntarily admitted.

Second, as shown in this thesis and as supported by other researchers (38;39;178;179), the

phenomenon of MR is multifaceted, and the reduction in the number of MR episodes thus requires

interventions in many different areas. Accordingly, many MR-preventive initiatives such as several

individual initiatives or programs encompassing several initiatives must be implemented in practise.


57

Third, the selection process of MR-preventive factors to implement in practise should address both

those supported by the most obvious research and those most transferrable to Danish or Norwegian

conditions.

Fourth, any implementation of MR-preventive initiatives in practise must be followed by an

investigation of eventual adverse effects, e.g. injuries to patients or staff, rises in other coercive

measures, or the decreased psychological wellbeing of patients or staff.

The results from this thesis arise from a Danish and a Norwegian cohort; thus, although the

generalisability is acceptable, the grade of the evidence is not high, indicating that causality has not

been established. Thus, the clinical implications of the results of this thesis might include

investigating the effects and adverse effects of implementing a program containing the identified

MR-preventive factors in such a way that the effects of individual elements (organisational factors,

MR-preventive methods and tools), on the number of MR episodes can be addressed.

11.2 Future Research Recommendations

Although some RCTs have been conducted, they are biased due to the complex nature of the MR

phenomenon. Furthermore, three of these studies are cluster RCTs, and one of them involves

optional randomisation, because involuntary admitted patients in most countries cannot be included

in trials. This limitation suggests that a very well-designed and conducted RCT cannot by itself

reach the highest grade of recommendation; instead, this level of evidence requires several studies

be conducted before a meta-synthesis can be performed to reach the highest evidence grade.

Furthermore, the conducted RCTs should be transferrable to different Danish and Norwegian

settings.

Therefore, the future recommendations for research are to conduct more cluster RCTs in different

settings in Denmark and Norway on specific MR-preventive initiatives, methods, and tools such as

mandatory review, patient involvement, less crowding, higher staff education, less substitute staff

use, a better work environment, increased number of staff per patient, and the identification of the

patient’s crisis triggers. Also, conducting cluster RCTs, investigating programs or strategies

containing several of the before mentioned MR-preventive initiatives might be recommended.


58

12 Reference List

(1) Bowers L, Van Der Werf B, Vokkolainen A, Muir-Cochrane E, Allan T, Alexander J. International variation

in containment measures for disturbed psychiatric inpatients: A comparative questionnaire survey. Int J Nurs

Stud 2007;44(3):357-64.

(2) Hornby AS. Oxford Advanced Learner´s Dictionary of Current English. 8 ed. Oxford: Oxford University

Press; 2010.

(3) Janssen WA, van de Sande R, Noorthoorn EO, Nijman HLI, Bowers L, Mulder CL, et al. Methodological

issues in monitoring the use of coercive measures. Int J Law Psychiatry 2011;34(6):429-38.

(4) Bak J, Aggernæs H. Coercion within Danish psychiatry compared with 10 other European countries. Nord J

Psychiatry 2011;66(5):297-302.

(5) Ministeriet for Sundhed og Forebyggelse, Minis. Bekendtgørelse om tvangsprotokoller og optegnelser samt

registrering og indberetning af tvang samt udskrivningsaftaler og koordinationsplaner på psykiatriske

afdelinger. BEK nr 1342. 2-12-2010.

Ref Type: Statute

(6) Ministeriet for Sundhed og Forebyggelse. Bekendtgørelse om anvendelse af tvang i psykiatrien. LBK nr

1729. 2-12-2010.

Ref Type: Statute

(7) Reiter PJ. Sindsyge og sindsygebehandling. En håndbog for sygeplejersker. København: Levin og

Munksgaards forlag; 1927.

(8) Pontoppidan K. Betragtninger over vort Sindsygevæsen. Bibliotek for Læger 1901;523-5.

(9) Sundhedsstyrelsen. Cirkulære om anvendelse af tvangsmidler og beskyttelsesforanstaltninger over for

patienter på psykiatriske sygehuse og afdelinger samt de dertil knyttede plejehjem m.v. 63. 28-6-1985.

Ref Type: Statute

(10) Sundhedsstyrelsen. Tvangsforanstaltninger i psykiatrien - en udredning. København: Sundhedsstyrelsen;

2009.

(11) United Nations. The protection of persons with mental illness and the improvement of mental health care.

RES 46/119. 17-12-1991.

Ref Type: Statute

(12) The European Committee for the Prevention of Torture and Inhumane or Degrading Treatment or

Punishment. Report to the Government of Denmark on the visit to Denmark carried out by the European

Committee for the Prevention of Torture and Inhumane or Degrading Treatment or Punishment. Strasburg:

The European Committee for the Prevention of Torture and Inhumane or Degrading Treatment or

Punishment; 2002.





Punishment; 2008.





59


Punishment; 2014.

(15) Hem E, Steen O, Opjordsmoen S. Thrombosis associated with physical restraints. Acta Psychiatr Scand

2001;103(1):73-6.

(16) Laursen SB, Jensen TN, Bolwig T, Olsen NV. Deep venous thrombosis and pulmonary embolism following

physical restraint. Acta Psychiatr Scand 2005;111(4):324-7.

(17) Hick JL, Smith SW, Lynch MT. Metabolic Acidosis in Restraint associated Cardiac Arrest - A Case Series.

Acad Emerg Med 1999;6(3):239-43.

(18) Morrison A, Sadler D. Death of a psychiatric patient during physical restraint. Excited delirium-a case

report. Med Sci Law 2001;41(1):46-50.

(19) Nielsen AS. Deep venous thrombosis and fatal pulmonary embolism in a physically restrained patient.

Ugeskr Laeger 2005;167(21):2294.

(20) Bonner G, Lowe T, Rawcliffe D, Wellman N. Trauma for all: a pilot study of the subjective experience of

physical restraint for mental health inpatients and staff in the UK. J Psychiatr Ment Health Nurs

2002;9(4):465-73.

(21) Bower FL, McCullough CS, Timmons ME. A Synthesis of What We Know About the Use of Physical

Restraints and Seclusion with Patients in Psychiatric and Acute Care Settings: 2003 Update. J Knowl Synth

Nurs 2003;10(1):1-29.

(22) Cusack KJ, Frueh BC, Hiers T, Suffoletta-Maierle S, Bennett S. Trauma within the psychiatric setting: A

preliminary empirical report. Adm Policy Ment Health Ment Health Serv Res 2003;30(5):453-60.

(23) Frueh BC, Knapp RG, Cusack KJ, Grubaugh AL, Sauvageot JA, Cousins VC, et al. Special section on

seclusion and restraint: Patients' reports of traumatic or harmful experiences within the psychiatric setting.

Psychiatr Serv 2005;56(9):1123-33.

(24) Steinert T, Bergbauer G, Schmid P, Gebhardt RP. Seclusion and restraint in patients with schizophrenia:

clinical and biographical correlates. J Nerv Ment Dis 2007;195(6):492-6.

(25) Jacobsen AT, Krarup F. Vejledning i sindsygepleje. København: Sundhedsstyrelsen; 1920.

(26) Ministeriet for Sundhed og Forebyggelse. Opgørelse over anvendelse af tvang i psykiatrien 2001-2013.

København: Ministeriet for Sundhed og Forebyggelse; 2014.

(27) Keski-Valkama A, Sailas E, Eronen M, Koivisto AM, Lönnqvist J, Kaltiala-Heino R. A 15-year national

follow-up: legislation is not enough to reduce the use of seclusion and restraint. Soc Psychiatry Psychiatr

Epidemiol 2007;42(9):747-52.

(28) Nijman HLI, Campo JMLG, Ravelli DP, Merckelbach HLGJ. A tentative model of aggression on inpatient

psychiatric wards. Psychiatr Serv 1999;50(6):832.

(29) Bowers L. Safewards: a new model of conflict and containment on psychiatric wards. J Psychiatr Ment

Health Nurs 2014;21(6):499-508.

(30) Peen J, Schoevers RA, Beekman AT, Dekker J. The current status of urbanrural differences in psychiatric

disorders. Acta Psychiatr Scand 2010;121(2):84-93.

(31) Munk-Jørgensen P, Mortensen PB. The Danish Psychiatric Central Register. Dan Med Bull 1997;44(1):82.

(32) Norredam M, Garcia Lopez A, Keiding N, Krasnik A. Excess use of coercive measures in psychiatry among

migrants compared with native Danes. Acta Psychiatr Scand 2010;121(2):143-51.


60

(33) Beghi MASS, Peroni FEDE, Gabola PIER, Rossetti AURO, Cornaggia CM. Prevalence and risk factors for

the use of restraint in psychiatry: a systematic review. Rivista di psichiatria 2013;48(1):10-22.

(34) Kaltiala-Heino R, Tuohimäki C, Korkeila J, Lehtinen V. Reasons for using seclusion and restraint in

psychiatric inpatient care. Int J Law Psychiatry 2003;26(2):139-49.

(35) Knutzen M, Sandvik L, Hauff E, Opjordsmoen S, Friis S. Association between patients' gender, age and

immigrant background and use of restraint-A 2-year retrospective study at a department of emergency

psychiatry. Nord J Psychiatry 2007;61(3):201-6.

(36) Knutzen M, Mjosund NH, Eidhammer G, Lorentzen S, Opjordsmoen S, Sandvik L, et al. Characteristics of

Psychiatric Inpatients Who Experienced Restraint and Those Who Did Not: A Case-Control Study. Psychiatr

Serv 2011;62(5):492-7.

(37) Whitehead PD, Liljeros F. Heavy-Tailed Distribution of Seclusion and Restraint Episodes in a State

Psychiatric Hospital. J Am Acad Psychiatry Law 2011;39(1):93-9.

(38) Johnson ME. Violence and restraint reduction efforts on inpatient psychiatric units. Issues Ment Health Nurs

2010;31(3):181-97.

(39) Scanlan JN. Interventions To Reduce the Use of Seclusion and Restraint in Inpatient Psychiatric Settings:

What We Know So Far a Review of the Literature. Int J Soc Psychiatry 2010;56(4):412-23.

(40) Steinert T, Lepping P, Bernhardsgrütter R, Conca A, Hatling T, Janssen W, et al. Incidence of seclusion and

restraint in psychiatric hospitals: a literature review and survey of international trends. Soc Psychiatry

Psychiatr Epidemiol 2010;45:889-97.

(41) Stewart D, Van der Merve M, Bowers L, Simpson A, Jones J. A review of interventions to reduce

mechanical restraint and seclusion among adult psychiatric inpatients. Issues Ment Health Nurs

2010;31(6):413-24.

(42) Strout TD. Perspectives on the experience of being physically restrained: An integrative review of the

qualitative literature. Int J Ment Health Nurs 2010;19(6):416-27.

(43) National Institute for Health and Clinical Excellence. The guidelines manual. London: National Institute for

Health and Clinical Excellence; 2007.

(44) The National Collaborating Centre for Mental Health. Self-harm, The short-term physical and psychological

management and secondary prevention of self-harm in primary and secondary care. London: The National

Institute for Health and Clinical Excellence; 2004. Report No.: Clinical Guideline 16.

(45) Fisher WA. Elements of successful restraint and seclusion reduction programs and their application in a

large, urban, state psychiatric hospital. J Psychiatr Prac 2003;9(1):7.

(46) Helllerstein DJ, Staub AMYB, Leqesne E. Decreasing the use of restraint and seclusion among psychiatric

inpatients. J Psychiatr Prac 2007;13(5):308.

(47) Bjørkly S, Knutzen M, ., Furre A, Sandvik L. Innsamling og analyse av data om bruk av tvangsmidler og

vedtak om skjerming i det psykiske helsevernet for voksne i 2009. Oslo: Kompetansesenter for sikkerhets-,

fengsels- og rettspsykiatri, Helseregion Sør-Øst; 2011. Report No.: 2011-1.

(48) Sekretariatet for Referenceprogrammer. Vejledning i udarbejdelse af referenceprogrammer. København:

Sundhedsstyrelsen; 2004.

(49) Sandelowski M, Barroso J. Handbook for synthesizing qualitative research. Berlin: Springer; 2007.

(50) Kearney MH. Levels and applications of qualitative research evidence. Res Nurs Health 2001;24(2):145-53.


61

(51) Harbour RT. SIGN 50 - A guideline developer´s handbook. Edinburg: Scottisch Intercollegiate Guidelines

Network; 2008.

(52) Critical Appraisal Skills Programme. Appraisal tools. Public Health Resource Unit 2006 [cited 2008 Jul

7];Available from: URL: http://www.phru.nhs.uk/Pages/PHD/resources.htm

(53) Health Care Practice Research & Development Unit. HCPRDU Tools - Critical appraisal tools. The Salford

Centre for Nursing, Midwifery and Collaborative Research 2005 [cited 2008 Jul 5];Available from: URL:

http://usir.salford.ac.uk/12969/

(54) Liddle J, Williamsom M, Irwing L. Method for evalueting research and guideline evidence. Sydney: NSW

Health Depatement; 1996.

(55) World Health Organization. Policies and practices for mental health in Europe meeting the challenges.

2008.

(56) Salize HJ, Dressing H, Peitz M. Compulsory admission and involuntary treatment of mentally ill patients -

legislation and practice in EU-member states. Mannheim, Germany: Central Institute of Mental Health;

2002. Report No.: Final Report.

(57) Aggernæs H, Cubli-Bauer A, Zarotti G, Schudel W, Saliba J, Pietrowski A, et al. Psychiatric Legislation in

Europe 1998. Report of the UEMS Section for Psychiatry. Prague: European Board of Psychiatry; 2001.

(58) Zinkler M, Priebe S. Detention of the mentally ill in Europe-a review. Acta Psychiatr Scand 2002;106(1):3-

8.

(59) Salize HJ, Dressing H. Epidemiology of involuntary placement of mentally ill people across the European

Union. Br J Psychiatry 2004;184:163-8.

(60) Salize HJ, Dressing H, Kief C. Placement and Treatment of Mentally Ill Offenders - Legislation and Practice

in EU Member States. Mannheim, Germany: Central Institute of Mental Health; 2005.

(61) Salize HJ, Dressing H, Kief C. Mentally Disordered Persons in European Prison Systems - Needs,

Programmes and Outcome (EUPRIS). Mannheim, Germany: Central Institute of Mental Health; 2007.

(62) Steinert T, Lepping P. Legal provisions and practice in the management of violent patients. A case vignette

study in 16 European countries. Eur Psychiatr 2008;24(2):135-41.

(63) Olsen H. Guide til gode spørgeskemaer. København: Socialforskningsinstituttet; 2006.

(64) Indenrigs- og Sundhedsministeriet. Bekendtgørelse af lov om anvendelse af tvang i psykiatrien.

Psykiatriloven, LBK nr. 1729. 2010.

Ref Type: Statute

(65) Helse- og omsorgsdepartementet. Lov om etablering og gjennomføring av psykiak helsevern. Lov nr. 62, -§

4-8. 1999. 2-7-1999.

Ref Type: Statute

(66) Sundhedsstyrelsen. Nye tal fra sundhedsstyrelsen. Anvendelse af tvang i psykiatrien 2007. København:


(67) Bremmes R, Hatling T, Bjørngaard JH. Bruk av tvangsmidler i psykisk helsevern i 2001, 2003, 2005 og

2007. Trondheim: SINTEF Helse; 2008.

(68) Willis GB, Lessler JT. Question Appraisal System (QAS-99). Atlanta: Research Triangle Institute; 1999.

(69) Willis GB. Cognitive Intervewing. A Tool for Improving Questionnaire Design. London: SAGE

Publications Ltd.; 2005.

http://www.phru.nhs.uk/Pages/PHD/resources.htm

http://usir.salford.ac.uk/12969/


62

(70) Watt T, Rasmussen Å, Groenvold M, Bjorner JB, Watt SH, Bonnema SJ, et al. Improving a newly

developed patient-reported outcome for thyroid patients, using cognitive interviewing. Qual Life Res

2008;17(7):1009-17.

(71) Schaeffer NC, Presser S. The Science of Asking Questions. Ann Rev Soc 2003;29:65-89.

(72) Lykke J, Austin SF, Morch MM. Cognitive milieu therapy and restraint within dual diagnosis populations.

Ugeskr Laeger 2008;170(5):339-43.

(73) Thyrsting K, Hall EOC. Prevention of coercion in psychiatric units: a secondary analysis of a practitioner

research project. Klin Sygepleje 2008;22(2):78-86.

(74) Khadivi AN, Patel RC, Atkinson AR, Levine JM. Association between seclusion and restraint and patient-

related violence. Psychiatr Serv 2004;55(11):1311-2.

(75) Sullivan AM, Bezmen J, Barron CT, Rivera J, Curley-Casey L, Marino D. Reducing restraints: alternatives

to restraints on an inpatient psychiatric service--utilizing safe and effective methods to evaluate and treat the

violent patient. Psychiatr Q 2005;76(1):51-65.

(76) Taxis JC. Ethics and praxis: alternative strategies to physical restraint and seclusion in a psychiatric setting.

Issues Ment Health Nurs 2002;23(2):157-70.

(77) Fletcher E, Stevenson C. Launching the Tidal Model in an adult mental health programme. Nurs Stand

2001;15(49):33-6.

(78) Sullivan D, Wallis M, Lloyd C. Effects of patient-focused care on seclusion in a psychiatric intensive care

unit. Int J Ther Rehabil 2004;(11):503-8.

(79) Jonikas JA, Cook JA, Rosen C, Laris A, Kim JB. A program to reduce use of physical restraint in psychiatric

inpatient facilities. Psychiatr Serv 2004;55(7):818-20.

(80) McCain M, Kornegay K. Behavioral Health Restraint: The Experience and Beliefs of Seasoned Psychiatric

Nurses. J Nurses Staff Dev 2005;21(5):236.

(81) Meehan T, McIntosh W, Bergen H. Aggressive behaviour in the high secure forensic setting: the perceptions

of patients. J Psychiatr Ment Health Nurs 2006;13(1):19-25.

(82) Wynn R, Mykleburst LH. Patients´ Satisfaction and Self-Rated Improvement Following Coercive

Interventions. Psychiatr Psychol Law 2006;13(2):199-202.

(83) Abderhalden C, Needham I, Dassen T, Halfens R, Haug HJ, Fischer JE. Structured risk assessment and

violence in acute psychiatric wards: randomised controlled trial. Br J Psychiatry 2008;193(1):44-50.

(84) Cochrane-Brink KA, Lofchy JS, Sakinofsky I. Clinical rating scales in suicide risk assessment. Gen Hosp

Psych 2000;22(6):445-51.

(85) Dernevik M, Grann M, Johansson S. Violent behaviour in forensic psychiatric patients: Risk assessment and

different risk-management levels using the HCR-20. Psychol Crim Law 2002;8(1):93-111.

(86) Palmstierna T, Olsson D. Violence from young women involuntarily admitted for severe drug abuse. Acta

Psychiatr Scand 2007;115(1):66-72.

(87) Donat DC. Impact of a mandatory behavioral consultation on seclusion/restraint utilization in a psychiatric

hospital. J Behav Ther Exp Psychiatry 1998;29(1):13-9.

(88) Prescott DL, Madden LM, Dennis M, Tisher P, Wingate C. Reducing mechanical restraints in acute

psychiatric care settings using rapid response teams. J Behav Health Serv Res 2007;34(1):96-105.


63

(89) Bowers L, Nijman H, Allan T, Simpson A, Warren J, Turner L. Prevention and Management of Aggression

Training and Violent Incidents on UK Acute Psychiatric Wards. Psychiatr Serv 2006;57(7):1022.

(90) Delaney KR, Johnson ME. Keeping the Unit Safe: Mapping Psychiatric Nursing Skills. J Am Psychiatr Nurs

Assoc 2006;12(4):198-207.

(91) Hahn S, Abderhalden C, Duxbury J, Halfens RJG, Needham I. The effect of a training course on mental

health nurses' attitudes on the reasons of patient aggression and its management. J Psychiatr Ment Health

Nurs 2006;13(2):197-204.

(92) Lee S, Wright S, Sayer J, Parr A, Gray R, Gournay K. Physical restraint training for nurses in English and

Welsh psychiatric intensive care and regional secure units. J Ment Health 2001;10(2):151-62.

(93) Mackay I, Paterson B, Cassells C. Constant or special observations of inpatients presenting a risk of

aggression or violence: nurses' perceptions of the rules of engagement. J Psychiatr Ment Health Nurs

2005;12(4):464-71.

(94) Needham I, Dassen T, Halfens RJG, Haug HJ, Fisher JE. The effect of a training course in aggression

management on mental health nurses perceptions of aggression: a cluster randomised controlled trial. Int J

Nurs Stud 2005;42(6):649-55.

(95) Spokes K, Bond K, Lowe T, Jones J, Illingworth P. HOVIS - Hertfordshire/Oxfordshire violent incident

study. J Psychiatr Ment Health Nurs 2002;9(2):199-209.

(96) Wright S, Sayer J, Parr AM, Gray R, Southern D, Gournay K. Breakaway and physical restraint techniques

in acute psychiatric nursing: Results from a national survey of training and practice. J Forens Psychiatry

Psychol 2005;16(2):380-98.

(97) Allen MH, Carpenter D, Sheets JL, Miccio S, Ross R. What do consumers say they want and need during a

psychiatric emergency? J Psychiatr Pract 2003;9(1):39-58.

(98) Bowers L, Brennan G, Flood C, Lipang M, Oladapo P. Preliminary outcomes of a trial to reduce conflict and

containment on acute psychiatric wards: City Nurses. J Psychiatr Ment Health Nurs 2006;13(2):165-72.

(99) Cardell R, Pitula CR. Suicidal Inpatients' Perceptions of Therapeutic and Nontherapeutic Aspects of

Constant Observation. Psychiatr Serv 1999;50(8):1066.

(100) Duxbury J. An evaluation of staff and patient views of and strategies employed to manage inpatient

aggression and violence on one mental health unit: a pluralistic design. J Psychiatr Ment Health Nurs

2002;9(3):325-37.

(101) Gillig PM, Markert R, Barron J, Coleman F. A Comparison of Staff and Patient Perceptions of the Causes

and Cures of Physical Aggression on a Psychiatric Unit. Psychiatr Q 1998;69(1):45-60.

(102) Robins CS, Sauvageot JA, Cusack KJ, Suffoletta-Maierle S, Frueh BC. Special Section on Seclusion and

Restraint: Consumers' Perceptions of Negative Experiences and "Sanctuary Harm" in Psychiatric Settings.

Psychiatr Serv 2005;56(9):1134.

(103) Smith GM, Davis RH, Bixler EO, Lin HM, Altenor A, Altenor RJ, et al. Pennsylvania State Hospital

System's seclusion and restraint reduction program. Psychiatr Serv 2005;56(9):1115.

(104) Wynn R. Staff's attitudes to the use of restraint and seclusion in a Norwegian university psychiatric hospital.

Nord J Psychiatry 2003;57(6):453-9.

(105) Flannery Jr RB, Farley E, Rego S, Walker AP. Characteristics of Staff Victims of Pschiatric Patient

Assaults: 15-Year Analysis of the Assaulted Staff Action Program (ASAP). Psychiatr Q 2007;78(1):25-37.

(106) Gim YS, Cheng SOB, Hong TL, Chin MWB, Hen WH, Lee YJ. Nurses experiences in identifying and

managing violence precipitants in a psychiatric in-patient setting. Singapore Nurs J 1999;26(1):21-4.


64

(107) Bowers L, Jeffery D, Simpson A, Daly C, Warren J, Nijman H. Junior staffing changes and the temporal

ecology of adverse incidents in acute psychiatric wards. J Adv Nurs 2007;57(2):153-60.

(108) Janssen W, Noorthoorn E, Linge R, Lendemeijer B. The influence of staffing levels on the use of seclusion.

Int J Law Psychiatry 2007;30(2):118-26.

(109) Vatne S, Fagermoen MS. To correct and to acknowledge: two simultaneous and conflicting perspectives of

limit-setting in mental health nursing. J Psychiatr Ment Health Nurs 2007 Feb;14(1):41-8.

(110) Goedhard LE, Stolker JJ, Nijman HLI, Egberts ACG, Heerdink ER. Aggression of Psychiatric Patients

Associated with the Use of As-needed Medication. Pharm Psykiat 2007;40(1):25.

(111) Thapa PB, Palmer SL, Owen RR, Huntley AL, Clardy JA, Miller LH. P.R.N. (As-needed) orders and

exposure of psychiatric inpatients to unnecessary psychotropic medications. Psychiatr Serv

2003;54(9):1282-6.

(112) Bowers L, Simpson A, Eyres S, Nijman H, Hall C, Grange A, et al. Serious untoward incidents and their

aftermath in acute inpatient psychiatry: the Tompkins Acute Ward study. Int J Ment Health Nurs

2006;15(4):226-34.

(113) Alexander J. Patients' feelings about ward nursing regimes and involvement in rule construction. J Psychiatr

Ment Health Nurs 2006;13(5):543-53.

(114) Thomas B, Jones M, Johns P, Trauer T. Prn medication use in a psychiatric high dependency unit following

the introduction of a nurse led activity programme. Int J Ment Health Nurs 2006;15(4):266-71.

(115) Bowers L, Allan T, Simpson A, Jones J, Whittington R. Morale is high in acute inpatient psychiatry. Soc

Psychiatry Psychiatr Epidemiol 2009;44(1):39-46.

(116) Daffern M, Mayer M, Martin T. Staff gender ratio and aggression in a forensic psychiatric hospital. Int J

Ment Health Nurs 2006;15(2):93-9.

(117) Owen C, Tarantello C, Jones M, Tennant C. Violence and Aggression in Psychiatric Units. Psychiatr Serv

1998;49(11):1452.

(118) Whittington D, McLaughlin C. Finding time for patients: an exploration of nursesÆ time allocation in an

acute psychiatric setting. J Psychiatr Ment Health Nurs 2000;7(3):259-68.

(119) Vaaler AE, Morken G, Fløvig JC, Iversen VC, Linaker OM. Effects of a psychiatric intensive care unit in an

acute psychiatric department. Nord J Psychiatry 2006;60(2):144-9.

(120) Bowers L, Allan T, Simpson A, Nijman H, Warren J. Adverse incidents, patient flow and nursing workforce

variables on acute psychiatric wards: The Tompkins Acute Ward Study. Int J Soc Psychiatry 2007;53(1):75.

(121) Lind M, Kaltiala-Heino R, Suominen T, Leino-Kilpi H, Valimaki M. Nurses' ethical perceptions about

coercion. J Psychiatr Ment Health Nurs 2004;11(4):379-85.

(122) Van Doeselaar M., Sleegers P, Hutschemaekers G. Professionals' attitudes toward reducing restraint: the

case of seclusion in the Netherlands. Psychiatr Q 2008;79(2):97-109.

(123) D'Orio BM, Purselle D, Stevens D, Garlow SJ. Reduction of episodes of seclusion and restraint in a

psychiatric emergency service. Psychiatr Serv 2004;55(5):581-3.

(124) Duxbury J, Whittington R. Causes and management of patient aggression and violence: staff and patient

perspectives. J Adv Nurs 2005;50(5):469-78.

(125) Vaaler AE, Morken G, Linaker OM. Effects of different interior decorations in the seclusion area of a

psychiatric acute ward. Nord J Psychiatry 2005;59(1):19-24.


65

(126) Janelli LM, Kanski G. Music for untying restrained patients. J N Y State Nurses Assoc 1998;29(1):13-5.

(127) Psykiatrian erikoisalan laitoshoito 2007. National Research and Develope Centre for Welfare and Health;

2008. Report No.: 36/2008.

(128) Beskrivning av vårdutnyttjande i psykiatrin - En rapport baserad på hälsodataregistren vid Socialstyrelsen.

Socialstyrelsen; 2008. Report No.: 2008-131-31.

(129) Den psykiatriske tvångvården - Verksamhetstillsyn av vården enligt LPT och LVR i södra

sjukvårdsregionen. Malmö: Socialstyrelsens regionala tillsynsenhet i Malmö; 2008. Report No.: 2006-126-

60.

(130) Barbato A, D'Avanzo B. Involuntary placement in Italy. Br J Psychiatry 2005;186:542.

(131) Bremnes R, Hatling T., Bjørngaard JH. Tvungent psykisk helsevern med døgnophold i perioden 2001-2006.

Trondheim: SINTEF Helse; 2008. Report No.: SINTEF A4319.

(132) Danish Nurses' Associatio. Member statistics - January 2009. Copenhagen: Danish Nurses´Association;

2009.

(133) Inspectie voor de Gezondheidszorg. Tabellenboek Wet Bopz 2002-2006. Haag: Staatstoezicht op de

volksgezondheid; 2007.

(134) Indenrigs- og Sundhedsministeriet. Bekendtgørelse om information og samtykke ved inddragelse af

forsøgspersoner i biomedicinske forskningsprojekter. LBK nr. 806. 12-7-2004.

Ref Type: Statute

(135) Northern Nurse Federation. Ethical guidelines for nursing research in the Nordic Countries. Vard Nord

2003;23(4):1-19.

(136) Council of Europe. White Paper Regarding a Draft Recommendation on Legal Protection of Persons

Suffering from Mental Disorder Especially Those Placed as Involuntary Patients. Strasbourg: Council of

Europe; 2000.

(137) Chang NA, Grant PM, Luther L, Beck AT. Effects of a Recovery-Oriented Cognitive Therapy Training

Program on Inpatient Staff Attitudes and Incidents of Seclusion and Restraint. Commun Ment Health J

2014;50(4):415-21.

(138) Oestrich I, Holm L. En kognitiv model for miljøterapi. Kognitiv miljøterapi. 1 ed. København: Dansk

psykologisk Forlag; 2001. p. 19-79.

(139) Soininen P, Välimäki M, Noda T, Puukka P, Korkeila J, Joffe G, et al. Secluded and restrained patients'

perceptions of their treatment. Int J Ment Health Nurs 2013;22(1):47-55.

(140) Azeem MW, Aujla A, Rammerth M, Binsfeld G, Jones RB. Effectiveness of six core strategies based on

trauma informed care in reducing seclusions and restraints at a child and adolescent psychiatric hospital. J

Child Adolesc Psychiatr Nurs 2011;24(1):11-5.

(141) Borckardt JJ, Madan A, Grubaugh AL, Danielson CK, Pelic CG, Hardesty SJ, et al. Systematic investigation

of initiatives to reduce seclusion and restraint in a state psychiatric hospital. Psychiatr Serv 2011;62(5):477-

83.

(142) Goetz SB, Taylor-Trujillo A. A Change in Culture Violence Prevention in an Acute Behavioral Health

Setting. J Am Psychiatr Nurs Assoc 2012;18(2):96-103.

(143) Huckshorn KA. Reducing seclusion and restraint use in mental health settings: A phenomenological study of

hospital leader and staff experiences [PhD thesis] Capella University; 2013.


66

(144) Lewis M, Taylor K, Parks J. Crisis prevention management: a program to reduce the use of seclusion and

restraint in an inpatient mental health setting. Issues Ment Health Nurs 2009;30(3):159-64.

(145) Putkonen A, Kuivalainen S, Louheranta O, Repo-Tiihonen E, Ryynänen O-P, Kautiainen H, et al. Cluster-

Randomized Controlled Trial of Reducing Seclusion and Restraint in Secured Care of Men With

Schizophrenia. Psychiatr Serv 2013;64(9):850-5.

(146) Smith GM, Ashbridge DM, Davis RH, Steinmetz W. Correlation Between Reduction of Seclusion and

Restraint and Assaults by Patients in Pennsylvanias State Hospitals. Psychiatr Serv 2014;Early online.

(147) Wale JB, Belkin GS, Moon R. Reducing the Use of Seclusion and Restraint in Psychiatric Emergency and

Adult Inpatient Services - Improving Patient-Centered Care. Perm J 2011;15(2):57.

(148) Barker P, Buchanan-Barker P. The Tidal Model of Mental Health Recovery and Reclamation: Application in

Acute Care Settings. Issues Ment Health Nurs 2010;31(3):171-80.

(149) van de Sande R, Nijman HLI, Noorthoorn EO, Wierdsma AI, Hellendoorn E, van der Staak C, et al.

Aggression and seclusion on acute psychiatric wards: effect of short-term risk assessment. Br J Psychiatry

2011;199(6):473-8.

(150) Danske Regioner. Retspsykiatri. Kvalitet og sikkerhed. 2010.

(151) Knutzen M, Björkly S, Eidhammer G, Lorentzen S, Mjøsund NH, Opjordsmoen S, et al. Mechanical and

pharmacological restraints in acute psychiatric wards-Why and how are they used? Psychiatry research

2013;209(1):91-7.

(152) Sundhedsstyrelsen. Nye tal fra sundhedsstyrelsen. Anvendelse af tvang i psykiatrien 2009. København:


(153) Bremmes R, Lilleeng SE, Pedersen PB, Hellevik V, Bergesen F. Bruk av tvang i psykisk helsevern for

voksne i 2013. Oslo: Helsedirektoratet; 2014. Report No.: IS 2243.

(154) Bowers L, Callaghan P, Clark N, Evers C. Comparisons of psychotropic drug prescribing patterns in acute

psychiatric wards across Europe. Eur J Clin Pharmacol 2004;60:29-35.

(155) Donat DC. An analysis of successful efforts to reduce the use of seclusion and restraint at a public

psychiatric hospital. Psychiatr Serv 2003;54(8):1119-23.

(156) McCue RE, Urcuyo L, Lilu Y, Tobias T, Chambers MJ. Reducing restraint use in a public psychiatric

inpatient service. J Behav Health Serv Res 2004;31(2):217-24.

(157) Pollard R, Yanasak EV, Rogers SA, Tapp A. Organizational and unit factors contributing to reduction in the

use of seclusion and restraint procedures on an acute psychiatric inpatient unit. Psychiatr Q 2007;78(1):73-

81.

(158) van der Schaaf PS, Dusseldorp E, Keuning FM, Janssen WA, Noorthoorn EO. Impact of the physical

environment of psychiatric wards on the use of seclusion. Br J Psychiatry 2013;202(2):142-9.

(159) Virtanen M, Vahtera J, Batty GD, Tuisku K, Pentti J, Oksanen T, et al. Overcrowding in psychiatric wards

and physical assaults on staff: data-linked longitudinal study. Br J Psychiatry 2011;198(2):149-55.

(160) Flannery Jr RB, LeVitre V, Rego S, Walker AP. Characteristics of Staff Victims of Psychiatric Patient

Assaults: 20-Year Analysis of the Assaulted Staff Action Program. Psychiatr Q 2011;82(1):11-21.

(161) Bowers L, Allan T, Simpson A, Jones J, Van Der Merwe M, Jeffery D. Identifying key factors associated

with aggression on acute inpatient psychiatric wards. Issues Ment Health Nurs 2009;30(4):260-71.


67

(162) De Benedictis L, Dumais A, Sieu N, Mailhot MP, Tran MAM, Stikarovska I, et al. Staff Perceptions and

Organizational Factors as Predictors of Seclusion and Restraint on Psychiatric Wards. Psychiatr Serv

2011;62(5):484-91.

(163) Hanrahan NP, Aiken LH, McClaine L, Hanlon AL. Relationship between psychiatric nurse work

environments and nurse burnout in acute care general hospitals. Issues Ment Health Nurs 2010;31(3):198-

207.

(164) Camerino D, Estryn-Behar M, Conway PM, van Der Heijden BIJM, Hasselhorn HM. Work-related factors

and violence among nursing staff in the European NEXT study: a longitudinal cohort study. Int J Nurs Stud

2008;45(1):35-50.

(165) Donat DC. Impact of improved staffing on seclusion/restraint reliance in a public psychiatric hospital.

Psychiatr Rehab J 2002;25(4):413-6.

(166) Fluttert FAJ, Van Meijel B, Nijman H, Bjørkly S, Grypdonck M. Preventing aggressive incidents and

seclusions in forensic care by means of the Early Recognition Method. J Clin Nurs 2010;19(11):1529-37.

(167) Higgins JPT, Green S. Cochrane Handbook for Systematic Reviews of Interventions Version 5.0.0 [updated

February 2008]. The Cochrane Collaboration; 2008.

(168) Hennekens CH, Buring JE. Epidemiology in medicine. Philadelphia: Lippencott Williams and Wilkins;

1987.

(169) Gerolamo AM. Measuring adverse outcomes in inpatient psychiatry: the reliability of nurse recall. Arch

Psychiatr Nurs 2008;22(2):95-103.

(170) Hartman JM, Forsen Jr JW, Wallace MS, Neely JG. Tutorials in clinical research: Part IV: Recognizing and

controlling bias. Laryngoscope 2002;112(1):23-31.

(171) Choi BCK, Pak AWP. A catalog of biases in questionnaires. Prev Chronic Dis 2005;2(1):1-13.

(172) Nijman HLI, Palmstierna T, Almvik R, Stolker JJ. Fifteen years of research with the Staff Observation

Aggression Scale: a review. Acta Psychiatr Scand 2005;111(1):12-21.

(173) Goleman D. Emotional Intelligence. New York: Bantam Books; 1995.

(174) Georgieva I, Mulder CL, Whittington R. Evaluation of behavioral changes and subjective distress after

exposure to coercive inpatient interventions. BMC Psychiatry 2012;12(1):54.

(175) Bergk J, Einsiedler B, Flammer E, Steinert T. A Randomized Controlled Comparison of Seclusion and

Mechanical Restraint in Inpatient Settings. Psychiatr Serv 2011;62(11):1310-7.

(176) Higgins J, Purvis K. A comparison of the Kennedy Axis V and the Global Assessment of Functioning Scale.

J Psychiatr Prac 2000;6(2):84-90.

(177) Henderson C, Flood C, Leese M, Thornicroft G, Sutherby K, Szmukler G. Effect of joint crisis plans on use

of compulsory treatment in psychiatry: single blind randomised controlled trial. BMJ 2004;329(7458):136.

(178) Gaskin CJ, Elsom SJ, Happell B. Interventions for reducing the use of seclusion in psychiatric facilities:

review of the literature. Br J Psychiatry 2007;191:298-303.

(179) Sailas E, Wahlbeck K. Restraint and seclusion in psychiatric inpatient wards. Curr Opin Psychiatry

2005;18(5):555-9.


68

13 Appendix

13.1 Appendix A

Appraisal, Grading, and Combining the Recommendations of the MR-preventive Factors

No

MR-preventive factor Reference

Qu

an

tita

tiv

e ev

iden

ce l

evel

(ef

fect

)

Qu

ali

tati

ve

evid

ence

lev

el

Qu

an

tita

tiv

e re

com

men

da

tio

n g

rad

e

Qu

ali

tati

ve

reco

mm

end

ati

on

gra

de

Co

mb

ined

rec

om

men

da

tio

n g

rad

e

Co

mb

ined

eff

ect

1 Implementation of cognitive

milieu therapy reduces the number

of mechanical restraint episodes,

among other things through

patient involvement and

empowerment.

Lykke J, Austin SF, Morch MM. Cognitive

milieu therapy and restraint within dual

diagnosis populations. Ugeskr Laeger 2008

Jan 28;170(5):339-43.

3

(87%)

4 3 3 1

Thyrsting K, Hall EOC. Prevention of

coercion in psychiatric units: a secondary

analysis of a practitioner research project.

Klinisk sygepleje 2008;22(2):78-86.

2

2 Combined intervention

programmes are effective in

avoiding mechanical restraints e.g.

patient participation, patient

education, staff education,

programmatic changes, high-level

administrative endorsement,

cultural changes, data analysis,

etc.

Fisher WA. Elements of successful restraint

and seclusion reduction programs and their

application in a large, urban, state

psychiatric hospital. Journal of psychiatric

practice 2003;9(1):7.

-2

(67%)

3 N.a. 3 2

Hellerstein DJ, Staub AB, Lequesne E.

Decreasing the use of restraint and seclusion

among psychiatric inpatients. Journal of

psychiatric practice 2007;(5):308-17.

+2

(91%)

Khadivi AN, Patel RC, Atkinson AR,

Levine JM. Association between seclusion

and restraint and patient-related violence.

Psychiatric Services 2004;55(11):1311-2.

4

(52%)

Sullivan AM, Bezmen J, Barron CT, Rivera

J, Curley-Casey L, Marino D. Reducing

restraints: alternatives to restraints on an

inpatient psychiatric service--utilizing safe

and effective methods to evaluate and treat

the violent patient. Psychiatr Q

2005;76(1):51-65.

-2

(72%)

Taxis JC. Ethics and praxis: alternative

strategies to physical restraint and seclusion

in a psychiatric setting. Issues in Mental

Health Nursing 2002;23(2):157-70.

4

(94%)

3 Implementation of patient-centred

care with a higher degree of

patients´ positive involvement in

their own care reduces the number

Fletcher E, Stevenson C. Launching the

Tidal Model in an adult mental health

programme. Nurs Stand 2001 Aug

22;15(49):33-6.

-2

(67%)

4 3 3 2


69

of mechanical restraint episodes. Sullivan D, Wallis M, Lloyd C. Effects of

patient-focused care on seclusion in a

psychiatric intensive care unit. Int J Ther

Rehabil 2004 Nov;11 2004;503-8.

-2

(36%)





2

4 The identification of the patients’

crisis triggers unique calming

techniques and decisions on

specific crisis interventions early

in the admission procedure,

thereby reducing the number of

mechanical restraints.

Jonikas JA, Cook JA, Rosen C, Laris A,

Kim JB. A program to reduce use of

physical restraint in psychiatric inpatient

facilities. Psychiatr Serv 2004

Jul;55(7):818-20.

+2

(50%)

4 3 3 3

McCain M, Kornegay K. Behavioral Health

Restraint: The Experience and Beliefs of

Seasoned Psychiatric Nurses. Journal for

Nurses in Staff Development

2005;21(5):236.

3

Meehan T, McIntosh W, Bergen H.

Aggressive behaviour in the high secure

forensic setting: the perceptions of patients.

Journal of Psychiatric and Mental Health

Nursing 2006;13(1):19-25.

4

Wynn R, Mykleburst LH. Patients´

Satisfaction and Self-Rated Improvement

Following Coercive Interventions.

Psychiatry, Psychology and Law

2006;13(2):199-202.

-2

(–%)

5 Introduction of validated risk

assessment systems including

suggestions of care interventions

reduce the number of mechanical

restraints.

Abderhalden C, Dassen T, Fisher JE,

Halfens RJG, Haug HJ, Needham I.

Structured risk assessment and violence in

acute psychiatric wards: randomised

controlled trial. The British Journal of

Psychiatry 2008;193(1):44.

-1

(27%)

2 N.a. 2 4

Cochrane-Brink KA, Lofchy JS, Sakinofsky

I. Clinical rating scales in suicide risk

assessment. General Hospital Psychiatry

2000;22(6):445-51.

-2

(–%)

Dernevik M, Grann M, Johansson S. Violent

behaviour in forensic psychiatric patients:

Risk assessment and different risk-

management levels using the HCR-20.

Psychology, Crime & Law 2002;8(1):93-

111.

3

(–%)

6 Effective treatment of substance

abuse and psychosis reduces the

number of mechanical restraint

episodes; the treatment of mental

illnesses is also described

elsewhere in the review.

Fletcher E, Stevenson C. Launching the

Tidal Model in an adult mental health

programme. Nurs Stand 2001 Aug

22;15(49):33-6.

-2

(67%)

2 N.a. 4 3

Palmstierna T, Olsson D. Violence from

young women involuntarily admitted for

severe drug abuse. Acta Psychiatrica

Scandinavica 2007;115(1):66-72.

3

(–%)

Sullivan D, Wallis M, Lloyd C. Effects of

patient-focused care on seclusion in a

psychiatric intensive care unit. Int J Ther

Rehabil 2004 Nov;11 2004;503-8.

-2

(36%)





(this paper ads very little to the synthesis)

–


70

7 Mandatory review from qualified

clinical experts, e.g. as post-

incident review minimizes the

number of mechanical restraint

episodes.

Donat DC. Impact of a mandatory

behavioral consultation on

seclusion/restraint utilization in a

psychiatric hospital. Journal of Behavior

Therapy and Experimental Psychiatry 1998

Mar;29(1):13-9.

-2

(62%)

4 N.a. 4 3

Prescott DL, Madden LM, Dennis M, Tisher

P, Wingate C. Reducing mechanical

restraints in acute psychiatric care settings

using rapid response teams. Journal of

Behavioral Health Services & Research

2007 Jan;34(1):96-105.

-2

(36%)

8 Better staff training reduces the

number of mechanical restraints

(longer duration of the training,

more staff participating, use of

different and involving education

methods, higher quality of the

content and greater diversity of

the topics, e.g. legislation, crisis

management, physical

intervention strategies,

antecedents, warning signs, de-

escalation techniques etc.).

Bonner G, Lowe T, Rawcliffe D, Wellman

N. Trauma for all: a pilot study of the

subjective experience of physical restraint

for mental health inpatients and staff in the

UK. Journal of Psychiatric and Mental

Health Nursing 2002;9(4):465-73.

4 4 3 3 4

Bowers L, Nijman H, Allan T, Simpson A,

Warren J, Turner L. Prevention and

Management of Aggression Training and

Violent Incidents on UK Acute Psychiatric

Wards. Psychiatric Services

2006;57(7):1022.

3

(–%)

Delaney KR, Johnson ME. Keeping the Unit

Safe: Mapping Psychiatric Nursing Skills.

Journal of the American Psychiatric Nurses

Association 2006 Aug 1;12(4):198-207.

2

Hahn S, Abderhalden C, Duxbury J, Halfens

RJG, Needham I. The effect of a training

course on mental health nurses' attitudes on

the reasons of patient aggression and its

management. Journal of Psychiatric &

Mental Health Nursing 2006;13(2):197-204.

-2

(–%)





Jul;55(7):818-20.

+2

(34%)

Lee S, Wright S, Sayer J, Parr A, Gray R,

Gournay K. Physical restraint training for

nurses in English and Welsh psychiatric

intensive care and regional secure units.

Journal of Mental Health 2001

Apr;10(2):151-62.

-2

(–%)

Mackay I, Paterson B, Cassells C. Constant

or special observations of inpatients

presenting a risk of aggression or violence:

nurses' perceptions of the rules of

engagement. Journal of Psychiatric &


4

Needham I, Dassen T, Halfens RJG, Haug

HJ, Fisher JE. The effect of a training

course in aggression management on mental

health nurses perceptions of aggression: a

cluster randomised controlled trial.

International Journal of Nursing Studies

2005;42(6):649-55.

-1

(–%)

Spokes K, Bond K, Lowe T, Jones J,

Illingworth P. HOVIS -

Hertfordshire/Oxfordshire violent incident

study. Journal of Psychiatric and Mental

Health Nursing 2002;9(2):199-209.

4


71

Wright S, Sayer J, Parr AM, Gray R,

Southern D, Gournay K. Breakaway and

physical restraint techniques in acute

psychiatric nursing: Results from a national

survey of training and practice. Journal of

Forensic Psychiatry & Psychology

2005;16(2):380-98.

-2

(–%)

9 Staff attitudes such as positive

care for the patients, respect,

collaboration, communication,

humanism, empathy and less

control and involuntary

procedures, reduce the number of


Allen MH, Carpenter D, Sheets JL, Miccio

S, Ross R. What do consumers say they

want and need during a psychiatric

emergency? J Psychiatr Pract 2003

Jan;9(1):39-58.

4

(–%)

4 4 4 4

Bowers L, Brennan G, Flood C, Lipang M,

Oladapo P. Preliminary outcomes of a trial

to reduce conflict and containment on acute

psychiatric wards: City Nurses. Journal of

Psychiatric & Mental Health Nursing

2006;13(2):165-72.

-2

(27%)

Bowers L, Alexander J, Simpson A, Ryan

C, Carr-Walker P. Student psychiatric

nurses' approval of containment measures:

relationship to perception of aggression and

attitudes to personality disorder.

International Journal of Nursing Studies

2007;44(3):349-56.

3

(–%)

Cardell R, Pitula CR. Suicidal Inpatients'

Perceptions of Therapeutic and

Nontherapeutic Aspects of Constant

Observation. Psychiatric Services

1999;50(8):1066.

4

Duxbury J. An evaluation of staff and

patient views of and strategies employed to

manage inpatient aggression and violence

on one mental health unit: a pluralistic

design. Journal of Psychiatric and Mental


3

(–%)

Gillig PM, Markert R, Barron J, Coleman F.

A Comparison of Staff and Patient

Perceptions of the Causes and Cures of

Physical Aggression on a Psychiatric Unit.

Psychiatric Quarterly 1998;69(1):45-60.

3

(–%)





Nursing 2006;13(1):19-25.

4

Robins CS, Sauvageot JA, Cusack KJ,

Suffoletta-Maierle S, Frueh BC. Special

Section on Seclusion and Restraint:

Consumers' Perceptions of Negative

Experiences and" Sanctuary Harm" in

Psychiatric Settings. Psychiatric Services

2005;56(9):1134.

4





Health Nursing 2002;9(2):199-209

4

10 Higher staff-patient ratio reduces

the number of mechanical

restraints.

Smith GM, Davis RH, Bixler EO, Lin HM,

Altenor A, Altenor RJ, et al. Special section

on seclusion and restraint: Pennsylvania

State Hospital System's seclusion and

restraint reduction program. Psychiatric

Services 2005;56(9):1115.

3

(22%)

4 N.a. 4 4


72

Wynn R. Staff's attitudes to the use of

restraint and seclusion in a Norwegian

university psychiatric hospital. Nordic

Journal of Psychiatry 2003;57(6):453-9.

3

(–%)

11 The higher educated staff, the

fewer episodes of mechanical

restraints.




Association 2006 Aug 1;12(4):198-207.

2 4 3 3 5

Flannery Jr RB, Stone P, Rego S, Walker

AP, Farley E. Characteristics of Staff

Victims of Patient Assault: 15-year Analysis

of the Assaulted Staff Action Program

(ASAP). Psychiatric Quarterly 2006.

-2

(–%)

Gim YS, Cheng SOB, Hong TL, Chin

MWB, Hen WH, Lee YJ. Nurses

experiences in identifying and managing

violence precipitants in a psychiatric in-

patient setting. Singapore Nursing Journal

1999;26(1):21-4.

3

(–%)







4





Health Nursing 2002;9(2):199-209.

4

12 The more experienced staff the

fewer episodes of mechanical

restraints.

Bowers L, Jeffery D, Simpson A, Daly C,

Warren J, Nijman H. Junior staffing changes

and the temporal ecology of adverse

incidents in acute psychiatric wards. Journal

of Advanced Nursing 2007;57(2):153-60.

3

(–%)

4 3 3 5




Association 2006 Aug 1;12(4):198-207.

2






-2

(–%)

Gim YS, Cheng SOB, Hong TL, Chin

MWB, Hen WH, Lee YJ. Nurses

experiences in identifying and managing

violence precipitants in a psychiatric in-

patient setting. Singapore Nursing Journal

1999;26(1):21-4.

3

(–%)







4





Health Nursing 2002;9(2):199-209.

4


73

13 Limit-setting. If nurses change

their focus from considering the

patient as deviant to being a

resource in their own treatment it

would decrease the number of

mechanical restraint.

Vatne S, Fagermoen MS. To correct and to

acknowledge: two simultaneous and

conflicting perspectives of limit-setting in

mental health nursing. J Psychiatr Ment

Health Nurs 2007 Feb;14(1):41-8.

4 N.a. 3 3 5

14 Early and adequate administration

of oral p.n. medication guided by

the individual patients´ problems,

medication experiences, and

preferences leads to fewer

mechanical restraint episodes.

Allen MH, Carpenter D, Sheets JL, Miccio

S, Ross R. What do consumers say they

want and need during a psychiatric

emergency? J Psychiatr Pract 2003

Jan;9(1):39-58.

4

(–%)

4 4 4 5

Goedhard LE, Stolker JJ, Nijman HLI,

Egberts ACG, Heerdink ER. Aggression of

Psychiatric Patients Associated with the Use

of As-needed Medication.

Pharmacopsykiatry 2007;40(1):25.

-2

(–%)

McCain M, Kornegay K. Behavioral Health

Restraint: The Experience and Beliefs of

Seasoned Psychiatric Nurses. Journal for

Nurses in Staff Development

2005;21(5):236.

3





Nursing 2006;13(1):19-25.

4

Thapa PB, Palmer SL, Owen RR, Huntley

AL, Clardy JA, Miller LH. P.R.N. (As-

needed) orders and exposure of psychiatric

inpatients to unnecessary psychotropic

medications. Psychiatr Serv 2003

Sep;54(9):1282-6.

3 (–50%)

15 Debriefing, defusing and crisis

intervention minimize the number

of mechanical restraint episodes.







4 4 5 4 5

Bowers L, Simpson A, Eyres S, Nijman H,

Hall C, Grange A, et al. Serious untoward

incidents and their aftermath in acute

inpatient psychiatry: the Tompkins Acute

Ward study. Int J Ment Health Nurs 2006

Dec;15(4):226-34.

4






-2

(–%)





Jul;55(7):818-20.

+2

(15%)

16 Patients´ response to rules can be

positive and lead to fewer cases of

mechanical restraint. For instance,

if staff helps patients to

Alexander J. Patients' feelings about ward

nursing regimes and involvement in rule

construction. J Psychiatr Ment Health Nurs

2006 Oct;13(5):543-53.

3

(–%)

4 5 4 5


74

understand not only the rules, but

also the rationale for the presence

of such rules. If the staff manages

to make patients understand that

the rules are imposed out of

concern for their welfare, then

patients may be less reluctant to

following these rules.

Duxbury J. An evaluation of staff and

patient views of and strategies employed to

manage inpatient aggression and violence

on one mental health unit: a pluralistic

design. Journal of Psychiatric and Mental

Health Nursing 2002;9(3):325-37 (this

paper ads only a little to the synthesis).

3

(–%)

Gillig PM, Markert R, Barron J, Coleman F.

A Comparison of Staff and Patient

Perceptions of the Causes and Cures of

Physical Aggression on a Psychiatric Unit.

Psychiatric Quarterly 1998;69(1):45-60.

3

(–%)





Nursing 2006;13(1):19-25.

4

Robins CS, Sauvageot JA, Cusack KJ,

Suffoletta-Maierle S, Frueh BC. Special

Section on Seclusion and Restraint:

Consumers' Perceptions of Negative

Experiences and" Sanctuary Harm" in

Psychiatric Settings. Psychiatric Services

2005;56(9):1134.

4

17 Activity-based nursing

interventions e.g. physical

activities are effective in avoiding






Nursing 2006;13(1):19-25.

4 4 5 4 5

Thomas B, Jones M, Johns P, Trauer T. Prn

medication use in a psychiatric high-

dependency unit following the introduction

of a nurse-led activity programme.

International Journal of Mental Health

Nursing 2006;15(4):266.

+2

(–%)

18 Several mental and physical

working environmental factors

influence positively the number of

mechanical restraints e.g.: Good

management as a high degree of

order, organisation or support,

Adequate staffing, Psychiatric

training, Rotation, Education, No

crowding, Emergency backup, etc.

Several of these factors are

described elsewhere in the review.







4 4 5 4 5

Bowers L, Allan T, Simpson A, Jones J,

Whittington R. Morale is high in acute

inpatient psychiatry. Soc Psychiatry

Psychiatr Epidemiol 2008 Jul 5.

-2

(–%)

Daffern M, Mayer M, Martin T. Staff

gender ratio and aggression in a forensic

psychiatric hospital. International Journal of


3

(–%)

Owen C, Tarantello C, Jones M, Tennant C.

Violence and Aggression in Psychiatric

Units. Psychiatric Services

1998;49(11):1452.

4

(–%)

Whittington D, McLaughlin C. Finding time

for patients: an exploration of nurses' time

allocation in an acute psychiatric setting.

Journal of Psychiatric & Mental Health

Nursing 2000;7(3):259-68.

3

(–%)

19 Less crowding e.g. less stimuli,

less environmental stress and

more personal space decrease the

number of mechanical restraints.





Nursing 2006;13(1):19-25.

4 4 5 4 5


75

Vaaler AE, Morken G, Fløvig JC, Iversen

VC, Linaker OM. Effects of a psychiatric

intensive care unit in an acute psychiatric

department. Nordic Journal of Psychiatry

2006;60(2):144-9.

-2

(–%)

20 Higher regular staff and lower

substitute ratio reduce the number

of mechanical restraints.







4 4 5 4 5

Bowers L, Allan T, Simpson A, Nijman H,

Warren J. Adverse Incidents, Patient Flow

and Nursing Workforce Variables on Acute

Psychiatric Wards: The Tompkins Acute

Ward Study. International Journal of Social

Psychiatry 2007;53(1):75.

3

(–%)





Health Nursing 2002;9(2):199-209.

(this paper ads only a little to the synthesis).

4

21 Staffs´ opposition towards

mechanical restraint, due to the

perception that mechanical

restraint is ethically problematic,

decrease the number of


Lind M, Kaltiala-Heino R, Suominen T,

Leino-Kilpi H, Valimaki M. Nurses'ethical

perceptions about coercion. Journal of

Psychiatric & Mental Health Nursing

2004;11(4):379-85.

-2

(–%)

4 N.a. 4 5

Van Doeselaar M., Sleegers P,

Hutschemaekers G. Professionals' attitudes

toward reducing restraint: the case of

seclusion in the Netherlands. Psychiatr Q

2008 Jun;79(2):97-109.

-2

(–%)

22 Alarm systems e.g. video

surveillance and radio systems and

well trained response teams,

reduce the number of mechanical

restraints.

D'Orio BM, Purselle D, Stevens D, Garlow

SJ. Reduction of episodes of seclusion and

restraint in a psychiatric emergency service.

Psychiatric Services 2004;55(5):581-3.

+2

(16%)

4 N.a. 4 5

23 Better care-alliance between

patients and the nursing staff

reduces the number of mechanical

restraints.

Wynn R. Staff's attitudes to the use of

restraint and seclusion in a Norwegian

university psychiatric hospital. Nordic

Journal of Psychiatry 2003;57(6):453-9.

-2

(–%)

4 N.a. 4 5

24 Communication especially

communicative de-escalation

reduces the number of mechanical

restraint episodes.

Duxbury J, Whittington R. Causes and

management of patient aggression and

violence: staff and patient perspectives.

Journal of Advanced Nursing 2005

Jun;50(5):469-78.

-3

(–%)

4 N.a. 4 5

25 Well-maintained and familiar

surroundings reduce the number


Vaaler AE, Morken G, Linaker OM. Effects

of different interior decorations in the

seclusion area of a psychiatric acute ward.

Nordic Journal of Psychiatry 2005;59(1):19-

24.

+2

(–%)

4 N.a. 4 5

26 Music as diversion and relaxation

before and after mechanical

restraint reduces the number of


Janelli LM, Kanski G. Music for untying

restrained patients. Journal of the New York

State Nurses Association 1998;29(1):13-5.

4

(–%)

4 N.a. 4 5

27 Separation of acutely disturbed

patients reduces the number of






Nursing 2006;13(1):19-25.

(This factor is graded, in the review as 5, it

should have been 4).

4 N.a. 5 5 5

No refers to the number in the hierarchy for which 1 is highest/best.


76

Factor refers to synthesized description of the MR-preventive factor. The final synthesis was guided by the following three

questions: Which interventions match patients’ views and experiences? Which interventions have yet to be evaluated in clinical

settings? Does the effect of the intervention seem plausible from a clinical point of view?

Reference refers to the literature reference in Vancouver style.

Quantitative evidence level describes the appraisal of individual quantitative papers. Appraising the quantitative research papers was

conducted as outlined by the Danish Reference Programme Secretariat (Sekretariatet for Referenceprogrammer, 2004), Method for

Evaluating Research and Guideline Evidence (MERGE; Liddle et al., 1996), and SIGN (Harbour, 2008); 1++, 1+, 1-, 2++, 2+, 2-, 3,

4, where 1++ is the best level. Checklists for evaluating the methodological quality of the papers can be acquired from the authors. In

brackets, the reduction in the number of mechanical restraints (in a few cases combined MR and seclusion rates) is displayed as a

percentage, if the reduction was caused by more than one intervention, it was downgraded accordingly. – describes no available

figures.

Qualitative evidence level describes the appraisal of individual qualitative papers. Appraising the qualitative research papers was

conducted as outlined by Sandelowski and Barroso (2007). The papers were rated according to the complexity levels by Kearney in

which 1 = dense explanatory description, 2 = depiction of experimental variation, 3 = shared pathway or meaning, 4 = descriptive

categories, 5 = findings restricted by a priori frameworks (level 5 papers were excluded), 1 is the “best” level. The checklist for

evaluating the methodological quality of the papers can be required from the author.

Quantitative recommendation grade from 1 to 5, in which 1 is the highest/best. N.a. refers to no available paper. The hierarchy of

the quantitative grading was developed from the DRPS (2004, NERGE (1996), and SIGN (2008).

Qualitative recommendation grade from 1 to 5, for which 1 is the highest/best. N.a. referred to no available paper. The hierarchy of

the qualitative grading can be found below.

Combined recommendation grade describes the whole body of evidence regarding the level of scientific “truth” or its potential for

clinical application of a specific intervention ranked from 1 to 5, for which 1 is the highest/best. Combining the quantitative and

qualitative recommendations was conducted by taking the better of the two.

Combined effect refers to the reduction in the number of MRs, from 1 to 5, with 1 being the best (80% to 100% reduction), 2

representing a reduction of 60% to 80% and so on. These were combined as follows: no paper, the level was set to 5 (0–20%

reduction); one paper, the level was directly transferred; two or more papers, the level was influenced by: (1) the paper with the best

recommendation grade, (2) the paper best describing the intervention, and (3) a simple average.

Assessment system for quantitative research

Evaluation criteria for the study (paper)

(MERGE-1996)

Evaluation criteria are coded according to the extent to which the criteria in the specific checklist

are fulfilled: Code

Criterion entirely fulfilled. a

Criterion mostly fulfilled. b1

Criterion mostly not fulfilled. b2

Criterion not at all fulfilled. c

Criterion not described adequately to classify as a, b1, b2 or c. ?

Criterion not applicable. n/a

Overall assessment of the study (paper)

(DRPS-2004)

This table is used to assess the general methodological quality of the study, in three grades: Code

All or the most criterion are fulfilled. Criterion witch are not fulfilled, will very rarely change

the conclusions of the study.

++

Some criterions are fulfilled. Criterion witch are not fulfilled or thorough described, will rarely

change the conclusions of the study.

+

Few or none criterions are fulfilled. The conclusions of the study can easily change. -

Levels of evidence for quantitative studies

(Insp. from SIGN-2008, MERGE-1996 and DRPS-2004)

1++ High-quality meta-analyses, systematic reviews of RCTs, or RCTs with a very low risk of bias

1+ Well-conducted meta-analyses, systematic reviews of RCTs, or RCTs with a low risk of bias

1- Meta-analyses, systematic reviews of RCTs, or RCTs with a high risk of bias or Controlled trials

without randomisation.

2++ High-quality systematic reviews of case–control or cohort studies

High-quality case–control or cohort studies with a very low risk of confounding, bias or chance and a


77

high probability that the relationship is causal.

2+ Well-conducted case–control or cohort studies or multiple time series or before and after studies

(Preferably from more than one centre or research group) with a low risk of confounding, bias or

chance and a moderate probability that the relationship is causal. Diagnostic tests (diagnostic

method).

2- Case–control or cohort studies or multiple time series or before and after studies with a high risk of

confounding, bias, or chance and a significant risk that the relationship is not causal. High quality

descriptive studies. Diagnostic tests (Nosographic method).

3 Moderate quality descriptive studies.

Case reports, case series, decision analyses.

4 Expert opinion. Formal consensus. Low quality descriptive studies.

Grades of recommendation for quantitative research

(Insp. from SIGN-2008, MERGE-1996 and DRPS-2004)

1 At least one meta-analysis, systematic review or RCT rated as 1++, and directly applicable to the

target population, or a systematic review of RCTs or a body of evidence consisting principally of

studies rated as 1+, directly applicable to the target population and demonstrating overall consistency

of results.

2 A body of evidence including studies rated as 1- and 2++, directly applicable to the target population

and demonstrating overall consistency of results, or extrapolated evidence from studies rated as 1++

or 1+.

3 A body of evidence including studies rated as 2+, directly applicable to the target population and

demonstrating overall consistency of results, or extrapolated evidence from studies rated as 2++.

4 A body of evidence including studies rated as 2- or 3, directly applicable to the target population and

demonstrating overall consistency of results, or extrapolated evidence from studies rated as 2+.

5 Opinions of respected authorities, based on clinical experience, or reports of expert committees rated

as 4.

Assessment system for qualitative research

Evaluation criteria for the study (paper)

(Sandelowski and Barroso – 2007)

Evaluation criteria are coded according to the extent to which the criteria in the specific checklist

are fulfilled: Code

Presence +/-

Relevance +/-

Overall assessment of the study (paper)

(Sandelowski and Barroso – 2007)

This table is used to evaluate the quality of the study: Code

Acceptable (Signal > noise) +

Questionable (Noise > signal) -

Levels of evidence for qualitative studies

(Kearny – 2004)

No Description Notes

1 Dense explanatory

description

The highest level of complexity and discovery might be termed the qualitative gold

standard. These findings may be seen as representing the characteristics of the “experiential

variation” category, when achieved to the highest degree, but in doing so, these findings

appear qualitatively different. Clifford Geertz (1973) exemplified this level of findings in

his definition of “thick description,” but this level of explanatory detail can be seen outside

ethnography as well. These findings are a rich evocation of a situated understanding of a

multifaceted and varied human phenomenon in a unique situation. With dense factual and

descriptive detail, they portray the full depth and range of complex influences that propel


78

persons to make one choice over another, to speak one way versus another, and to view life

one way rather than another. Physical and social context is colorfully conveyed, and the

experience of time is captured at the levels of social, cultural, and political history as well

as at the level of personal progress. The role of the researcher in this particular context and

interaction also is clearly apparent. Dense explanatory description may be achieved using a

number of qualitative methods including ethnography, phenomenology, and grounded

theory.

2 Depiction of

experiential

variation

An even greater degree of qualitative complexity is achieved in findings that not only

describe the main pathway or essence of an experience but also portray how that

experience or pathway varies depending on individuality and context. Portraying or

explaining variation in a human experience requires considerable breadth and depth of

sampling and data collection and a high level of analytic expertise, which can produce a

high degree of discovery of new insights or perspectives on human phenomena. Although

grounded theories are the type of findings most commonly thought of as portraying

variation, this level of complexity can be seen in other approaches as well. The rich detail

in a fully realized phenomenology or ethnography can capture a variety of viewpoints and

realizations of a human experience and the contextual sources of that variety, whether

political, cultural, familial, or intrapersonal.

3 Shared pathway or

meaning

The third category shows an increase in complexity. Here, the investigator's interpretation

produces a synthesis of a shared experience or process. It is distinguished from the

previous grouping by the investigator's integration of concepts or themes into a linked and

logical portrayal. The commonality, core, or essence of the experience is captured by the

analyst. This synthesis has the potential to reveal something previously undescribed about

the phenomenon that would not be readily apparent in a series of unlinked categories.

The increased complexity enables greater discovery. Shared pathways or meanings can be

seen in findings achieved with a variety of qualitative approaches. For example, in a

qualitative description, data clusters are linked in a holistic picture of the experience. In a

grounded-theory study, concepts are connected in a model of influences and strategies or

actions, with each relationship substantiated by data. In a phenomenological study, themes

are experiential components integrated into a narrative depiction of a multifaceted

phenomenon. The analyst has moved from describing parts of a data set to explaining how

these are components of a larger social or experiential whole.

4 Descriptive

categories

The simplest level of complexity of qualitative evidence, but one that nonetheless may

portray discovery is a series of labelled data categories. Clusters of data are labelled with

brief headings that indicate the topic or type of data contained therein, as in a book's table

of contents. Content analysis techniques often produce findings in this format.

5 Findings restricted

by a priori

frameworks

Findings that are produced by applying an existing set of ideas to qualitative data without

identifying new insights or enriching, extending, or revising existing theory may offer a

certain degree of complexity but little discovery, and consequently they provide little or no

evidence for practice.

Grades of recommendation for qualitative research

(Insp. by Sandelowski and Barroso – 2007, Kearney – 2004, MERGE – 1996, DRPS – 2004, and SIGN – 2008)

1 At least one high quality meta-synthesis, systematic review or meta-summary rated as 1 and directly applicable

to the target population and context, or a body of evidence consisting principally of papers rated as 1, directly

applicable to the target population and context and demonstrating overall consistency of findings.

2 A body of evidence including papers rated as 2 or one paper rated as 1, directly applicable to the target

population and context and demonstrating overall consistency of findings, or extrapolated evidence from papers

rated as 1.


population and context and demonstrating overall consistency of findings, or extrapolated evidence from papers

rated as 2.


population and context and demonstrating overall consistency of findings.

5 One or two papers rated as 4.

Paper 1

Mechanical Restraint

Which Interventions Prevent Episodes of Mechanical Restraint?

A Systematic Review.

Mechanical Restraint—Which Interventions Prevent Episodesof Mechanical Restraint?—A Systematic Reviewppc_307 83..94

Jesper Bak, RN, SD, MPH, Stud. PhD, Mette Brandt-Christensen, MD, PhD,Dorte Maria Sestoft, MD, PhD, and Vibeke Zoffmann, RN, MPH, PhD

Jesper Bak, RN, SD, MPH, Stud. PhD, is a Head Nurse, Mental Health Centre Sct. Hans, Roskilde, Denmark; Mette Brandt-Christensen, MD, PhD, isEmployed Senior Consultant, Forensic Psychiatric Centre Glostrup, Copenhagen, Denmark; Dorte Maria Sestoft, MD, PhD, is Head of Department,Ministry of Justice, Clinic of Forensic Psychiatry, Copenhagen, Denmark; and Vibeke Zoffmann, RN, MPH, PhD, is a Senior Researcher, Steno DiabetesCenter, Gentofte, Denmark.

Search terms:Mechanical restraint, mental disorder, nursing,psychiatry, systematic review

Author contact:[email protected], with a copy to theEditor: [email protected]

Conflict of Interest StatementThe authors report no actual or potentialconflicts of interest. No external or intramuralfunding was received.

First Received August 11, 2010; Final Revisionreceived February 10, 2011; Accepted forpublication February 18, 2011.

doi: 10.1111/j.1744-6163.2011.00307.x

PURPOSE: To identify interventions preventing mechanical restraints.DESIGN AND METHODS: Systematic review of international research papersdealing with mechanical restraint. The review combines qualitative and quantitativeresearch inanewway,describingthequalityof evidenceandtheeffectof intervention.FINDINGS: Implementation of cognitive milieu therapy, combined interventions,and patient-centered care were the three interventions most likely to reduce thenumber of mechanical restraints.PRACTICE IMPLICATIONS: There is a lack of high-quality and effective interven-tion studies. This leaves patients and metal health professionals with uncertaintywhen choosing interventions in an attempt to prevent mechanical restraints.

In some countries, mechanical restraint is performed accord-ing to the law when psychiatric inpatients pose a risk to them-selves or to others (Indenrigs- og Sundhedsministeriet, 2006).Other countries do not allow the use of mechanical restraint;for example, in the United Kingdom, only the use of seclusionand holding (physical restraint) is allowed (except in excep-tional circumstances in special hospital environments).Therefore, what constitutes the use of mechanical restraintdiffers a lot across Europe and the rest of the world. Nursingstaff has the primary role in carrying out the practical proce-dure. Leather or fiber belts are used to fixate the patient to thebed, occasionally together with straps around the hands andfeet. In this paper, mechanical restraint is defined: The use ofrestraining straps, belts, or other equipment to restrict move-ment, as suggested by Bowers, van der Werf, et al. (2007). Eventhough mechanical restraint is used within the law, thoroughethical considerations regarding patients’ right of self-determination and human rights, in general, should be madeon how much this intervention should be used. The EuropeanCommittee for the Prevention of Torture and Inhuman orDegrading Treatment or Punishment has, after visits toDenmark, even stated that applying instruments of physicalrestraint to psychiatric patients for days cannot have anymedical justification and amounts to ill treatment (The Euro-

pean Committee for the Prevention of Torture and Inhumanor Degrading Treatment or Punishment, 2002; Hauksson,2008). Physically, applying mechanical restraints is a riskyintervention where patients can develop deep vein throm-boses, pulmonary embolisms, and lactic acidosis, whichcould contribute to cardiovascular collapse, broken bones,etc. (Chan, Neuman, Clausen, Eisele, & Vilke, 2004; Hem,Steen, Opjordsmoen, & Moussaoui, 2001; Hick, Smith, &Lynch, 1999; Laursen, Jensen, Bolwig, & Olsen, 2005; Morri-son & Sadler, 2001; Nielsen, 2005). Against this are also thenegative psychological impact of mechanical restraint onpatients feeling helpless, angry, trapped, sad, fearsome,unpleasant, or offended, and patients often regard it as aninjustice or see it as punishment, in the worst case developinglifetime trauma and possibly post-traumatic stress disorder(Bower, McCullough, & Timmons, 2003; Cusack, Frueh,Hiers, Suffoletta-Maierle, & Bennett, 2003; Frueh et al., 2005;Steinert, Bergbauer, Psych, Schmid, & Gebhardt, 2007). Thesenegative consequences are inconsistent with healthcare pro-viders’ basic attitudes regarding“doing good”for the patients.Weighty arguments for using mechanical restraint are,however, that staff members have to allow for the patient’sown health and life, as well as the health, lives, and rights ofothers, for example, violence toward fellow patients, relatives,

Perspectives in Psychiatric Care ISSN 0031-5990

83Perspectives in Psychiatric Care 48 (2012) 83–94© 2011 Wiley Periodicals, Inc.

mailto:[email protected]


and staff. Altogether, the dilemma is not easily solved, but it iswidely agreed that the number of mechanical restraintsshould be reduced as much as possible (Sundhedsstyrelsen,2009).

Scientists and clinicians agree that many factors influencethe complex phenomenon of mechanical restraint. Thefactors, which influence the aggressive or self-destructivebehavior of the patient and, secondarily, the frequency ofrestraints inside a psychiatric ward, could be relative to thepatient, ward, and staff, or to“situational/interactional condi-tion” (Nijman, Campo, Ravelli, & Merckelbach, 1999). Manyof these factors are related to daily activities when socializingwith other persons in the psychiatric ward. Daily living is thearea where nurses and other nursing staff members have acrucial role in reducing the use of mechanical restraint. Also,the rate of coercive interventions is, in many countries, con-sidered an indicator of the quality of psychiatric inpatienttreatment (Institut for Kvalitet og Akkreditering i Sundheds-væsenet, 2009). However, research into the type of nursinginterventions required, to prevent episodes of mechanicalrestraint, is scarce (Leitner, Barr, Jones, McGuire, & Whitting-ton, 2006; National Institute for Health and Clinical Excel-lence [NICE], 2004, 2005).

Numerous actions are recommended in order to reduce theuse of mechanical and physical restraint: risk management;physical environment and alarm systems; staff attitudes andcare alliance; prediction: antecedents, warning signs, and riskassessment; training, education, skills, and competencies;working with psychiatric patients; de-escalation methods andtechniques; post-incident review; and staff environmentalfactors, both mental and physical working environment;medication; effective treatment of illness and abuse; effectiveobservation; and physical training (Department of Health,2007; NICE, 2005). Though, locally, some of these actionshave been shown to reduce the numbers of restraint (Fisher,2003; Hellerstein, Staub, & Lequesne, 2007), the nationwidefigures still fail to decrease in some countries (Bremmes,Hatling, & Bjørngaard, 2008; Sundhedsstyrelsen, 2008).

Purpose

In order to provide a basis for choosing and developingnursing interventions, under which the number of mechani-cal restraint episodes are decreased, the following will bereviewed:

Which conditions in nursing and which nursinginterventions have been shown to reduce the frequencyof mechanical restraint episodes?

Design and Methods

It was recognized very early during the process that only veryfew randomized clinical trials existed, and no meta-analyses

were found (Muralidharan & Fenton, 2006). Therefore,searches were not limited to these study designs.Also, many ofthe areas under investigation could not be covered fromquantitative papers solely. Therefore, we developed a way tocombine quantitative and qualitative papers into ranked rec-ommendations in order to deduce maximum informationfrom the available papers.

Theory

The method was based on principles outlined by the DanishReference Programme Secretariat (Sekretariatet for Referen-ceprogrammer, 2004) and NICE (2007) involving evaluationof the quantitative research; by Sandelowski and Barroso(2007) and Kearney (2001) involving evaluation of the quali-tative research; and by Scottish Intercollegiate GuidelinesNetwork (SIGN; Harbour, 2008) inspiring the process of thecombined evaluation of both qualitative and quantitativeresearch. The combination of principles from the mentionedsources has been necessary because no one in itself coveredboth quantitative and qualitative research quantification. Thestages of the review process can be found in Table 1.

Inclusion Criteria

Original peer-reviewed papers, covering the care of adult psy-chiatric inpatients who have been physically restrained, wereincluded in the review. The interventions and conditionscovered in the review can be found in Table 2.

Exclusion Criteria

Reviews were excluded to avoid some papers to be includedtwice or more, spoiling the calculations of effect and recom-

Table 1. Stages of the Review Process

Developing scope of the review.Developing review questions with clinical focus, and inclusion andexclusion criteria.Developing search strategy and search phrases.Identifying sources of evidence.Carrying out the searches.Sifting the evidence.Extracting relevant data from studies meeting methodological andclinical criteria.Evaluating methodological quality using relevant checklists.Compiling papers in relevant evidence tables.Interpreting and grading each paper.Abstracting interventions and summarizing intervention levels.Grading the recommendations for the quantitative and qualitativeevidence separately.Summarizing the whole body of evidence.Drawing up presentation of the result.Formulating conclusions and implications for further research.Drawing up the review article.

Mechanical Restraint—Which Interventions Prevent Episodes of Mechanical Restraint?—A Systematic Review

84 Perspectives in Psychiatric Care 48 (2012) 83–94© 2011 Wiley Periodicals, Inc.

mendation. Papers describing interventions and conditionsto prevent physical restraint in nonpsychiatric inpatient set-tings, learning disability settings, nursing homes, and prisonswere excluded.

Time Frame

The study included papers published after 1998. The paperswere obtained between August and November 2008, with anupdate in April 2009.

Search Strategies

The searches included English, Danish, Norwegian, andSwedish language citations, all with English abstracts.

Key words used in different combinations and inflectionsin the search were active, admitted, adolescence, affect, aged,aggression, alarm, and alliance, and the antecedents wereanxiety, appreciative, atmosphere, attitude, bipolar, careplan, check, child, committed, communication, coping,debriefing, de-escalation, delusion, depression, detained,diffusion, developmental disability, disorder, diversion, dualdiagnosis, education, elder, environmental therapy, forensic,hospitalised, house rule, hypo, hysterical, impulse, inpatient,intervention, learning disability, manage, manic, medica-tion, mental health, milieu, mood, music, nonaversive,

nonconfrontational, nursing, nursing home, nursing theory,observation, old, one on one, one to one, outpatient, panic,paranoid, personality disorder, pharmacology, physical,prediction, prevention, prison, p.r.n./p.n. (pro re nata/pro necessitate—as needed or as the situation arises)medication, psychiatry, rapid tranquillisation, rehabilita-tion, reorganizing, restraint, review, risk assessment, schizo-phrenia, seclusion, strategy, surveillance, talk down, therapy,time out, training, treatment, violence, warning signs, andyoung.

The structure of the search strategy is:

• An overarching strategy for conditions and interventionsacross selected databases• A search to identify guidance and reports, for example, gov-ernmental reports, not indexed in the major databases• A topic-specific search strategy on PubMed• For each condition or intervention, evidence of effective-ness or harm was sought• Manual searching was performed in key areas not electroni-cally indexed (e.g., PhD thesis)• Reference lists of the most relevant papers were checked forarticles not otherwise found.

One example from the searches is provided below inTable 3; the whole search strategy, from the 32 databases, canbe required from the authors.

Table 2. Interventions and Conditions Covered

Care Pathways (MR Sequence)

Fundamental causes (psychosis,conflicts, abuse)

Negative aggressive feelings.Toward the patienthimself/herself or towardothers.

Negative aggressive actions.Toward the patienthimself/herself or towardothers.

The mechanical restraint. Framework andaftercare/prevention.

Interventions and Conditions

Fundamental causes:• Degree of illness and

substance abuse• Working with psychiatric patients,

for example, in environmenttherapy or using nursing theory

• Background variables such asdiagnosis, gender distribution,ethnicity, types of staff beinginjured, reasons for beingmechanically restrained, etc.

Negative aggressive feelings:• Patients’ and relatives’

knowledge, feelings, andattitudes

• Staff attitudes and carealliance

• Prescription and use of p.n.medication, administeredby nurses

Negative aggressive actions:• Prediction such as

antecedents, warningsigns, and risk assessment

• De-escalation methods andtechniques

• Observation• Limit setting

The mechanical restraint:• Alarm systems• Physical environment• Physical staff training• Debriefing, defusing, or post-incident

reviews (patients and staff)• Potential health risk because of MR• Staff training, education, skills, and

competencies• Staff environmental factors, both mental

and physical working environment

MR, mechanical restraint; p.n., pro necessitate (as needed).

Table 3. Example of a Search String

CINAHL and AMED/EBSCOhost: (Restrained OR Restraint OR Restraints OR Secluded OR Seclusion OR (“restraint”)) or (MH “Psychiatry”) not(Nursing-home OR Prison OR Learning-disability OR Learning-disabilities OR Rehabilitation OR Outpatient OR developmental-disability ORdevelopmental-disabilities), Limiters Applied, Publication Year from: 1998-; Inpatients; Age Groups: Adult, 19–44 years, Middle Age, 45–64 years.



Checklists and Evidence Tables

Checklists for evaluating the methodological quality of thepapers and evidence tables for compiling studies can berequired from the authors (Critical Appraisal Skills Pro-gramme, 2006; Health Care Practice Research and Develop-ment Unit, 2005; Kearney, 2001; Liddle, Williamsom, &Irwing, 1996; NICE, 2007; Sandelowski & Barroso, 2007; Sek-retariatet for Referenceprogrammer, 2004).

Evidence Grading

Quantitative Research Grading. Once individual papers hadbeen assessed for methodological quality and relevance interms of the clinical questions, they were graded accordingto the levels of evidence. The available evidence for eachintervention/condition was then compiled into tables.Grading of quantitative research papers was conducted asoutlined by the Danish Reference Programme Secretariat(Sekretariatet for Referenceprogrammer, 2004). Because ofthe small number of eligible quantitative studies, it was notpossible to perform meta-analyses. The levels of evidence andgrades of recommendation (the grade of recommendationdescribes the whole body of evidence regarding a specificquantitative or qualitative topic [here interventions] and itsability to influence the phenomenon, “mechanical restraint,”in the clinical setting) were developed from the Danish Refer-ence Programme Secretariat (Sekretariatet for Referencepro-grammer, 2004), Method for Evaluating Research andGuideline Evidence (MERGE; Liddle et al., 1996), and SIGN(Harbour, 2008); the developed descriptions can be requiredfrom the authors.

Qualitative Research Grading. Once individual papers hadbeen assessed for methodological quality and relevance interms of the clinical questions, they were graded according to

the levels of complexity in the evidence described by Kearney(2001) (Kearney argues that: “The greater the complexity anddiscovery within qualitative findings, the stronger may be thepotential for clinical application”). The available evidencewas then compiled into tables. Appraising the qualitativeresearch papers was conducted as outlined by Sandelowskiand Barroso (2007). The new hierarchy of the recommenda-tion grades for the qualitative evidence was developed for thisreview by the first author, Jesper Bak, in order to provide aunited and structured evidence base. The developmentwas inspired by Kearney, Sandelowski, and Barroso, theDanish Reference Programme Secretariat (Sekretariatet forReferenceprogrammer, 2004), MERGE (Liddle et al., 1996),and SIGN (Harbour, 2008); the developed hierarchy can beseen in Table 4.

Synthesizing the Papers

The final synthesis was based on consensus decisions by thefour authors and was guided by the following three questions:Which interventions match patients’ views and experiences?,which interventions have yet to be evaluated in clinical set-tings?, and does the effect of the intervention seem plausiblefrom a clinical point of view?

Sample

The search yielded a total of 2,885 papers. This number wasreduced to 358 after sifting the material that potentially meteligibility criteria on the basis of the title/abstract. After thefull papers were reviewed for relevance/eligibility or for vaguerelevance/eligibility, the number of papers was furtherreduced to 268. Lastly, the papers were critically appraisedand quality checked, leading to a final number of 59 papers.Out of these, 48 were quantitative papers and 11 were qualita-tive papers.

Table 4. Grades of Recommendation for Qualitative Research

1 At least one high-quality meta-synthesis, systematic review, or meta-summary rated as 1 and directly applicable to the target population andcontext, or a body of evidence consisting principally of papers rated as 1, directly applicable to the target population and context, anddemonstrating overall consistency of findings.

2 A body of evidence including papers rated as 2 or one paper rated as 1, directly applicable to the target population and context, and demonstratingoverall consistency of findings, or extrapolated evidence from papers rated as 1.

3 A body of evidence including papers rated as 3 or one paper rated as 2, directly applicable to the target population and context, and demonstratingoverall consistency of findings, or extrapolated evidence from papers rated as 2.

4 A body of evidence including papers rated as 4 or one paper rated as 3, directly applicable to the target population and context, and demonstratingoverall consistency of findings.

5 One or two papers rated as 4.

The papers were rated according to the complexity levels by Kearney where 1 = dense explanatory description, 2 = depiction of experimental variation, 3= shared pathway or meaning, 4 = descriptive categories, 5 = findings restricted by a priori frameworks (level 5 papers were excluded).The basic assumption for the recommendation model was that the greater complexity and discovery, the more trustworthy the evidence was, and themore researchers that found independently the same finding (here intervention), the more reason there was to believe that it would work in the clinicalsetting.If the evidence was not directly applicable to the target population and context, it was downgraded.



Displaying the Findings

The development of the combined recommendation gradingmade it possible to display the registered interventions hierar-chically (see Table 5).

The Hierarchy. In Table 5, the first column shows the numberin the hierarchy. The second column shows the description ofthe synthesized intervention, and the third column shows thecombined quantitative and qualitative recommendationgrade. The recommendation grade describes the interven-tion’s ability to exert an effect on reducing the number ofmechanical restraints in the clinical setting, not the degreeof how much it would reduce the number of mechanicalrestraints, but if it would. This is described in five levels from 1to 5, with 1 being the best. 1 indicates an intervention thatmost certainly would have an effect on the phenomenon(mechanical restraint), and 5 indicates a high degree ofuncertainty whether or not the intervention would have animpact on, or have any relation at all to, the phenomenon.Combining the quantitative and qualitative recommenda-tions was done by taking the better of the two, to show whatkind of evidence had influence on the combined recommen-dation; the individual recommendations are shown in brack-ets, first, the qualitative recommendation grade and, second,the quantitative recommendation grade.

This is followed by the fourth column showing the effect ofthe intervention, where the reduction in the number ofmechanical restraints was divided into five levels from 1 to 5,with 1 being the best, from 80% to 100% reduction, 2 from60% to 80%, etc. The effect of the intervention was extrapo-lated from the quantitative papers and was combined in thefollowing way:

• No paper: The level was set to 5 (0–20% reduction)• One paper: The level was transferred directly• Two or more papers: The level was influenced by:

1. The paper with the best recommendation grade.2. The paper best describing the intervention.3. A simple average.

The Matrix. To illustrate which interventions were the mostlikely to reduce the use of mechanical restraints, every inter-vention was placed in a matrix of effect and recommendationgrade (represented as numbers in Figure 1). The closer theintervention was placed to the upper right corner, the morelikely it could reduce the occurrence of mechanical restraints(the “better” the intervention was).

Approval and Ethics

Literature reviews do not need approval from the scientific–ethical committee system (Indenrigs- og Sundhedsminis-

teriet, 2004). The study followed the Ethical Guidelines forNursing Research in the Nordic Countries (Northern NurseFederation, 2003) and the White Paper Regarding a Draft Rec-ommendation on Legal Protection of Persons Suffering FromMental Disorder Especially Those Placed as InvoluntaryPatients (Council of Europe, 2000).

Findings

No interventions reached the highest degree of recommenda-tion combined with the highest effect. One interventionreached the highest effect, with a decrease in mechanicalrestraints of 87%, but only with a recommendation grade of 3(Table 5, no. 1), and another with a recommendation grade of2, but only with decrease in mechanical restraints of 27%(Table 5, no. 5).

The Three Highest Ranked Interventions

The highest ranked intervention found was implementationof cognitive milieu therapy (CMT) (Lykke, Austin, &Morch, 2008), where the reduction in mechanical restraintswas 87%. CMT was an active, structured, problem-orientated, psycho-educational, and dynamic treatmentform. The cognitive treatment model in the environmentgave the patient opportunity to find new alternatives toinappropriate patterns of reactions and develop new skills(Oestrich, Lykke, & Holm, 2001). One of the core issues inthe CMT was a schema-focused dialog between the patientand the nursing staff, in the situation just before a potentialmechanical restraint. Here, the patient had the opportunityto try out new actions. This sort of patient involvement isalso found elsewhere (Robins, Sauvageot, Cusack, Suffoletta-Maierle, & Frueh, 2005; Thyrsting & Hall, 2008) to havepositive influence on the occurrence of mechanical restraint.Especially one of these papers (Robins et al., 2005) had ahigh evidence quality, and for that reason, the combinationof CMT and patient involvement is regarded as the “best”intervention in this review (Table 5, no. 1).

The second highest ranked intervention is not one particu-lar intervention but programs including several interven-tions, for instance, patient education, staff education, patientparticipation, programmatic changes, high-level administra-tive endorsement, cultural changes, data analyses, etc. Thereare several papers describing different programs (Fisher,2003; Hellerstein et al., 2007; Khadivi, Patel, Atkinson, &Levine, 2004; Sullivan et al., 2005; Taxis, 2002). An average ofthese gives a 76% decrease of mechanical restraints (Table 5,no. 2). The evidence consisted mostly of “before- and after”papers, and because some of these are well founded, the rec-ommendation grade is the same as in no. 1.

The third highest ranked intervention is implementationof patient-centered care (Fletcher & Stevenson, 2001; Robins



Table 5. Intervention Hierarchy

No. Intervention Recommendation Effect Literature Reference

1 Implementation of cognitive milieu therapy reduces the number

of mechanical restraint episodes, among other things, through

patient involvement and empowerment.

3 (3–4) 1 Lykke et al., 2008; Robins et al., 2005; Thyrsting and Hall, 2008

2 Combined intervention programs are effective in avoiding

mechanical restraints, for example, patient participation,

patient education, staff education, programmatic changes,

high-level administrative endorsement, cultural changes, data

analysis, etc.

3 (n.a.–3) 2 Fisher, 2003; Hellerstein et al., 2007; Khadivi et al., 2004; Sullivan

et al., 2005; Taxis, 2002

3 Implementation of patient-centered care with a higher degree of

patients’ positive involvement in their own care reduces the

number of mechanical restraint episodes.

3 (3–4) 2 Fletcher and Stevenson, 2001; Robins et al., 2005; Sullivan et al.,

2004; Thyrsting and Hall, 2008

4 The identification of the patients’ crisis triggers unique calming

techniques and decisions on specific crisis interventions early in

the admission procedure, thereby reducing the number of


3 (4–3) 3 Jonikas, Cook, Rosen, Laris, and Kim, 2004; McCain and

Kornegay, 2005; Meehan, Mcintosh, and Bergen, 2006; Wynn

and Mykleburst, 2006

5 Introduction of validated risk assessment systems including

suggestions of care interventions reduces the number of


2 (n.a.–2) 4 Abderhalden et al., 2008; Cochrane-Brink, Lofchy, and

Sakinofsky, 2000; Dernevik, Grann, and Johansson, 2002

6 Effective treatment of substance abuse and psychosis reduce the

number of mechanical restraint episodes; the treatment of

mental illnesses is also described elsewhere in the review.

4 (n.a.–4) 3 Fletcher and Stevenson, 2001; Palmstierna and Olsson, 2007;

Sullivan et al., 2004; Thyrsting and Hall, 2008

7 Mandatory review from qualified clinical experts, for example, as

post-incident review, minimizes the number of mechanical

restraint episodes.

4 (n.a.–4) 3 Donat, 1998; Prescott, Madden, Dennis, Tisher, and Wingate,

2007

8 Better staff training reduces the number of mechanical restraints

(longer duration of the training, more staff participating, use of

different and involving education methods, higher quality of

the content, and greater diversity of the topics, for example,

legislation, crisis management, physical intervention strategies,

antecedents, warning signs, de-escalation techniques, etc.).

3 (3–4) 4 Bonner, Lowe, Rawcliffe, and Wellman, 2002; Bowers, Nijman,

et al., 2006; Cardell and Pitula, 1999; Delaney and Johnson,

2006; Hahn, Abderhalden, Duxbury, Halfens, and Needham,

2006; Jonikas et al., 2004; Lee et al., 2001; Mackay, Paterson,

and Cassells, 2005; Needham, Dassen, Halfens, Haug, and

Fisher, 2005; Spokes et al., 2002; Wright et al., 2005

9 Staff attitudes such as positive care for the patients, respect,

collaboration, communication, humanism, empathy, and less

control and involuntary procedures reduce the number of


4 (4–4) 4 Allen, Carpenter, Sheets, Miccio, and Ross, 2003; Bowers,

Alexander, Simpson, Ryan, and Carr-Walker, 2007; Bowers,

Brennan, Flood, Lipang, and Oladapo, 2006; Duxbury, 2002;

Gillig, Markert, Barron, and Coleman, 1998; Meehan et al.,

2006; Robins et al., 2005; Spokes et al., 2002

10 Higher patient–staff ratio reduces the number of mechanical

restraints.

4 (n.a.–4) 4 Smith et al., 2005; Wynn, 2003

11 The higher educated the staff, the fewer episodes of mechanical

restraints.

3 (3–4) 5 Bonner et al., 2002; Cardell and Pitula, 1999; Delaney and

Johnson, 2006; Flannery, Stone, Rego, Walker, and Farley,

2006; Gim et al., 1999; Mackay et al., 2005; Spokes et al.,

2002

12 The more experienced the staff, the fewer episodes of


3 (3–4) 5 Bonner et al., 2002; Bowers, Jeffery, et al., 2007; Cardell and

Pitula, 1999; Delaney and Johnson, 2006; Flannery et al.,

2006; Janssen, Noorthoorn, Linge, and Lendemeijer, 2007;

Mackay et al., 2005; Spokes et al., 2002

13 If nurses change their focus from considering the patient as

deviant to being a resource in his/her own treatment, it would

decrease the number of mechanical restraint.

3 (3–n.a.) 5 Vatne and Fagermoen, 2007

14 Early and adequate administration of oral p.n. medication guided

by the individual patients’ problems, medication experiences,

and preferences leads to fewer mechanical restraint episodes.

4 (4–4) 5 Allen et al., 2003; Goedhard, Stolker, Nijman, Egberts, and

Heerdink, 2007; McCain and Kornegay, 2005; Meehan et al.,

2006; Thapa et al., 2003

15 Debriefing, defusing, and crisis intervention minimize the number


4 (5–4) 5 Bonner et al., 2002; Bowers, Simpson, et al., 2006; Flannery

et al., 2006; Jonikas et al., 2004

16 Patients’ response to rules can be positive and lead to fewer cases

of mechanical restraint, for instance, if staff helps patients to

understand not only the rules, but also the rationale for the

presence of such rules. If the staff manages to make patients

understand that the rules are imposed out of concern for their

welfare, then patients may be less reluctant to following these

rules.

4 (5–4) 5 Alexander, 2006; Duxbury, 2002; Gillig et al., 1998; Meehan

et al., 2006; Robins et al., 2005

17 Activity-based nursing interventions, for example, physical

activities, are effective in avoiding mechanical restraints.

4 (5–4) 5 Meehan et al., 2006; Thomas, Jones, Johns, and Trauer, 2006



et al., 2005; Sullivan, Wallis, & Lloyd, 2004; Thyrsting & Hall,2008) (Table 4, no. 3). Here, implementation of the TidalModel (TM) reduces the number of mechanical restraints by68%. TM is a philosophical approach to the discovery ofmental health. TM focuses on helping people who have expe-rienced some metaphorical“breakdown”to recover their livesas fully as possible, by reclaiming the personal story of theirdistress and difficulty (Barker & Buchanan-Barker, 2010).One of the core elements in TM is empowerment of thepatient. Empowerment is only possible if the patients areinvolved in their own care. Patient involvement is also foundelsewhere (Robins et al., 2005; Thyrsting & Hall, 2008) to havea positive influence on the occurrence of mechanicalrestraint. Therefore, the recommendation grade is the same asin the first intervention described.

Discussion

The highest ranked intervention was CMT combined withpatient involvement. To our knowledge, this result has not

Table 5. Continued


18 Several mental and physical working environmental factors

influence positively the number of mechanical restraints, for

example, good management as a high degree of order,

organization, or support; adequate staffing; psychiatric

training; rotation; education; no crowding; emergency backup,

etc. Several of these factors are described elsewhere in the

review.

4 (5–4) 5 Bonner et al., 2002; Bowers, Allan, Simpson, Jones, and

Whittington, 2008; Daffern, Mayer, and Martin, 2006; Owen,

Tarantello, Jones, and Tennant, 1998; Whittington and

McLaughlin, 2000

19 Less crowding, for example, less stimuli and less environmental

stress, and more personal space decrease the number of


4 (5–4) 5 Meehan et al., 2006; Vaaler, Morken, Fløvig, Iversen, and Linaker,

2006

20 Higher regular staff and lower substitute ratio reduce the number


4 (5–4) 5 Bonner et al., 2002; Bowers, Allan, Simpson, Nijman, and

Warren, 2007; Spokes et al., 2002

21 Staffs’ opposition toward mechanical restraint because of the

perception that mechanical restraint is ethically problematic

decreases the number of mechanical restraints.

4 (n.a.–4) 5 Lind, Kaltiala-Heino, Suominen, Leino-Kilpi, and Valimaki, 2004;

Van Doeselaar, Sleegers, and Hutschemaekers, 2008

22 Alarm systems, for example, video surveillance and radio systems,

and well-trained response teams reduce the number of


4 (n.a.–4) 5 D’Orio, Purselle, Stevens, and Garlow, 2004

23 Better care alliance between patients and the nursing staff

reduces the number of mechanical restraints.

4 (n.a.–4) 5 Wynn, 2003

24 Communication especially communicative de-escalation reduces

the number of mechanical restraint episodes.

4 (n.a.–4) 5 Duxbury and Whittington, 2005

25 Well-maintained and familiar surroundings reduce the number of


4 (n.a.–4) 5 Vaaler, Morken, and Linaker, 2005

26 Music as diversion and relaxation before and after mechanical

restraint reduces the number of mechanical restraint episodes.

4 (n.a.–4) 5 Janelli and Kanski, 1998

27 Separation of acutely disturbed patients reduces the number of


5 (5–n.a.) 5 Meehan et al., 2006

No. refers to the number in the hierarchy where 1 is highest/best. Recommendation describes the whole body of evidence regarding a specific intervention level of scientific

“truth” or its potential for clinical application. To show what kind of evidence that mostly influenced the combined recommendation, the individual recommendations are

shown in brackets, first, the qualitative recommendation grade and, second, the quantitative recommendation grade. All from 1 to 5, where 1 is the highest/best. n.a. referred

to no available paper. Effect refers to the reduction in the number of mechanical restraints, from 1 to 5, with 1 being the best, from 80% to 100% reduction, 2 from 60% to

80%, etc.

p.n., pro necessitate (as needed).

Figure 1. Effect and Recommendation Grade MatrixThe numbers in the matrix refer to a specific intervention (the firstnumber in the intervention hierarchy [Table 5]).



been identified in prior reviews, one probable cause beingthat the CMT paper was not published in English, and in theabstract, the effect was only described as a significant effect,not the exact figure of a staggering 87% reduction. Also, thefact that this paper is rather new (from 2008) might explain itslimited impact.

The second highest ranked intervention was combinedinterventions (many interventions in a program or apackage). This finding is supported by an earlier review(Gaskin, Elsom, & Happell, 2007) where the authors’ conclu-sion recommended a combination of several interventions toreduce the number of mechanical restraints.

The third highest ranked intervention was implementationof patient-centered care, which means that the patient isplaced in the center of his/her own care. Despite being inclosed units where restrictions to individual liberty oftenmust be exerted, the patient should be given as much influ-ence over his/her situation as possible. The positive influenceon mechanical restraint by patient-centered or patient-involving care is also supported in other reviews (Gaskinet al., 2007; Sailas & Wahlbeck, 2005).

No interventions reached the highest degree of recommen-dation combined with the highest effect, and to create strongevidence-based practice in this area, substantially furtherresearch is needed. This is necessary because implementingnot effective interventions could be both expensive andunethical.

Many apparently different interventions centered on thepatient could be more efficacious on reducing the number ofmechanical restraints. This point could lead us to speculatethat therapeutic interventions could allow us to reduce coer-cive measures in psychiatry.

Limitations

First, only one person was involved in the initial reviewprocess, until the drawing up of evidence statements; it mightbe argued that two or more persons reading all the materialfollowed by consensus evaluations would produce more reli-able results. However, all the evidence was rated in exactly thesame way (no inter-rater differences).

Second, the quantification combining qualitative andquantitative evidence is new and possibly controversial. Asimple argument in favor is that the more mechanicalrestraint and a specific intervention are interrelated (nomatter if this relation in nature is quantitative or qualitative),the greater the evidence for this relation becomes. So, in theauthors’ opinion, as mechanical restraint is a multifacetedphenomenon and the evidence is limited, it is reasonable tocombine well-established quantitative review procedureswith well-established quantified qualitative review proce-dures in order to deduce maximum information from theavailable papers.

Implication for Research

There is an obvious lack of high-quality intervention studies,which may be because of the vast difficulties in performingsuch studies because of the complexity of the phenomenonand because of the fact that the interventions must be per-formed in clinical settings. The findings of this review point tothe need for randomized clinical trials on specific interven-tions. The likelihood of implementing interventions inseveral units as a cluster-randomized trial would constitute asuitable study design for this purpose. Before choosing theintervention for the clinical trial, it could be interesting toinvestigate the “effect and recommendation grade matrix”through a cross-sectional study evaluating the potency of theindividual interventions in relation to mechanical restraintsin terms of hypotheses testing and falsification. This wouldstrengthen the choice of intervention.

Implications for Nursing Practice

The papers included in this review do not imply that the inter-ventions easily can be transferred to practice because in manycases, they describe specific settings and narrow populations.Neither did any of the interventions reach the highest degreeof recommendation, as far as influencing the number ofmechanical restraints in the clinical setting is concerned.However, no intervention seemed to have any negative orharmful side effects. Considering that mechanical restraint isone of the most intrusive care interventions, every effort toavoid it must be used until convincing evidence is provided.

Acknowledgments

For funding this project, we thank: the Mental Health CentreSct. Hans, Copenhagen University Hospital, the mentalhealth services in the capital region of Denmark, and theResearch Foundation in the capital region of Denmark.

The authors wish to thank Margit Asser (Deputy Head ofCentre, Mental Health Centre Sct. Hans, Roskilde, Denmark),Jeanett Bauer (Employed Senior Physician, Psychiatric CentreCopenhagen, Bispebjerg Hospital, Denmark), and ThomasWerge (Head of Research, Mental Health Centre Sct. Hans,Roskilde, Denmark) for valuable comments and carefulreading during the writing process.

References

Abderhalden, C., Dassen, T., Fisher, J. E., Halfens, R. J. G., Haug,H. J., & Needham, I. (2008). Structured risk assessment andviolence in acute psychiatric wards: Randomised controlledtrial. British Journal of Psychiatry, 193(1), 44.doi:10.1192/bjp.bp.107.045534

Alexander, J. (2006). Patients’ feelings about ward nursingregimes and involvement in rule construction. Journal of



Psychiatric and Mental Health Nursing, 13(5), 543–553.doi:10.1111/j.1365-2850.2006.00977.x

Allen, M. H., Carpenter, D., Sheets, J. L., Miccio, S., & Ross, R.(2003). What do consumers say they want and need during apsychiatric emergency? Journal of Psychiatric Practice, 9(1),39–58. doi:10.1097/00131746-200301000-00005

Barker, P., & Buchanan-Barker, P. (2010). The tidal model ofmental health recovery and reclamation: Application in acutecare settings. Issues in Mental Health Nursing, 31(3), 171–180.doi:10.3109/01612840903276696

Bonner, G., Lowe, T., Rawcliffe, D., & Wellman, N. (2002). Traumafor all: A pilot study of the subjective experience of physicalrestraint for mental health inpatients and staff in the UK.Journal of Psychiatric and Mental Health Nursing, 9(4),465–473. doi:10.1046/j.1365-2850.2002.00504.x

Bower, F. L., McCullough, C. S., & Timmons, M. E. (2003). Asynthesis of what we know about the use of physical restraintsand seclusion with patients in psychiatric and acute caresettings: 2003 update. Worldviews on Evidence-Based NursingPresents the Archives. Online Journal of Knowledge Synthesis forNursing, 10(1), 1–29. doi:10.1111/j.1524-475X.2003.00001.x

Bowers, L., Alexander, J., Simpson, A., Ryan, C., & Carr-Walker, P.(2007). Student psychiatric nurses’ approval of containmentmeasures: Relationship to perception of aggression and attitudesto personality disorder. International Journal of Nursing Studies,44(3), 349–356. doi:10.1016/j.ijnurstu.2005.03.002

Bowers, L., Allan, T., Simpson, A., Jones, J., & Whittington, R.(2008). Morale is high in acute inpatient psychiatry. SocialPsychiatry and Psychiatric Epidemiology, 44(1), 39–46.doi:10.1007/s00127-008-0396-z

Bowers, L., Allan, T., Simpson, A., Nijman, H., & Warren, J.(2007). Adverse incidents, patient flow and nursing workforcevariables on acute psychiatric wards: The Tompkins AcuteWard Study. International Journal of Social Psychiatry, 53(1), 75.

Bowers, L., Brennan, G., Flood, C., Lipang, M., & Oladapo, P.(2006). Preliminary outcomes of a trial to reduce conflict andcontainment on acute psychiatric wards: City nurses. Journal ofPsychiatric and Mental Health Nursing, 13(2), 165–172.doi:10.1111/j.1365-2850.2006.00931.x

Bowers, L., Jeffery, D., Simpson, A., Daly, C., Warren, J., & Nijman,H. (2007). Junior staffing changes and the temporal ecology ofadverse incidents in acute psychiatric wards. Journal ofAdvanced Nursing, 57(2), 153–160. doi:10.1111/j.1365-2648.2006.04101.x

Bowers, L., Nijman, H., Allan, T., Simpson, A., Warren, J., &Turner, L. (2006). Prevention and management of aggressiontraining and violent incidents on UK acute psychiatric wards.Psychiatric Services, 57(7), 1022. doi:10.1176/appi.ps.57.7.1022

Bowers, L., Simpson, A., Eyres, S., Nijman, H., Hall, C., Grange, A.,& Phillips, L. (2006). Serious untoward incidents and theiraftermath in acute inpatient psychiatry: The Tompkins AcuteWard Study. International Journal of Mental Health Nursing,15(4), 226–234. doi:10.1111/j.1447-0349.2006.00428.x

Bowers, L., van der Werf, B., Vokkolainen, A., Muir-Cochrane, E.,Allan, T., & Alexander, J. (2007). International variation in

containment measures for disturbed psychiatric inpatients: Acomparative questionnaire survey. International Journal ofNursing Studies, 44(3), 357–364.

Bremmes, R., Hatling, T., & Bjørngaard, J. H. (2008). Bruk avtvangsmidler i psykisk helsevern i 2001, 2003, 2005 og 2007.Trondheim, Norway: SINTEF Helse.

Cardell, R., & Pitula, C. R. (1999). Suicidal inpatients’ perceptionsof therapeutic and nontherapeutic aspects of constantobservation. Psychiatric Services, 50(8), 1066.

Chan, T. C., Neuman, T., Clausen, J., Eisele, J., & Vilke, G. M.(2004). Weight force during prone restraint and respiratoryfunction. American Journal of Forensic Medicine and Pathology,25(3), 185.

Cochrane-Brink, K. A., Lofchy, J. S., & Sakinofsky, I. (2000).Clinical rating scales in suicide risk assessment. GeneralHospital Psychiatry, 22(6), 445–451. doi:10.1016/S0163-8343(00)00106-7

Council of Europe. (2000). White paper regarding a draftrecommendation on legal protection of persons suffering frommental disorder especially those placed as involuntary patients.Strasbourg, France.

Critical Appraisal Skills Programme. (2006). Appraisal tools.Retrieved from http://www.sph.nhs.uk/sph-files/casp-appraisal-tools/?searchterm=casp

Cusack, K. J., Frueh, B. C., Hiers, T., Suffoletta-Maierle, S., &Bennett, S. (2003). Trauma within the psychiatric setting: Apreliminary empirical report. Administration and Policy inMental Health, 30(5), 453–460. doi:10.1023/A:1024697715409

D’Orio, B. M., Purselle, D., Stevens, D., & Garlow, S. J. (2004).Reduction of episodes of seclusion and restraint in a psychiatricemergency service. Psychiatric Services, 55(5), 581–583.

Daffern, M., Mayer, M., & Martin, T. (2006). Staff gender ratioand aggression in a forensic psychiatric hospital. InternationalJournal of Mental Health Nursing, 15(2), 93–99. doi:10.1111/j.1447-0349.2006.00408.x

Delaney, K. R., & Johnson, M. E. (2006). Keeping the unit safe:Mapping psychiatric nursing skills. Journal of the AmericanPsychiatric Nurses Association, 12(4), 198–207. doi:10.1177/1078390306294462

Department of Health, National Risk Management Programme.(2007). Best practice in managing risk. Retrieved fromhttp://webarchive.nationalarchives.gov.uk/+/www.dh.gov.uk/prod_consum_dh/groups/dh_digitalassets/@dh/@en/documents/digitalasset/dh_076512.pdf

Dernevik, M., Grann, M., & Johansson, S. (2002). Violentbehaviour in forensic psychiatric patients: Risk assessment anddifferent risk-management levels using the HCR-20. Psychology,Crime & Law, 8(1), 93–111. doi:10.1080/10683160208401811

Donat, D. C. (1998). Impact of a mandatory behavioralconsultation on seclusion/restraint utilization in a psychiatrichospital. Journal of Behavior Therapy and ExperimentalPsychiatry, 29(1), 13–19. doi:10.1016/S0005-7916(97)00039-6

Duxbury, J. (2002). An evaluation of staff and patient views ofand strategies employed to manage inpatient aggression andviolence on one mental health unit: A pluralistic design. Journal



http://www.sph.nhs.uk/sph-files/casp-appraisal-tools/?searchterm=casp

http://www.sph.nhs.uk/sph-files/casp-appraisal-tools/?searchterm=casp

http://webarchive.nationalarchives.gov.uk/+/www.dh.gov.uk/prod_consum_dh/groups/dh_digitalassets/@dh/@en/documents/digitalasset/dh_076512.pdf



of Psychiatric and Mental Health Nursing, 9(3), 325–337.doi:10.1046/j.1365-2850.2002.00497.x

Duxbury, J., & Whittington, R. (2005). Causes and managementof patient aggression and violence: Staff and patientperspectives. Journal of Advanced Nursing, 50(5), 469–478.doi:10.1111/j.1365-2648.2005.03426.x

The European Committee for the Prevention of Torture andInhumane or Degrading Treatment or Punishment. (2002).Report to the government of Denmark on the visit to Denmarkcarried out by the European Committee for the Prevention ofTorture and Inhumane or Degrading Treatment or Punishment.Strasbourg, France.

Fisher, W. A. (2003). Elements of successful restraint andseclusion reduction programs and their application in a large,urban, state psychiatric hospital. Journal of Psychiatric Practice,9(1), 7–15.

Flannery, R. B., Jr., Stone, P., Rego, S., Walker, A. P., & Farley, E.(2006). Characteristics of staff victims of psychiatric patientassaults: 15-year analysis of the Assaulted Staff Action Program(ASAP). Psychiatric Quarterly, 78(1), 25–37. doi:10.1007/s11126-006-9024-9

Fletcher, E., & Stevenson, C. (2001). Launching the tidal model inan adult mental health programme. Nursing Standard, 15(49),33–36.

Frueh, B. C., Knapp, R. G., Cusack, K. J., Grubaugh, A. L.,Sauvageot, J. A., Cousins, V. C., . . . Hiers, T. G. (2005). Specialsection on seclusion and restraint: Patients’ reports oftraumatic or harmful experiences within the psychiatricsetting. Psychiatric Services, 56(9), 1123. doi:10.1176/appi.ps.56.9.1123

Gaskin, C. J., Elsom, S. J., & Happell, B. (2007). Interventions forreducing the use of seclusion in psychiatric facilities: Review ofthe literature. British Journal of Psychiatry: The Journal ofMental Science, 191, 298–303. doi:10.1192/bjp.bp.106.034538

Gillig, P. M., Markert, R., Barron, J., & Coleman, F. (1998). Acomparison of staff and patient perceptions of the causes andcures of physical aggression on a psychiatric unit. PsychiatricQuarterly, 69(1), 45–60.

Gim, Y. S., Cheng, S. O. B., Hong, T. L., Chin, M. W. B., Hen, W. H.,& Lee, Y. J. (1999). Nurses experiences in identifying andmanaging violence precipitants in a psychiatric in-patientsetting. Singapore Nursing Journal, 26(1), 21–24.

Goedhard, L. E., Stolker, J. J., Nijman, H. L. I., Egberts, A. C. G., &Heerdink, E. R. (2007). Aggression of psychiatric patientsassociated with the use of as-needed medication.Pharmacopsychiatry, 40(1), 25. doi:10.1055/s-2007-961817

Hahn, S., Abderhalden, C., Duxbury, J., Halfens, R. J. G., &Needham, I. (2006). The effect of a training course on mentalhealth nurses’ attitudes on the reasons of patient aggressionand its management. Journal of Psychiatric and Mental HealthNursing, 13(2), 197–204. doi:10.1111/j.1365-2850.2006.00941.x

Harbour, R. T. (2008). SIGN 50—A guideline developer’shandbook. Edinburgh, UK: Scottish Intercollegiate GuidelinesNetwork.

Hauksson, P. (2008). Statement made by Mr Pétur Hauksson at themeeting with senior Danish officials on 20 February 2008, at theend of the 4th periodic visit to Denmark by the EuropeanCommittee for the Prevention of Torture and Inhuman orDegrading Treatment or Punishment. The European Committeefor the Prevention of Torture and Inhuman or DegradingTreatment or Punishment. Strasbourg, France.

Health Care Practice Research and Development Unit. (2005).HCPRDU tools—Critical appraisal tools. The Salford Centre forNursing, Midwifery and Collaborative Research. Manchester,UK.

Hellerstein, D. J., Staub, A. B., & Lequesne, E. (2007). Decreasingthe use of restraint and seclusion among psychiatric inpatients.Journal of Psychiatric Practice, 13(5), 308–317.doi:10.1097/01.pra.0000290669.10107.ba

Hem, E., Steen, O., Opjordsmoen, S., & Moussaoui, D. (2001).Thrombosis associated with physical restraints. ActaPsychiatrica Scandinavica, 103(1), 73–76. doi:10.1111/j.1600-0447.2001.00082.x

Hick, J. L., Smith, S. W., & Lynch, M. T. (1999). Metabolic acidosisin restraint-associated cardiac arrest: A case series. AcademicEmergency Medicine, 6(3), 239–243.doi:10.1111/j.1553-2712.1999.tb00164.x

Indenrigs- og Sundhedsministeriet. (2004). Bekendtgørelse ominformation og samtykke ved inddragelse af forsøgspersoner ibiomedicinske forskningsprojekter. LBK nr. 806.

Indenrigs- og Sundhedsministeriet. (2006). Bekendtgørelse af lovom anvendelse af tvang i psykiatrien. LBK nr. 1111.

Institut for Kvalitet og Akkreditering i Sundhedsvæsenet. (2009).Den Danske Kvalitetsmodel. Akkrediteringsstandarder forsygehuse (1. version).

Janelli, L. M., & Kanski, G. (1998). Music for untying restrainedpatients. Journal of the New York State Nurses’ Association,29(1), 13–15.

Janssen, W., Noorthoorn, E., Linge, R., & Lendemeijer, B. (2007).The influence of staffing levels on the use of seclusion.International Journal of Law and Psychiatry, 30(2),118–126. doi:10.1016/j.ijlp.2006.04.005

Jonikas, J. A., Cook, J. A., Rosen, C., Laris, A., & Kim, J. B. (2004).A program to reduce use of physical restraint in psychiatricinpatient facilities. Psychiatric Services, 55(7), 818–820.

Kearney, M. H. (2001). Levels and applications of qualitativeresearch evidence. Research in Nursing and Health, 24(2),145–153. doi:10.1002/nur.1017

Khadivi, A. N., Patel, R. C., Atkinson, A. R., & Levine, J. M. (2004).Association between seclusion and restraint and patient-relatedviolence. Psychiatric Services, 55(11), 1311–1312.

Laursen, S. B., Jensen, T. N., Bolwig, T., & Olsen, N. V. (2005).Deep venous thrombosis and pulmonary embolism followingphysical restraint. Acta Psychiatrica Scandinavica, 111(4),324–327. doi:10.1111/j.1600-0447.2004.00456.x

Lee, S., Wright, S., Sayer, J., Parr, A., Gray, R., & Gournay, K.(2001). Physical restraint training for nurses in English andWelsh psychiatric intensive care and regional secure units.Journal of Mental Health, 10(2), 151–162.



Leitner, M., Barr, W., Jones, S., McGuire, J., & Whittington, R.(2006). Systematic review of prevention and interventionstrategies for populations at high risk of engaging in violentbehaviour. Liverpool, UK: The National Forensic Mental HealthR&D Programme.

Liddle, J., Williamsom, M., & Irwing, L. (1996). Method forevaluating research and guideline evidence. Sydney, Australia:NSW Health Department.

Lind, M., Kaltiala-Heino, R., Suominen, T., Leino-Kilpi, H., &Valimaki, M. (2004). Nurses’ ethical perceptions aboutcoercion. Journal of Psychiatric and Mental Health Nursing,11(4), 379–385. doi:10.1111/j.1365-2850.2004.00715.x

Lykke, J., Austin, S. F., & Morch, M. M. (2008). Cognitive milieutherapy and restraint within dual diagnosis populations.Ugeskrift for Laeger, 170(5), 339–343.

Mackay, I., Paterson, B., & Cassells, C. (2005). Constant or specialobservations of inpatients presenting a risk of aggression orviolence: Nurses’ perceptions of the rules of engagement.Journal of Psychiatric and Mental Health Nursing, 12(4),464–471. doi:10.1111/j.1365-2850.2005.00867.x

McCain, M., & Kornegay, K. (2005). Behavioral health restraint:The experience and beliefs of seasoned psychiatric nurses.Journal for Nurses in Staff Development, 21(5), 236.

Meehan, T., Mcintosh, W., & Bergen, H. (2006). Aggressivebehaviour in the high-secure forensic setting: The perceptionsof patients. Journal of Psychiatric and Mental Health Nursing,13(1), 19–25. doi:10.1111/j.1365-2850.2006.00906.x

Morrison, A., & Sadler, D. (2001). Death of a psychiatric patientduring physical restraint. Excited delirium—A case report.Medicine, Science, and the Law, 41(1), 46–50.

Muralidharan, S., & Fenton, M. (2006). Containmentstrategies for people with serious mental illness. CochraneDatabase of Systematic Reviews, 3, doi:10.1002/14651858.CD002084

National Institute for Health and Clinical Excellence (NICE).(2004). Self-harm, the short-term physical and psychologicalmanagement and secondary prevention of self-harm in primaryand secondary care (clinical guideline 16). London: The RoyalCollege of Psychiatrists.

National Institute for Health and Clinical Excellence (NICE).(2005). Violence, the short-term management ofdisturbed/violent behaviour in in-patient psychiatric settings andemergency departments (clinical guideline 25). London: NationalInstitute for Clinical Excellence.

National Institute for Health and Clinical Excellence (NICE).(2007). The guidelines manual. London: Author. Retrieved fromwww.nice.org.uk

Needham, I., Dassen, T., Halfens, R. J. G., Haug, H. J., & Fisher,J. E. (2005). The effect of a training course in aggressionmanagement on mental health nurses’ perceptions ofaggression: A cluster randomised controlled trial. InternationalJournal of Nursing Studies, 42(6), 649–655.

Nielsen, A. S. (2005). Deep venous thrombosis and fatalpulmonary embolism in a physically restrained patient.Ugeskrift for Laeger, 167(21), 2294.

Nijman, H. L. I., Campo, J. M. L. G., Ravelli, D. P., &Merckelbach, H. L. G. J. (1999). A tentative model ofaggression on inpatient psychiatric wards. Psychiatric Services,50(6), 832–834.

Northern Nurse Federation. (2003). Ethical guidelines for nursingresearch in the Nordic countries. Vård i Norden, 23(4), 1–19.

Oestrich, I., Lykke, J., & Holm, L. (2001). En kognitiv model formiljøterapi. In I. Oestrich, & L. Holm (Eds.), Kognitivmiljøterapi (1st ed., pp. 19–79). København, Denmark: Danskpsykologisk Forlag.

Owen, C., Tarantello, C., Jones, M., & Tennant, C. (1998).Violence and aggression in psychiatric units. PsychiatricServices, 49(11), 1452.

Palmstierna, T., & Olsson, D. (2007). Violence from young womeninvoluntarily admitted for severe drug abuse. Acta PsychiatricaScandinavica, 115(1), 66–72. doi:10.1111/j.1600-0447.2006.00853.x

Prescott, D. L., Madden, L. M., Dennis, M., Tisher, P., & Wingate,C. (2007). Reducing mechanical restraints in acute psychiatriccare settings using rapid response teams. Journal of BehavioralHealth Services and Research, 34(1), 96–105.

Robins, C. S., Sauvageot, J. A., Cusack, K. J., Suffoletta-Maierle, S.,& Frueh, B. C. (2005). Special section on seclusion andrestraint: Consumers’ perceptions of negative experiences and“sanctuary harm” in psychiatric settings. Psychiatric Services,56(9), 1134–1138. doi:10.1176/appi.ps.56.9.1134

Sailas, E., & Wahlbeck, K. (2005). Restraint and seclusion inpsychiatric inpatient wards. Current Opinion in Psychiatry,18(5), 555–559.

Sandelowski, M., & Barroso, J. (2007). Handbook for synthesizingqualitative research. Berlin, Germany: Springer.

Sekretariatet for Referenceprogrammer. (2004). Vejledning iudarbejdelse af referenceprogrammer. København, Denmark:Sundhedsstyrelsen.

Smith, G. M., Davis, R. H., Bixler, E. O., Lin, H.-M., Altenor, A.,Altenor, R. J., . . . Kopchick, G. A. (2005). Pennsylvania statehospital system’s seclusion and restraint reduction program.Psychiatric Services, 56(9), 1115–1122. doi:10.1176/appi.ps.56.9.1115

Spokes, K., Bond, K., Lowe, T., Jones, J., Illingworth, P.,Brimblecombe, N. et al. (2002). HOVIS—TheHertfordshire/Oxfordshire Violent Incident Study. Journal ofPsychiatric and Mental Health Nursing, 9(2), 199–209.doi:10.1046/j.1365-2850.2002.00467.x

Steinert, T., Bergbauer, G., Psych, D., Schmid, P., & Gebhardt, R. P.(2007). Seclusion and restraint in patients with schizophrenia:Clinical and biographical correlates. Journal of Nervous andMental Disease, 195(6), 492. doi:10.1097/NMD.0b013e3180302af6

Sullivan, A. M., Bezmen, J., Barron, C. T., Rivera, J., Curley-Casey,L., & Marino, D. (2005). Reducing restraints: Alternatives torestraints on an inpatient psychiatric service—Utilizing safeand effective methods to evaluate and treat the violent patient.Psychiatric Quarterly, 76(1), 51–65. doi:10.1007/s11089-005-5581-3



Sullivan, D., Wallis, M., & Lloyd, C. (2004). Effects ofpatient-focused care on seclusion in a psychiatric intensive careunit. International Journal of Therapy and Rehabilitation, 11,503–508.

Sundhedsstyrelsen. (2008). Nye tal fra sundhedsstyrelsen. Anven-delse af tvang i psykiatrien 2007. København, Denmark: Author.

Sundhedsstyrelsen. (2009). Tvangsforanstaltninger ipsykiatrien—En udredning. København, Denmark: Author.

Taxis, J. C. (2002). Ethics and praxis: Alternative strategies tophysical restraint and seclusion in a psychiatric setting. Issues inMental Health Nursing, 23(2), 157–170.

Thapa, P. B., Palmer, S. L., Owen, R. R., Huntley, A. L., Clardy, J.A., & Miller, L. H. (2003). P.R.N. (as-needed) orders andexposure of psychiatric inpatients to unnecessary psychotropicmedications. Psychiatric Services, 54(9), 1282–1286.

Thomas, B., Jones, M., Johns, P., & Trauer, T. (2006). PRNmedication use in a psychiatric high-dependency unitfollowing the introduction of a nurse-led activity programme.International Journal of Mental Health Nursing, 15(4),266. doi:10.1111/j.1447-0349.2006.00433.x

Thyrsting, K., & Hall, E. O. C. (2008). Prevention of coercion inpsychiatric units: A secondary analysis of a practitionerresearch project. Klinisk Sygepleje, 22(2), 78–86.

Vaaler, A. E., Morken, G., Fløvig, J. C., Iversen, V. C., & Linaker,O. M. (2006). Effects of a psychiatric intensive care unit in anacute psychiatric department. Nordic Journal of Psychiatry,60(2), 144–149. doi:10.1080/08039480600583472

Vaaler, A. E., Morken, G., & Linaker, O. M. (2005). Effects of differ-ent interior decorations in the seclusion area of a psychiatric

acute ward. Nordic Journal of Psychiatry, 59(1), 19–24.doi:10.1080/08039480510018887

Van Doeselaar, M., Sleegers, P., & Hutschemaekers, G. (2008).Professionals’ attitudes toward reducing restraint: The case ofseclusion in the Netherlands. Psychiatric Quarterly, 79(2),97–109. doi:10.1007/s11126-007-9063-x

Vatne, S., & Fagermoen, M. S. (2007). To correct and toacknowledge: Two simultaneous and conflicting perspectives oflimit-setting in mental health nursing. Journal of Psychiatricand Mental Health Nursing, 14(1), 41–48.doi:10.1111/j.1365-2850.2007.01037.x

Whittington, D., & McLaughlin, C. (2000). Finding time forpatients: An exploration of nurses’ time allocation in an acutepsychiatric setting. Journal of Psychiatric and Mental HealthNursing, 7(3), 259–268. doi:10.1046/j.1365-2850.2000.00291.x

Wright, S., Sayer, J., Parr, A. M., Gray, R., Southern, D., &Gournay, K. (2005). Breakaway and physical restrainttechniques in acute psychiatric nursing: Results from a nationalsurvey of training and practice. Journal of Forensic Psychiatryand Psychology, 16(2), 380–398.doi:10.1080/14789940412331270735

Wynn, R. (2003). Staff’s attitudes to the use of restraint andseclusion in a Norwegian university psychiatric hospital. NordicJournal of Psychiatry, 57(6), 453–459.doi:10.1080/08039480310003470

Wynn, R., & Mykleburst, L. H. (2006). Patients’ satisfaction andself-rated improvement following coercive interventions.Psychiatry, Psychology and Law, 13(2), 199–202.doi:10.1375/pplt.13.2.199



Corrigendumppc_337 183..184

Bak, J., Brandt-Christensen, M., Sestoft, D. M., & Zoffmann, V. (2011). Mechanical restraint—Which interventions prevent epi-sodes of mechanical restraint?—A systematic review. Perspectives in Psychiatric Care, 48(2), 83–94.

Under “Findings” and sub-heading “The Three Highest Ranked Interventions”:

1. The citations for “Robins, Sauvageot, Cusack, Suffoletta-Maierle, & Frueh, 2005” and “(Robins et al., 2005)” should beremoved.

2. The following two sentences have been changed.

“The study included papers published after 1998.”

The correct statement is:

“The study included papers published from 1998.”

And,

“Especially one of these papers (Robins et al., 2005) had a high evidence quality, and for that reason, the combination of CMTand patient involvement is regarded as the ‘best’ intervention in this review (Table 5, no. 1).”

The correct statement is:

“Especially this paper had a high evidence quality and for that reason, the combination of CMT and patient involvement isregarded as the ‘best’ intervention in this review (Table 5, no. 1).”

3. Several references have been amended in Table 5. The corrected Table 5 is below:

Table 5. Intervention Hierarchy


1 Implementation of cognitive milieu therapy reduces the number

of mechanical restraint episodes, among other things, through

patient involvement and empowerment.

3 (3–4) 1 Lykke et al., 2008; Thyrsting and Hall, 2008

2 Combined intervention programs are effective in avoiding

mechanical restraints, for example, patient participation,

patient education, staff education, programmatic changes,

high-level administrative endorsement, cultural changes, data

analysis, etc.

3 (n.a.–3) 2 Fisher, 2003; Hellerstein et al., 2007; Khadivi et al., 2004; Sullivan

et al., 2005; Taxis, 2002

3 Implementation of patient-centered care with a higher degree of

patients’ positive involvement in their own care reduces the

number of mechanical restraint episodes.

3 (3–4) 2 Fletcher and Stevenson, 2001; Sullivan et al., 2004; Thyrsting and

Hall, 2008

4 The identification of the patients’ crisis triggers unique calming

techniques and decisions on specific crisis interventions early in

the admission procedure, thereby reducing the number of


3 (4–3) 3 Jonikas, Cook, Rosen, Laris, and Kim, 2004; McCain and

Kornegay, 2005; Meehan, Mcintosh, and Bergen, 2006; Wynn

and Mykleburst, 2006

5 Introduction of validated risk assessment systems including

suggestions of care interventions reduces the number of


2 (n.a.–2) 4 Abderhalden et al., 2008; Cochrane-Brink, Lofchy, and

Sakinofsky, 2000; Dernevik, Grann, and Johansson, 2002

6 Effective treatment of substance abuse and psychosis reduce the

number of mechanical restraint episodes; the treatment of

mental illnesses is also described elsewhere in the review.

4 (n.a.–4) 3 Fletcher and Stevenson, 2001; Palmstierna and Olsson, 2007;

Sullivan et al., 2004; Thyrsting and Hall, 2008

7 Mandatory review from qualified clinical experts, for example, as

post-incident review, minimizes the number of mechanical

restraint episodes.

4 (n.a.–4) 3 Donat, 1998; Prescott, Madden, Dennis, Tisher, and Wingate,

2007

8 Better staff training reduces the number of mechanical restraints

(longer duration of the training, more staff participating, use of

different and involving education methods, higher quality of

the content, and greater diversity of the topics, for example,

legislation, crisis management, physical intervention strategies,

antecedents, warning signs, de-escalation techniques, etc.).

3 (3–4) 4 Bonner, Lowe, Rawcliffe, and Wellman, 2002; Bowers, Nijman,

et al., 2006; Delaney and Johnson, 2006; Hahn, Abderhalden,

Duxbury, Halfens, and Needham, 2006; Jonikas et al., 2004;

Lee et al., 2001; Mackay, Paterson, and Cassells, 2005;

Needham, Dassen, Halfens, Haug, and Fisher, 2005; Spokes

et al., 2002; Wright et al., 2005

doi:10.1111/j.1744-6163.2012.00337.x



Table 5. Continued


9 Staff attitudes such as positive care for the patients, respect,

collaboration, communication, humanism, empathy, and less

control and involuntary procedures reduce the number of


4 (4–4) 4 Allen, Carpenter, Sheets, Miccio, and Ross, 2003; Bowers, Alexander,

Simpson, Ryan, and Carr-Walker, 2007; Bowers, Brennan, Flood,

Lipang, and Oladapo, 2006; Cardell and Pitula, 1999; Duxbury,

2002; Gillig, Markert, Barron, and Coleman, 1998; Meehan et al.,

2006; Robins et al., 2005; Spokes et al., 2002

10 Higher patient–staff ratio reduces the number of mechanical

restraints.

4 (n.a.–4) 4 Smith et al., 2005; Wynn, 2003

11 The higher educated the staff, the fewer episodes of mechanical

restraints.

3 (3–4) 5 Delaney and Johnson, 2006; Flannery, Stone, Rego, Walker, and

Farley, 2006; Gim et al., 1999; Mackay et al., 2005; Spokes et al.,

2002

12 The more experienced the staff, the fewer episodes of mechanical

restraints.

3 (3–4) 5 Bowers, Jeffery, et al., 2007; Delaney and Johnson, 2006;

Flannery et al., 2006; Janssen, Noorthoorn, Linge, and

Lendemeijer, 2007; Mackay et al., 2005; Spokes et al., 2002

13 If nurses change their focus from considering the patient as

deviant to being a resource in his/her own treatment, it would

decrease the number of mechanical restraint.

3 (3–n.a.) 5 Vatne and Fagermoen, 2007

14 Early and adequate administration of oral p.n. medication guided

by the individual patients’ problems, medication experiences,

and preferences leads to fewer mechanical restraint episodes.

4 (4–4) 5 Allen et al., 2003; Goedhard, Stolker, Nijman, Egberts, and

Heerdink, 2007; McCain and Kornegay, 2005; Meehan et al.,

2006; Thapa et al., 2003

15 Debriefing, defusing, and crisis intervention minimize the number


4 (5–4) 5 Bonner et al., 2002; Bowers, Simpson, et al., 2006; Flannery

et al., 2006; Jonikas et al., 2004

16 Patients’ response to rules can be positive and lead to fewer cases

of mechanical restraint, for instance, if staff helps patients to

understand not only the rules, but also the rationale for the

presence of such rules. If the staff manages to make patients

understand that the rules are imposed out of concern for their

welfare, then patients may be less reluctant to following these

rules.

4 (5–4) 5 Alexander, 2006; Duxbury, 2002; Gillig et al., 1998; Meehan

et al., 2006; Robins et al., 2005

17 Activity-based nursing interventions, for example, physical

activities, are effective in avoiding mechanical restraints.

4 (5–4) 5 Meehan et al., 2006; Thomas, Jones, Johns, and Trauer, 2006

18 Several mental and physical working environmental factors

influence positively the number of mechanical restraints, for

example, good management as a high degree of order,

organization, or support; adequate staffing; psychiatric training;

rotation; education; no crowding; emergency backup, etc.

Several of these factors are described elsewhere in the review.

4 (5–4) 5 Bonner et al., 2002; Bowers, Allan, Simpson, Jones, and

Whittington, 2008; Daffern, Mayer, and Martin, 2006; Owen,

Tarantello, Jones, and Tennant, 1998; Whittington and

McLaughlin, 2000

19 Less crowding, for example, less stimuli and less environmental

stress, and more personal space decrease the number of


4 (5–4) 5 Meehan et al., 2006; Vaaler, Morken, Fløvig, Iversen, and Linaker,

2006

20 Higher regular staff and lower substitute ratio reduce the number


4 (5–4) 5 Bonner et al., 2002; Bowers, Allan, Simpson, Nijman, and

Warren, 2007; Spokes et al., 2002

21 Staffs’ opposition toward mechanical restraint because of the

perception that mechanical restraint is ethically problematic

decreases the number of mechanical restraints.

4 (n.a.–4) 5 Lind, Kaltiala-Heino, Suominen, Leino-Kilpi, and Valimaki, 2004;

Van Doeselaar, Sleegers, and Hutschemaekers, 2008

22 Alarm systems, for example, video surveillance and radio systems,

and well-trained response teams reduce the number of


4 (n.a.–4) 5 D’Orio, Purselle, Stevens, and Garlow, 2004

23 Better care alliance between patients and the nursing staff reduces

the number of mechanical restraints.

4 (n.a.–4) 5 Wynn, 2003

24 Communication especially communicative de-escalation reduces

the number of mechanical restraint episodes.

4 (n.a.–4) 5 Duxbury and Whittington, 2005

25 Well-maintained and familiar surroundings reduce the number of


4 (n.a.–4) 5 Vaaler, Morken, and Linaker, 2005

26 Music as diversion and relaxation before and after mechanical

restraint reduces the number of mechanical restraint episodes.

4 (n.a.–4) 5 Janelli and Kanski, 1998

27 Separation of acutely disturbed patients reduces the number of


5 (5–n.a.) 5 Meehan et al., 2006

No. refers to the number in the hierarchy where 1 is highest/best. Recommendation describes the whole body of evidence regarding a specific interven-tion level of scientific “truth” or its potential for clinical application. To show what kind of evidence that mostly influenced the combined recommenda-tion, the individual recommendations are shown in brackets, first, the qualitative recommendation grade and, second, the quantitative recommendationgrade. All from 1 to 5, where 1 is the highest/best. n.a. referred to no available paper. Effect refers to the reduction in the number of mechanicalrestraints, from 1 to 5, with 1 being the best, from 80% to 100% reduction, 2 from 60% to 80%, etc.p.n., pro necessitate (as needed).

Corrigendum


Paper 2

Coercion within Danish psychiatry

compared with 10 other European countries.

Coercion within Danish psychiatry compared with 10 other European countries JESPER BAK , HELLE AGGERN Æ S

Bak J, Aggern æ s H. Coercion within Danish psychiatry compared with 10 other European countries. Nord J Psychiatry 2012;66:297–302.

Background: In 2008, the European Committee for the Prevention of Torture and Inhuman or Degrading Treatment or Punishment (CPT) criticized the use of mechanical restraint in Denmark and referred to it as ill-treatment. What do other European countries do better? To answer this question, we compared the use of coercive measures regarding psychiatric inpatients in Denmark and comparable European countries. Aims: Comparing coercive measures from Denmark, Sweden, Norway, Finland, Iceland, Belgium, The Netherlands, United Kingdom, Ireland, France and Italy. Methods: Review of international literature and a cross-sectional study performed as a questionnaire survey. Results: Denmark used more mechanical restraint and holding than Finland and Norway; however Sweden used twice as much as Denmark. Finland used more seclusion than did the other countries. Norway was the country that used the smallest amount of physical coercion. Only Norway, Finland, Sweden and Denmark had comparable representative data on coercion. Conclusions: Norway used less physical restraint than Denmark. We could not fi nd any obvious reasons for the differences in the use of physical restraint. Clinical implications: Comparing the factors surrounding coercion between countries may serve as a basis for minimizing coercion and carrying it out in the most acceptable manner for the patients, thereby providing better psychiatric treatment in Europe.

• Coercion , Mental health , Psychiatric code , Restraint.

Jesper Bak, Head Nurse, R.N., S.D., M.P.H., Ph.D. student, Mental Health Centre Sct. Hans, Boserupvej 2, 4000 Roskilde, Denmark, E-mail: [email protected]; Accepted 12 October 2011.

In the autumn of 2008, public interest on coercion was

rekindled following the European Committee for the

Prevention of Torture and Inhumane or Degrading Treat-

ment or Punishment ’ s visit to Denmark. The Committee

criticized the use of mechanical restraint in terms of

duration and frequency and labelled it ill-treatment.

Before changing the psychiatric legislation in Denmark,

the Minister of Health requested that a comparison

between Denmark and comparable European countries

were made on the use of coercive measures targeted at

psychiatric inpatients together with a professional expert

evaluation on the coercive measures already used in

Denmark.

The National Board of Health appointed the authors

to help prepare a clearing report (1). The report was

modifi ed according to discussions in an appointed group

of non-governmental organizations, academics and pro-

fessionals prior to publication.

Aim To compare the use of coercive measures on psychiatric

inpatients between Denmark (DK) and Sweden (SE),

Norway (NO), Finland (FI), Iceland (IC), Belgium (BE),

The Netherlands (NL), United Kingdom (UK), Ireland

(IE), France (FR) and Italy (IT), in order to share

knowledge and learn about what other countries do to

minimize coercion.

Material and methods Data were collected by:

1) Reviewing international reports, research literature

and national registers;

2) Performing a survey, asking government offi cials and

medical specialists to complete a questionnaire.

Literature review The review covered legislation and coercive measures

used in psychiatry across Europe comprising: forced

medication/long period, forced medication/short period

(in some papers defi ned acute medication), mechanical

restraint, seclusion/isolation, physical restraint/holding,

time out and constant observation.

© 2012 Informa Healthcare DOI: 10.3109/08039488.2011.632645

Nor

d J

Psyc

hiat

ry D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y D

anm

arks

Vet

erin

aer

& J

ordb

rugs

bibl

iote

k on

10/

22/1

2Fo

r pe

rson

al u

se o

nly.

JBAK0003

Tekstboks

J BAK & H AGGERNÆS

298 NORD J PSYCHIATRY·VOL 66 NO 5·2012

Major differences between countries were found in

almost every area under investigation:

• Number of psychiatric beds per 100,000 inhabitants:

from 8 in IT to 152 in BE.

• Number of forced admissions per 100,000 inhabitants:

from 31 in IT to 218 in FI.

• Mean length of stay in days: from 14 days in IT to

52 days in UK.

• Number of forensic psychiatric beds per 100,000

inhabitants: from 0 in NO to 10.3 in BE.

• Number of psychiatric beds in jails per 100,000

inhabitants: from 0 in several countries to 4.6 in NL.

• Year of last revision of the psychiatric code; from

1990 in FR to 2009 in NL.

• National register for coercion: four countries had

national registers and seven countries had none.

• Number of psychiatrists per 100,000 inhabitants: from

7 in IE to 26 in FI.

• Number of nurses working in Mental Health Care:

from 33 in IT to 163 in FI.

No formal statistics were performed because of the low

number of respondents.

The different use of coercive measures Table 2 shows the different types of coercive measures

that can be applied in the countries, as reported in the

questionnaire survey (2009) where specialists from EPSO

and UEMS have described the legislation and clinical use

of the coercive measures in their countries.

As shown in Table 2, all countries allowed the use of

forced medication in some form in 2009. In NL and FR,

only acute medication was allowed. UK was the only

country where mechanical restraint was not allowed and

DK was the only country where seclusion was not

allowed.

Physical restraint/holding was used in two different

ways: for a short period e.g. while the patient was

restrained either during forced medication or for longer

periods to calm the patient. Some countries argue that

holding a patient for a longer period, especially very a

psychotic patient, was more intrusive to the patient than

mechanical restraint and seclusion.

Time out and constant observation were evaluated

differently between countries. Some saw it as coercion,

others as “ best practice ” .

The number of physically coercive measures per 100,000 inhabitants Table 3 shows the numbers of the three most frequent

physical coercive measures: mechanical restraint, seclu-

sion and holding. The numbers were obtained from

national registers (DK, NO, FI, NL), national reports

(SE) and scientifi c papers (UK, IT) (12 – 19).

The coercive measures were mostly defi ned as sug-

gested by Bowers et al. (2).

The review was conducted between April and June

2009, with an update in March 2010, and covered the

years from 1999 to 2010.

The searches included literature in the languages:

English, Danish, Norwegian, Dutch, Finnish and Swedish.

The search strategy was structured:

1) An overall search for legislation and coercion across

selected databases;

2) A search to identify registers and reports, e.g. govern-

mental reports not indexed in the major databases;

3) A topic-specifi c search on PubMed;

4) Reference lists of the most relevant papers were

checked for articles not otherwise found.

All references were placed in an electronic reference-

managing program (Ref. Man 11).

Once the references were retrieved, the following sift-

ing process was initiated:

• First sift — sift for material that potentially met eligi-

bility criteria based on title/abstract.

• Second sift — full reference that appeared relevant and

eligible or where relevance/eligibility was not clear

from the abstract where scrutinized.

• Third sift — lastly, the references were critically

assessed.

The focused search strategy is shown in Supplementary

Appendix 1 to be found online at http://informahealthcare.

com/doi/abs/10.3109/08039488.2011.632645.

Questionnaire survey A questionnaire was developed and covered the same

areas as mentioned earlier in the literature review.

The questionnaires were sent to governmental health

authorities (European Platform of Supervisory Organisa-

tions, EPSO) and the Psychiatric Section of the European

Union of Medical Specialists (Union Europ é enne des

m é dicins Sp é cialistes, UEMS). An example from the

questionnaire appears in Supplementary Appendix 2 to

be found online at http://informahealthcare.com/doi/abs/

10.3109/08039488.2011.632645.

The questionnaire survey was conducted between May

and July 2009.

Results Background data The background data in Table 1 were collected from several

sources: national registers, governmental and specialist

reports, research papers and the questionnaire survey

(3 – 11).

Nor

d J

Psyc

hiat

ry D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y D

anm

arks

Vet

erin

aer

& J

ordb

rugs

bibl

iote

k on

10/

22/1

2Fo

r pe

rson

al u

se o

nly.

JBAK0003

Tekstboks

COERCION IN EUROPE

NORD J PSYCHIATRY·VOL 66 NO 5·2012 299

Table 1 . Background data.

DK SE NO FI IS BE NL UK IE FR IT

Psychiatric beds per 100,000 inhabitants (3) (2007) 61 54 119 72 n.a. 152 114 23 94 95 8

Number of forced admissions per 100,000 inhabitants (4 – 7) (1999 – 2001) 70 114 135 218 n.a. 47 44 93 74 103 31

Mean length of stay in days (3, 8) (1998 and 2007) 30 25 29 38 n.a. 25 n.a. 52 16 36 14

Forensic psychiatric beds per 100,000 inhabitants (9, 10) (2002 – 2007) 6.6 8.0 0 6.9 2.3 10.3 9.8 6.1 2.2 0.8 2.2

Psychiatric beds in jails per 100,000 inhabitants (9, 10) (2002 – 2007) 0 1.6 0 1.2 0 2.1 4.6 0 0 1.2 n.a.

Year of last revision of the psychiatric code (Survey, 2009) 2006 2008 2006 2001 n.a. 2007 2009 2007 2001 1990 n.a.

National register for coercion (Survey, 2009) X (X) X X X

Psychiatrists per 100,000 inhabitants (3) (2007) 11 24 16 26 n.a. 23 15 13 7 22 10

Nurses working in Mental Health Care per 100,000 inhabitants (3, 11) (2007) 64 73 n.a. 163 n.a. n.a. 122 52 126 n.a. 33

n.a.: No available information.

Year: The available data were recorded from 1998 to 2008. The numbers of forced admissions: DK: 2000, SE: 1998, NO: 2001, FI: 2000, BE: 1998,

NL: 1999, UK: 1998, IE: 1999, FR: 1999, IT: 2001. Mean length of stay was from 2007 only NO.UK and FR were from 1998.

Mean length of stay: The median number of days in mental hospitals and community-based psychiatric inpatient units.

DK: The number of forced admissions includes forced compulsory retention of voluntarily admitted patients; the number on forced admissions alone

would be 34. “ Nurses working in Mental Health Care ” refers to all nurses in the number of specialized nurses that was 9 (3, 11). The length of stay is

only from community-based psychiatric inpatient units.

SE: National register for coercion should start in 2009. The length of stay was only from community-based psychiatric inpatient units.

NO: The number of forced admissions is from 2001.

FI: The length of stay is only from mental hospitals.

IS: IS did not have a Psychiatric code, but a § in the law regarding guardianship “ L ö gr æ ð isl ö gin ” from 1997.

BE: The number of beds included some ambulatory inpatients.

UK: The number of psychiatric beds does not include Scotland. The number of forced admissions includes forced compulsory retention of voluntarily

admitted patients; the number of forced admissions alone would be 48. The mean length of stay includes forensic patients. The last version of the

approved mental health legislation in Scotland was before 2005.

FR: The number of forced admissions was adjusted according to Barbato ’ s comments.

IT: The number of beds only represented Sardinia. The number of forced admissions only represented the Lombardia region, but numbers from whole

IT in 1997 was 26 and quite similar. IT had psychiatric beds in jails; however, the number of beds was uncertain. The length of stay was only from

community-based psychiatric inpatient units. Year of last revision of the psychiatric code was uncertain, we only know that it was after 2005 (3).

Table 2 . Applied methods of coercion.


Forced medication/long period X X X X X X X

Forced medication/short period X X X X X X X X X X

Mechanical restraint X X X X X X X X X X

Seclusion/isolation X X X X X X X X X X

Physical restraint/holding X X X X X X X X

Time out X X X X X X X

Constant observation X X X X X X X X X

Are other types of coercion used X X X X X X X

DK: If patients complained about the use of forced medication, it would give 2 – 3 weeks delayed effect before medication could be initiated. During

mechanical restraint, constant observation was required. Seclusion had only been allowed since 2005 and only in a particularly high security department

for extraordinarily dangerous psychiatric patients. Physical restraint was only used few minutes in connection with other coercive measures. Constant

observation as a coercive measure was fi rst reported if it was prolonged more than 24 h.

SE: Seclusion was only used if the patient was quite dangerous for other patients. Time out was regarded as isolation (Time out was defi ned as being

placed in a room with the door unlocked).

NO: Seclusion was only allowed in acute situations and was not to be used if patients were under 16 years. Time out was not regarded as coercion.

Constant observation was fi rst regarded as a coercive measure after 12 – 24 h.

FI: During mechanical restraint, constant observation was required if patients were under 18 years. During seclusion, constant observation was required

if patients were under 18 years. Physical restraint was rarely used in adult psychiatry and could only be used in child psychiatry if parents agreed.

IS: Mechanical restraint has not been used as a routine measure for decades. Physical restraint was only used a few minutes in connection with forced

medication and to calm patients. Constant observation was not regarded as coercion.

BE: Beside three-point restraint (waist, hand and foot), chest belts and straitjackets were allowed. During mechanical restraint and seclusion, the staff

could use camera surveillance and should always be able to hear the patient. Time out was only allowed max 48 h.

NL: Today, mechanical restraint is rarely used in acute psychiatric wards. The use in chronic wards was unknown.

UK: The forced medication was only allowed if it was part of the total treatment plan. Seclusion was only allowed in acute situations.

FR: Seclusion was described in a clinical guideline. If seclusion was prolonged more than 24 h, the patient was transferred to a special unit (Unites pour

maladies diffi ciles). Physical restraint was used in relation to forced medication and acute situations to calm patients and to prevent seclusion.

Nor

d J

Psyc

hiat

ry D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y D

anm

arks

Vet

erin

aer

& J

ordb

rugs

bibl

iote

k on

10/

22/1

2Fo

r pe

rson

al u

se o

nly.

JBAK0003

Tekstboks

J BAK & H AGGERNÆS


NL, NO, FI and DK had national registers. The num-

bers from NL were reported as very uncertain and could

easily be twice as high (20). The numbers from SE covered

around 20% of the population (not including the capital

region); data were deemed reasonably representative. The

data from UK and IT came from small samples, so the

data were deemed not representative. From the remain-

ing countries, it was not possible to fi nd any data.

Comparing the data from the three countries with

national registers that is DK, NO, FI and the numbers

from SE, which covered a reasonable part of the country,

we could see that DK more often used mechanical restraint

and holding than FI and NO. SE more often used mechan-

ical restraint than the other countries. FI more often used

seclusion than the other countries. DK more often used

physical restraint/holding than the other countries. NO was

the country that used the smallest amount of physical coer-

cion. It was reasonable to compare these four countries

since they were somewhat similar with respect to legisla-

tion, culture and clinical use. The countries that were most

alike probably were NO and DK, whereas the culture and

legislation were different in FI and SE.

Most intrusive coercive measures Table 4 describes the most intrusive coercive measures

obtained from the questionnaire survey (2009). The

respondents were asked to rank the coercive measures,

from 1 to 8, 1 indicating the measure perceived as the

least intrusive and 8 the most intrusive.

Table 3 . Physical coercion/100,000 inhabitants.


Mechanical restraint 34 70 21 20 n.a. n.a. 14 0 n.a. n.a. 7

Seclusion/isolation 0 31 2 43 n.a. n.a. 34 4 n.a. n.a. n.a.

Physical restraint/

holding

18 0 6 8 n.a. n.a. n.a. 13 n.a. n.a. 7

n.a.: Not available.

DK: 2007. The numbers included geronto-, child/adolescence and forensic

psychiatry (12). In DK, it was possible for the patients to be subjected to

voluntary mechanical restraint, which happened in 10% of all incidents.

NO: 2007. The numbers included geronto-, adolescence and forensic

psychiatry (13).

FI: 2007. The numbers included geronto-, adolescence and forensic

psychiatry (14).

SE: 2007. The numbers came from control visits in the southern hospital

region made by the National Board of Health and Welfare. The numbers

were then extrapolated to the whole of SE. The numbers included geronto-,

adolescence and forensic psychiatry (15, 16).

NL: 2006. The numbers came from the Dutch hospital database (IGZ) (17).

Mechanical restraint was only used for protection in geronto-psychiatry.

UK: 2002. The numbers came from three hospitals in London. The numbers

were then extrapolated to the whole of the UK. The numbers did not

include geronto-, adolescence and forensic psychiatry (18). Steinert et al.

described a total of seclusion and restraint in 2005 – 2007 between 52 and

77 (Wales 40) (19).

IT: 2002. The numbers came from two hospitals in the Modena region.

The numbers were then extrapolated to the whole of IT (18).

Table 4 . Ranking of coercive measures with respect to perceived intrusion.


Forced medication/long

period

5 6 8 3 5 4 3

Forced medication/short

period

4 2 7 2 8 1 2

Mechanical restraint 7 4 6 3 5 7 7

Seclusion/isolation 6 5 4 2 6 1 5 6

Holding/physical restraint 3 3 3 1 4 3 2 3 4

Time out 1 1 1 1 2 1

Constant observation 2 2 2 2 5

Mechanically restraint

ambulatory

8 7 5 5 4

Blank: Coercive measures were not in use or allowed in the country or no

reply was given, or it was ranked as “ 0 ” .

Forced medication/short period: Was defi ned as medicine given against the

patient ’ s will in an emergency (not medical treatment for a longer time

according to treatment plan).

Forced medication/long period: Was defi ned as medicine given against the

patient ’ s will, as treatment (medical treatment for a longer time according

to treatment plan).

Mechanically restraint ambulatory: Was defi ned as the use of restraint

devices, which allow patients to be out of bed and to walk around.

IE: No answer.

IT: No answer.

The answers from the European countries refl ected

culture based differences in how coercion was perceived,

e.g. forced medication/long period was considered worst

in NO; forced medication/short period in BE; mechanical

restraint in FI, IS and FR; seclusion in the UK; holding/

physical restraint in NL and mechanical restraint ambu-

latory in SE, DK and IS.

The answers also showed that in the countries using

both seclusion and mechanical restraint, mechanical

restraint was regarded as the most intrusive.

Discussion and conclusions To our knowledge, this is the fi rst comparative study on

coercion in psychiatry trying to combine both scientifi c

papers and questionnaire surveys from 11 European

countries.

Background variables A statistical evaluation of the background data (Table 1)

was not possible because of sample size, differences in

registration practice, data collection, legislation and

framework. Furthermore, some countries allowed the pri-

vate sector to use coercion, others not. In some countries,

this also included the forensic area and data from the

private sector was not reported.

We found no explanation for the lower number of psy-

chiatric beds, forced admissions and shorter mean length

of stay in IT. One explanation might be, as mentioned

Nor

d J

Psyc

hiat

ry D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y D

anm

arks

Vet

erin

aer

& J

ordb

rugs

bibl

iote

k on

10/

22/1

2Fo

r pe

rson

al u

se o

nly.

JBAK0003

Tekstboks

COERCION IN EUROPE

NORD J PSYCHIATRY·VOL 66 NO 5·2012 301

these discrepancies did not affect the conclusions made

above.

The number of physically coercive measures per 100,000 inhabitants As mentioned in the results section, the only countries,

of the 11 under investigation, that provided national pub-

lic representative data were NO, FI, SE and DK. We

believe that NL after 2006 has qualifi ed their data and

SE yet has not started, as planned, collection of nation-

wide data. Why other countries in Europe, beside the

Nordic countries, are not able to provide data on physical

coercive measures are unknown. One reason could be

lack of public openness; another could be that they have

a minor tradition for measuring and working with contin-

uous quality improvement, or that some countries have

both private and public psychiatric hospitals, complicating

the collection of data, but we do not know.

What we can see is that NO used physical coercion

less than FI, SE and DK; the reason for this is unknown.

However, between 1998 and 2006, NO allocated an extra

US $ 576 million in running costs and US $ 789 million in

investments (in buildings and so on), a total of US $ 1.4

billion to the psychiatric area (23). During approximately

the same period, the economic growth in the psychiatric

area in DK was US $ 189 million (24). In addition, NO

has a long tradition of specialized education in environ-

mental therapy of nursing staff and the number of psy-

chiatrists in NO is 31% higher than in DK but 50%

lower than SE and 63% lower than FI. Also, legal treat-

ment practice in NO does not have the delaying effect

of patient ’ s complaints on forced medication. During this

2 – 3-week period, where the Danish patients are not med-

icated, they are probably exposed to more physical coer-

cion. How the legal treatment practices are in SE and FI

we do not know exactly. Lastly, NO used twice as much

forced admission as DK, and maybe some patients,

because of earlier treatment were exposed to less physi-

cal coercion; on the other hand FI uses even more forced

admissions. Yet the fi gures on forced admissions was

somewhat insecure because DK had, in 2010, 2638

forensic psychiatric patients (25), whereas NO had

mostly around 500 (we do not know the exact number)

and the forensic psychiatric patients were not a part of

the fi gure in Table 1. How many forensic psychiatric

patients FI and SE had we do not know but the number

of forensic psychiatric beds where quite similar to DK.

Finally, caution must be shown when trying to com-

pare coercive measures between the different countries.

The differences in legislation, nursing and treatment cul-

ture, framework and economics make it important that

the European countries on a national level produce pub-

lic data on coercion. If not, the basis for research, bench-

marking and the possibility to learn from each other is

made much more diffi cult.

earlier, not representative data, or the effect of factors not

under investigation in this paper, e.g. attitudes towards

authorities or the families ’ role in caring for the patients.

The different types of coercive measures All countries allowed the use of forced medication in

some form. Differences in culture and history are proba-

bly the main reason for allowing and using different

coercive measures. For instance, we do not know whether

holding is used for a longer or shorter period and what

the patients perceive as more or less intrusive. This dis-

cussion could also be infl uenced by economy, because

holding for a long period necessitates competent and

expensive staff resources.

There were few answers from UEMS and in some

cases; they were contrary to the answers from EPSO.

The reason could be that in some countries, legislation is

implemented in practice after a delay of several months

to years.

Most intrusive coercive measures Five of the nine countries in Europe answering the

question believe that mechanical restraint, in some form,

was the most intrusive coercive measure towards the

patients. An often-mentioned argument in countries

using mechanical restraint, against using seclusion, is

the high amount of medication used to calm the patients,

preparing them for seclusion. The only study we could

fi nd describing this problem was “ Comparisons of psy-

chotropic drug prescribing patterns in acute psychiatric

wards across Europe (21) ” , not describing the actual

relationship between coercion and medication but more

in general. What could support the above-mentioned

idea is that England is the country using the highest

dose of anti-psychotic drugs and at the same time the

only country not using mechanical restraint. On the

other hand, they used the smallest amount of anxiolytic/

bezodiazepine drugs. The study did not incorporate

depot-medication, it was a small-scale pilot study and

overall not able to confi rm or reject the above idea.

However, whether mechanical restraint or seclusion is

considered more or less intrusive towards patients is in

fact unknown. In a small randomized study, Steinert &

Bergk (22) describes that “ both from ethical and safety

aspects the results do not yield evidence to prefer or for-

bid one of the interventions. Clinical decisions should

take into account patients ’ preferences. ” Of course, con-

sidering patients ’ preferences requires that clinicians have

the option to choose between the two kinds of coercion,

which is not the case in the UK and DK.

In the questionnaire, the countries were asked to rank

all the coercive measures, not all countries did answer as

requested. Some countries used the same number twice,

some had not used all the numbers and some had not

ranked measures they did not use themselves. However,

Nor

d J

Psyc

hiat

ry D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y D

anm

arks

Vet

erin

aer

& J

ordb

rugs

bibl

iote

k on

10/

22/1

2Fo

r pe

rson

al u

se o

nly.

JBAK0003

Tekstboks

J BAK & H AGGERNÆS


Salize HJ, Dressing H, Peitz M. Compulsory admission and 6. involuntary treatment of mentally ill patients — Legislation and practice in EU-member states (Final Report). Mannheim, Germany: Central Institute of Mental Health; 2002. Salize HJ, Dressing H. Epidemiology of involuntary placement of 7. mentally ill people across the European Union. Br J Psychiatry 2004;184:163 – 8. Aggern æ s H, Cubli-Bauer A, Zarotti G, Schudel W, Saliba J, 8. Pietrowski A, Martin V. Psychiatric legislation in Europe 1998. Report of the UEMS Section for Psychiatry. Prague: European Board of Psychiatry; 2001. Salize HJ, Dressing H, Kief C. Placement and treatment of 9. mentally ill offenders — Legislation and practice in EU member states. Mannheim: Central Institute of Mental Health; 2005. Salize HJ, Dressing H, Kief C. Mentally disordered persons in 10. European Prison Systems — Needs, Programmes and Outcome (EUPRIS). Mannheim, Germany: Central Institute of Mental Health; 2007. Dansk Sygeplejer å d. Medlemsstatistik — Januar 2009. K ø benhavn: 11. Dansk Sygeplejer å d; 2009. Sundhedsstyrelsen. Nye tal fra sundhedsstyrelsen. Anvendelse af 12. tvang i psykiatrien 2007. K ø benhavn: Sundhedsstyrelsen; 2008. Bremnes R, Hatling T, Bj ø rngaard JH. Bruk av tvangsmidler i 13. psykisk helsevern i 2001, 2003, 2005 og 2007. Trondheim: SINTEF Helse; 2008. Report No. SINTEF A8231. Psykiatrian erikoisalan laitoshoito 2007. National Research and 14. Develope Centre for Welfare and Health; 2008. Report No. 36/2008. Beskrivning av v å rdutnyttjande i psykiatrin — En rapport baserad p å 15. h ä lsodataregistren vid Socialstyrelsen. Socialstyrelsen; 2008. Report No. 2008-131-31. Den psykiatriske tv å ngv å rden — Verksamhetstillsyn av v å rden enligt 16. LPT och LVR i s ö dra sjukv å rdsregionen. Malm ö : Socialstyrelsens regionala tillsynsenhet i Malm ö ; 2008. Report No. 2006-126-60. Inspectie voor de Gezondheidszorg. Tabellenboek Wet Bopz 17. 2002 – 2006. Haag: Staatstoezicht op de volksgezondheid; 2007. Bowers L, Douzenis A, Galeazzi G, Forghieri M, Tsopelas C, 18. Simpson A, Allan T. Disruptive and dangerous behaviour by patients on acute psychiatric wards in three European centres. Soc Psychiatry Psychiatr Epidemiol 2005;40:822 – 8. Steinert T, Lepping P, Bernhardsgr ü tter R, Conca A, Hatling T, 19. Janssen W, et al. Incidence of seclusion and restraint in psychiatric hospitals: A literature review and survey of international trends. Soc Psychiatry Psychiatr Epidemiol 2010;45:889 – 97. Janssen WA, Noorthoorn EO, De Vries WJ, Hutschemeakers GJ, 20. Lendemeijer HH, Widdershoven GA. The use of seclusion in the Netherlands compared to countries in and outside Europe. Int J Law Psychiatry 2008;31:463 – 70. Bowers L, Callaghan P, Clark N, Evers C. Comparisons of 21. psychotropic drug prescribing patterns in acute psychiatric wards across Europe. Eur J Clin Pharmacol 2004;60:29 – 35. Steinert T, Bergk J. A randomised study comparing seclusion and 22. mechanical restraint in people with serious mental illness. Eur Psychiatry 2007;22:220. Helse- og omsorgsdepartementet. Om optrappingsplan for psykisk 23. helse 1999 – 2006 Endringer i statsbudsjettet for 1998. 1998. Report No. St. prp. nr. 63. Danske Regioner. Behandlingsgaranti i psykiatrien — Udfordringer 24. og muligheder. K ø benhavn: Danske Regioner; 2006. Danske Regioner. Retspsykiatri. Kvalitet og sikkerhed. K ø benhavn: 25. Danske Regioner; 2011. Jan B, Erich F, Tilman S. Coercion Experience Scale (CES) — 26. Validation of a questionnaire on coercive measures. BMC Psychiatry 2010;10:5.

Jesper Bak, Head Nurse, R.N., S.D., M.P.H., Ph.D. student, Research Institute, Mental Health Centre Sct. Hans, Boserupvej 2, 4000 Roskilde, Denmark. Helle Aggern æ s, Head of Clinic, M.D., Mental Health Centre Amager, Digevej 10, 2300 Copenhagen, Denmark.

Regarding research in the area, it is evident that there

is a lack of high-quality intervention studies. One of the rea-

sons is probably the lack of instruments that can measure

the degree of patients ’ intrusive experiences of coercion.

Only one validated instrument, the Coercion Experience

Scale (26), was found.

It is very important to have scales that measure the

degree of intrusive experience, because not all mechani-

cal restraint episodes are intrusive. From a public point

of view, it is worth noting that some people voluntarily

accept mechanical restraint, given that 10% of the psy-

chiatric patients that were mechanically restrained in DK

were restrained voluntarily. Despite their psychotic con-

dition, some patients seem to have such a high sense of

reality that they prefer coercion to the risk of hurting

themselves and/or others.

One could hope that in future psychiatry, we will

know when it is inevitable to choose the kind of coer-

cion that not only is the most acceptable to the individ-

ual patient, but also convincingly is the means of least

restriction. It is obvious from the reported diversity of

the coercive measures in the present data that no country

holds the optimal solution for dealing with coercion in

psychiatry.

Acknowledgements—The authors wish to thank Mette Dons (Head of

Department, the National Board of Health, Copenhagen), Torben

H æ rslev (Senior Specialist, the National Board of Health, Copenha-

gen) and Louise Gj ø rup (Administrative Assistant, the National

Board of Health, Copenhagen) for discussions and participation, and

Thomas Werge (Head of Research, Mental Health Centre Sct. Hans,

Roskilde) for valuable comments and careful reading during the

writing process.

Disclosure of interest: The project was partly funded by

the National Board of Health. There are no known con-

fl icts of interest, fi nancial, personal, political and aca-

demic or others. The authors are responsible for the

contents and writing the paper.

References Sundhedsstyrelsen. Tvangsforanstaltninger i psykiatrien — en 1. udredning. K ø benhavn: Sundhedsstyrelsen; 2009. Bowers L, van der Werf B, Vokkolainen A, Muir-Cochrane E, 2. Allan T, Alexander J. International variation in containment measures for disturbed psychiatric inpatients: A comparative questionnaire survey. Int J Nurs Stud 2007;44:357 – 64. World Health Organization. Policies and practices for mental 3. health in Europe — Meeting the challenges. Copenhagen: WHO Regional Offi ce for Europe; 2008. Barbato A, D ’ Avanzo B. Involuntary placement in Italy. Br J 4. Psychiatry 2005;186:542. Bremnes R, Hatling T., Bj ø rngaard JH. Tvungent psykisk 5. helsevern med d ø gnophold i perioden 2001 – 2006. Trondheim: SINTEF Helse; 2008. Report No. SINTEF A4319.

Supplementary material available online Supplementary Appendix 1 and 2.

Nor

d J

Psyc

hiat

ry D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y D

anm

arks

Vet

erin

aer

& J

ordb

rugs

bibl

iote

k on

10/

22/1

2Fo

r pe

rson

al u

se o

nly.

JBAK0003

Tekstboks

1

Supplementary Appendix 1: Focused search strategy No/database/host Search phrase/ ” -string ” Date Result 1st sift 2nd sift 3rd sift

GOOGLE Scholar Europa lovgivning psykiatri OR psykiatrisk OR psykisk OR psykotisk

tvangsfi ksering OR b æ ltefi ksering OR fi ksering OR tvang OR

tvangsmidler. Limiters — Publication Year from: 1999-, Area: Medicin,

farmakologi og veterin æ rvidenskab, Samfundsvidenskab, kunst og

humaniora

08 April 2009 24 8 4 1

GOOGLE Scholar Restrained OR Restraint OR Restraints OR Secluded OR Seclusion OR

Coercion OR Coercive OR Containment Legislation OR Legal

Comparison OR Differences OR Variations European OR Europe OR

Europeans Mental OR Psychiatric OR Psychiatry OR Mental health OR

Mentally ill, Limiters — Publication Year from: 1999-, Area: Medicin,

farmakologi og veterin æ rvidenskab, Samfundsvidenskab, kunst og

humaniora

15 April 2009 200 21 6 2

CINAHL and

AMED/

EBSCOhost

(Restrained OR Restraint OR Restraints OR Secluded OR Seclusion OR

Coercion OR Coercive OR Containment) and (Legislation OR Legal)

and (Mental OR Psychiatric OR Psychiatry OR Mental health OR

Mentally ill) and (Europe OR European OR Europeans), Limiters —

Publication Year from: 1999 – , Expanders — Also search within the full

text of the articles, Search modes — Boolean/Phrase

15 April 2009 21 3 1 1

/PubMed/

Silverplatter

(Ment * OR Psychiat * OR Psycho * ) AND (Restrain * OR Seclu * OR

Coerc * OR Containt * ) AND (Europe * ) AND (Compar * OR Difference *

OR Variat * )

15 April 2009 21 3 0 0

Criminal justice

abstracts/

CSA-

ILLUMINA

(Ment * OR Psychiat * OR Psycho * ) AND (Restrain * OR Seclu * OR


OR Variat * ), Date Range: 1999 to 2010

15 April 2009 2 1 1 1

British Journal of

Criminology

((Ment * OR Psychiat * OR Psycho * ) AND (Restrain * OR Seclu * OR


OR Variat * )) AND PDN ( � 1/1/1999)

15 April 2009 10 1 1 1

Cochrane Library “ Restrain * OR Seclu * OR Coerc * OR Containt * and Compar * OR

Difference * OR Variat * and Europe * and Ment * OR Psychiat * OR

Psycho * and Legislat * OR Legal * , from 1999 to 2009 in Cochrane

Database of Systematic Reviews ”

15 April 2009 13 0 0 0

PSYCINFO/

WebSPIRS/

OvidSP

(Restrain * or Seclusion or Coercion or Contain * ) and (Comparison or

Difference * or Variation * ) and Europe * and (Ment * or Psychiat * or

Psycho * ) and (Legislation * or Legal * ) (Including Related Terms)

(english language and abstracts and yr � “ 1999 – 2009 ” )

15 April 2009 42 9 3 2

ISI web of

knowledge

Topic � ((Ment * OR Psychiat * ) AND (Restrain * OR Seclu * OR Coerc *

OR Containt * ) AND (Legisla * OR Legal * ) AND (Compar * OR

Difference * OR Variat * ) AND Europ * )

15 April 2009 10 0 0 0

SCIRUS http://

www.scirus.

com/srsapp/

(Ment * OR Psychiat * ) AND (Restrain * OR Seclu * OR Coerc * OR

Containt * ) AND (Legisla * OR Legal * ) AND (Compar * OR Difference *

OR Variat * ) AND Europ * , 1999 – 2009, Law, Medicine, Neuroscience,

Pharmacology, Psychology, Social and Behavioral Sciences, Sociology

15 April 2009 14 0 0 0

GOOGLE

different

Universities,

National Health

Boards and

others

Restrained OR Restraint OR Restraints OR Secluded OR Seclusion OR

Coercion OR Coercive OR Containment OR Psychiatric Legislation

AND Denmark OR Sweden OR Norway OR Finland OR Iceland OR

The Nederlands OR United Kingdom OR Scotland OR Ireland OR

France OR Italy (Part of titles from earlier found papers reference lists,

ex: Psykiatrian erikoisalan laitoshoito, Anvendelse af tvang i

psykiatrien, Den psykiatriske tv å ngv å rden, Inspectie voor de

Gezondheidszorg, Bruk av tvangsmidler i psykisk helsevern and so on)

April – June

2009

? ? 63 40

Update March 2010 8 3

Sum 357 46 87 51

Supplementary material for Bak J, Aggern æ s H. Coercion within Danish psychiatry compared with 10 other European

countries. Nord J Psychiatry 2012;66:297–302.

2

Supplementary Appendix 2 Example from the questionnaire, the whole questionnaire can be required from the authors

Paper 3

Mechanical Restraint in Psychiatry:

Preventive Factors in Theory and Practice.

A Danish-Norwegian Association Study.

Mechanical Restraint in Psychiatry: Preventive Factors inTheory and Practice. A Danish–Norwegian Association StudyJesper Bak, RN, SD, MPH, Vibeke Zoffmann, RN, MPH, PhD, Dorte Maria Sestoft, MD, PhD,Roger Almvik, RN, RMN, Dr. Philos, and Mette Brandt-Christensen, MD, PhD

Jesper Bak, RN, SD, MPH, is PhD student, Head Nurse, Mental Health Centre Sct. Hans, Roskilde, Denmark; Vibeke Zoffmann, RN, MPH, PhD, is SeniorResearcher, Steno Diabetes Centre, Gentofte, Denmark; Dorte Maria Sestoft, MD, PhD, is Head of Department, Clinic of Forensic Psychiatry, Ministryof Justice, Copenhagen, Denmark, Roger Almvik, RN, RMN, Dr. Philos, is Research Director, Centre for Research & Education in Forensic Psychiatry, St.Olavs University Hospital, Forensic Department Bröset, Trondheim, Norway; and Mette Brandt-Christensen, MD, PhD, is Head of Research & Educationin Forensic Psychiatry, Mental Health Centre Sct. Hans, Roskilde, Denmark.

Search terms:Coercion, mental health, nursing, physicalrestraint, psychiatry

Author contact:[email protected], with a copy to theEditor: [email protected]

Conflict of Interest StatementThere are no known financial, personal, ethical,political, academic, or other conflicts ofinterest.

First Received April 5, 2013; Final Revisionreceived June 10, 2013; Accepted forpublication June 27, 2013.

doi: 10.1111/ppc.12036

PURPOSE: To examine how potential mechanical restraint preventive factors inhospitals are associated with the frequency of mechanical restraint episodes.DESIGN AND METHODS: This study employed a retrospective association design,and linear regression was used to assess the associations.FINDINGS: Three mechanical restraint preventive factors were significantly associ-ated with low rates of mechanical restraint use: mandatory review (exp[B] = .36,p < .01), patient involvement (exp[B] = .42, p < .01), and no crowding (exp[B] = .54,p < .01).PRACTICE IMPLICATIONS: None of the three mechanical restraint preventivefactors presented any adverse effects; therefore, units should seriously considerimplementing these measures.

In some countries, mechanical restraint (MR) can be per-formed legally when psychiatric inpatients pose a risk tothemselves or to others (Psykiatriloven, LBK nr. 1729, 2010).The procedure is primarily performed by nursing staff andinvolves the use of leather or fiber belts to secure the patient toa bed. The straps are also occasionally attached around thepatient’s wrists and ankles. MR exposes patients to potentialphysical risks, such as deep vein thrombosis, pulmonaryembolism, pressure sores, bone fractures, and even death(Hem, Steen, & Opjordsmoen, 2001; Hick, Smith, & Lynch,1999; Laursen, Jensen, Bolwig, & Olsen, 2005; Morrison &Sadler, 2001; Nielsen, 2005). For many patients, MR producesnegative psychological effects, such as feelings of helplessness,anger, being trapped, sadness, fear, displeasure, or offense.Patients often regard the use of MR as an injustice or even apunishment (Bonner, Lowe, Rawcliffe, & Wellman, 2002;Bower, McCullough, & Timmons, 2003; Cusack, Frueh,Hiers, Suffoletta-Maierle, & Bennett, 2003; Frueh et al.,2005). In the worst cases, MR can cause post-traumatic stressand lifelong trauma (Cusack et al., 2003; Frueh et al., 2005;Steinert, Bergbauer, Schmid, & Gebhardt, 2007). These nega-

tive consequences are inconsistent with healthcare providers’basic sense of “doing good” for the patients. Equally impor-tant arguments in favor of using MR are that healthcare pro-fessionals are obliged to secure and care for not only theindividual patient’s health and life, but also the health, lives,and rights of others. This second obligation requires provid-ers to prevent violence toward other patients, relatives, andstaff. Previous studies indicate that up to 10% of patientsrequest MR to prevent themselves from acting out (Bak &Aggernæs, 2012). Nevertheless, the use of MR generally posesproblems, requiring that greater effort be undertaken toreduce its frequency (Sundhedsstyrelsen, 2009).

Many factors or conditions probably influence the eventsthat lead to the use of MR. A lot of these conditions resultfrom daily activities while socializing with others in the psy-chiatric ward, and they can potentially be changed or pre-vented by mental-healthcare providers, who thereby have acrucial role in reducing the use of MR. Numerous actionshave been recommended to reduce the use of mechanical andphysical restraints (National Risk Management Programme,2007; The National Collaborating Centre for Nursing and

bs_bs_banner





Supportive Care, 2005). More than 30 reviews on the manage-ment of agitation, aggression, violence, and coercion (includ-ing MR, seclusion, manual restraint, and forced medication,among others) were identified; 10 explicitly included MR, andall described the no high-quality research. Thus, no causalrelationship between preventive interventions and the reduc-tion of MR episodes has been established (Johnson, 2010;Scanlan, 2010; Steinert et al., 2010; Stewart, Van der Merve,Bowers, Simpson, & Jones, 2010; Strout, 2010).

The most recent review (Bak, Brandt-Christensen, Sestoft,& Zoffmann, 2012) identified numerous MR-preventivefactors based on studies specifically targeting MR prevention.No MR-preventive factors found in the review were based onthe highest scientific-recommendation grade (see gradinglevels in the review). Furthermore, several of these MR-preventive factors might be effective only in controlledresearch environments. Thus, the applicability and impact ofthese MR-preventive factors in everyday practice remainunknown.

In 2008, a critical public debate concerning the use of MRin Denmark led to a comparative study on the use of coercionin 11 European countries, including Denmark (Bak &Aggernæs, 2012). Few of the included countries had nationalstatistics available on the use of MR. Finland, Norway, andDenmark had valid nationwide statistics, as did the Nether-lands and Sweden, to some extent (Bak & Aggernæs, 2012).Norway had the lowest rate of physical coercion (MR, seclu-sion, and physical restraint/holding), with 29 people per100,000 inhabitants having experienced physical coercion.Compared to Denmark, which had 122 episodes of MRper 100,000 inhabitants, Norway had 87 episodes of MRper 100,000 inhabitants in 2007 (Bremmes, Hatling, &Bjørngaard, 2008; Sundhedsstyrelsen, 2008). Because thelower frequency of MR episodes in Norway might indicatethat more effective MR-preventive factors have been imple-mented there compared to Denmark, we presumed thatincluding both countries in this study might generate newknowledge. In addition, a cross-country study was possiblebecause the national legislations, administrative structures,and cultures of these nations are similar.

Purpose

The goal of this study was to provide a basis for developingMR-preventive practices by answering the following ques-tion: How are presumed MR-preventive factors of nonmedi-cal origin associated with the frequency of MR episodes inDenmark and Norway?

Design and Methods

This paper uses Bowers et al.’s (2007) definition of MR: Theuse of restraining straps, belts, or other equipment to restrict

movement (Bowers et al., 2007). Note that this definitionrefers only to the restraint of inpatients in psychiatric hospi-tals.A retrospective association study was conducted to inves-tigate the degree to which the nonmedical MR-preventivefactors identified in a previous review (Bak et al., 2012) wereimplemented in the psychiatric units of Denmark andNorway. Next, we investigated whether the chosen factorswere associated with the number of MR episodes.

The following conditions must be present to legally initiateMR according to the Danish Mental Health Act (translated bythe authors):

Mechanical restraint may be used only when necessary toprevent a patient from1. exposing their body or health or the body or health ofothers to danger,2. pursuing or in any other way grossly molesting fellowpatients, or3. committing significant acts of vandalism.(Psykiatriloven, LBK nr. 1729, 2010)

These conditions are consistent with Norwegian legislation(translated by the authors):

Restraints may be used on a patient only when absolutelynecessary to prevent him or her from damaging him orherself or others or to prevent damage to buildings,clothing, inventory or other things. (Lov nr. 62, 1999)

The MR-preventive factors identified by the literature review(Bak et al., 2012) were included in an online questionnairesent to the clinical nurse managers for the purpose of quanti-fying whether they were implemented in the units. Thesemanagers were chosen because they had exclusive access tothe requested information. The units were the smallest orga-nizational collection of beds led by a clinical nurse manager.

Time Frame

The questionnaire was administered between December 23,2009, and January 27, 2011. The collected data pertained to2008 in Denmark and 2010 in Norway.

Included Participants

All psychiatric hospital units in Denmark (87) and Norway(96) that treated adult inpatients (18–65 years) in which MRhad been used for approximately two years prior to the year ofinvestigation were included in the study. The clinical nursemanager of each unit provided the requested information byanswering the questionnaire. The units were identified bycontacting the clinical heads of the departments. InDenmark, the departments were selected from the nationalregister on coercion; in Norway, they were selected by con-tacting the administrators of the health regions and theresearchers already working in the area.

Mechanical Restraint in Psychiatry: Preventive Factors in Theory and Practice. A Danish–Norwegian Association Study


Excluded Participants

The special high-security department in Denmark wasexcluded because specific legislation concerning seclusion/isolation was in effect. MR did occur in Norway in settingsother than hospitals; however, because only 1.5% of recordedNorwegian MR episodes in 2007 occurred in these settings(Bremmes et al., 2008) and because the primary focus of thisstudy was hospitals, these units were also excluded.

Excluded Outcome Data

If MR was used more than 25 times on the same patientduring the investigation period, then these data wereexcluded. Chance determined the units to which thesepatients were admitted, but their presence significantlyaffected the unit MR frequency, distorting the results.

Data Collection: The Questionnaire

A questionnaire was developed based on the results from anearlier systematic literature review that focused on factorsthat might reduce the use of MR (Bak et al., 2012). The foundMR-preventive factors can be found in Table 1. For a morethorough description, see Bak et al. (2012).

Every MR-preventive factor was“translated”into questionsto identify the degree to which the factor was present in theunits. Six factors were not included in the present study.Number 2 (combined intervention programs) was omittedbecause it was too complex to measure. The following factorswere omitted because they were not considered to be mea-sured appropriately by the questionnaire: Number 6 (treat-ment of substance abuse and psychosis), Number 13 (patientsperceived as a resource in their own treatment), Number 14

(pro necessitate/as-needed medication guided by individualpatients), Number 23 (care alliance), and Number 26 (musicused as a diversion and for relaxation). Numbers 1 (cognitivemilieu therapy and patient involvement) and 3 (patient-centered care and patient involvement) were separated intothree factors: cognitive milieu therapy, patient-centered care,and patient involvement. Although patient involvement ispart of cognitive milieu therapy and patient-centered care, itmight also occur by itself.

The questionnaire was developed by conceptualizing andoperationalizing the MR-preventive factors. To increasevalidity and reliability, we followed the Question AppraisalSystem (QAS-99; Willis & Lessler, 1999) that was developed toidentify problems in questionnaires, including the cognitiveprocesses inherent to the question-answering process.

The construction and pretesting of the questionnaireinvolved an eight-step appraisal of each item (QAS-99):reading, instructions, clarity, assumption, knowledge/memory, sensitivity/bias, response categories, and others.1

The pretest was conducted by creating a cognitive-test proto-col that included probes for each item. The probes covered theaforementioned eight steps and consisted of seven types:general, specific, understanding/interpretation, safety,

1 Step 1: Reading: Determine if it is difficult for the interviewers to read the

question uniformly to all respondents. Step 2: Instructions: Look for problems

with any introductions, instructions, or explanations from the respondent’s

point of view. Step 3: Clarity: Identify problems related to communicating the

intent or meaning of the question to the respondent. Step 4: Assumptions:

Determine whether there are problems with assumptions or the underlying

logic. Step 5: Knowledge/Memory: Check whether respondents are likely to

not know or have trouble remembering information. Step 6: Sensitivity/Bias:

Assess questions for sensitive nature or wording, and for bias. Step 7: Response

categories: Assess the adequacy of the range of responses to be recorded. Step

8: Other: Look for problems not identified in Steps 1–7 (Willis & Lessler,

1999).

Table 1. Hierarchy of Preventive Factors That Might Decrease the Use of MR

Rank MR-Preventive Factor Rank MR-Preventive Factor

1 Cognitive milieu therapy and patient involvement 15 Debriefing, defusing, and crisis intervention2 Combined intervention programs 16 Patient’s positive responses to rules3 Patient-centered care and patient involvement 17 Activity-based nursing4 Identification of the patient’s crisis triggers 18 Working environmental factors5 Risk assessment 19 No crowding6 Treatment of substance abuse and psychosis 20 Regular staff–substitute ratio7 Mandatory review (e.g., post-incident review) 21 Staff opposition to MR8 Staff training 22 Alarm systems9 Staff attitudes 23 Care alliance

10 Patient–staff ratio 24 Communication (communicative de-escalation)11 Staff education 25 Well-maintained and familiar surroundings12 Staff experience 26 Music used as a diversion and for relaxation13 Patients perceived as a resource in their own treatment 27 Separation of acutely disturbed patients14 p.n. medication guided by individual patients

Note. Rank refers to the decreasing order in the hierarchy of effect and recommendation grade (Bak et al., 2012). p.n. medication, pro necessitate/as-needed medication; MR, mechanical restraint.



paraphrasing, and memory (e.g., how did you reach thisanswer, how do you define physical activity, etc.).

The pretest consisted of cognitive interviewing (Willis,2005) four clinical nurse managers individually using concur-rent probing (Watt et al., 2008) in which verbal probes wereasked immediately after a question. This protocol led to therevision of seven questions that were subsequently pretestedin a group interview.

Data collection, administration, and processing, as well aserror identification and the creation of electronic data files(Olsen, 2006; Schaeffer & Presser, 2003), were administeredusing the electronic web-based survey system EnalyzerSurvey Solution, Version 2010 (Enalyzer A/S, Copenhagen,Denmark).

The questionnaire was checked for test–retest reliability byasking three clinical nurse managers to answer the question-naire a second time after a 3-month interval (the respondentswere instructed not to check their previous answers). Theestimated mean correlation between all of the first and secondanswers was reasonable (Kendall’s Taub [τKen,b] = .75). Withregard to the Norwegian portion of the study, the question-naire was translated and retranslated by JB and then checkedby two healthcare professionals who were familiar with thesubject. A pretest was conducted to test face validity throughcognitive interviewing seven clinical nurse managers. Thistest led to the revision of eight questions, mostly due toimprecise translations. Moreover, one question was added tothe questionnaire due to different staff competences: Psy-chologists in Norway are partially responsible for theirpatients’ treatment, which is not the case in Denmark.

Later, the outcome variable (i.e., the number of MR epi-sodes) was reexamined. This procedure was necessary due tothe confusing wording of a heading in the Danish question-naire and the lack of precise numbers regarding the Norwe-gian patients who experienced MR more than 25 times (14answer categories instead of the actual number were pro-vided, which made the exclusion imprecise). The reexamina-tion was accomplished by contacting all of the clinical nursemanagers in Denmark and four of the clinical nurse managersin Norway. The respondent rate was 94%, and the resultslowered the overall frequency of MR by 3.2%. A high correla-tion coefficient (τKen,b = .89) was found between the twoanswers, revealing that the confusing headline and the catego-rization error did not have an important effect.

From Review to Data Analyses

The flowchart in Figure 1 shows an example of transformingthe findings in the literature review to the final analyses. TheDanish and Norwegian questionnaire can be obtained fromthe authors.

All dichotomized MR-preventive factors can be found inTable 2.

The Outcome Variable

The outcome variable was the number of MR episodes perunit over 1 year.

Potential Confounders

Several conditions may be associated with both the frequencyof MR use and the presence of the preventive factors in theunits; therefore, they must be treated as covariates in the esti-mation of the associations between the preventive factors andthe outcome variable.

These conditions were geography (Bremmes et al., 2008;Keski-Valkama et al., 2007), unit size, bed occupancy, type ofbed, reason for the MR, forensic provision, ethnicity other

Data analysesFinally, we analyzed whether associations existed among the outcome variable (the numbers of MR incidents), the background variables, and the preventive factor.

QuestionnaireThe questionnaire was created using a conversion process in which the factor was “translated” into a question (e.g., did you audit, review or conduct a similar systematic follow-up after the MR events to analyze and learn from them?)The question was measured using five levels:

1. Yes, after all MR incidents2. Yes, after ¾ of all MR incidents3. Yes, after ½ of all MR incidents4. Yes, after ¼ of all MR incidents5. No

Literature reviewThe literature review revealed that two studies found that the preventative factor “Mandatory review” was most likely associated with the use of MR. Specifically, MR-reduction rates were between 36% and 62%.

Combining the questions into a dichotomized factorThe dichotomization of the five levels in the question was accomplished by examining the two papers from the literature review. One review was conducted after all MR episodes, and the other was conducted after six events or one long-lasting MR episode. The unit and department staff conducted the review in both papers. The above question does not answer whether this is the case in the actual unit. Therefore, we decided that the dichotomization should be conducted after all MR incidents, rather than after those that are fewer or newer.

Figure 1. Flowchart “From review to Data Analyses”An example of the possible MR-preventive factor “Mandatory review”from the literature review. Qualified clinical experts should conductmandatory post-incident reviews because they minimize the number ofMR episodes. MR, mechanical restraint.



Table 2. A Description of How the MR-Preventive Factors Were Measured

MR-Preventive Factors Measurement Description

Cognitive milieu therapy vs. nocognitive milieu therapy

The clinical nurse manager described the care in his or her unit as “cognitive behaviour therapy” or “cognitivemilieu therapy,” and the staff had participated in 40 hr of training or more in this care method—The caremethod was not described.

Patient involvement vs. no patientinvolvement

One of the following factors was present: the patients participated in ward rounds/conferences, influencedthe rules imposed by the unit, or influenced what was written in the nurse’s journal—Patients did notparticipate in these areas.

Patient-centered care vs. nopatient-centered care

The clinical nurse manger described the care in his or her unit using the words “individual,” and“cooperative,” or “patient-centred,” and the staff had participated in 40 hr of training or more in this caremethod—These words were not used to describe this care method.

Identification of the patient′s crisistriggers vs. no identification

The initial nurse evaluation of at-risk patients always contained a written evaluation of aggressive,self-mutilating, or suicidal behaviors (e.g., crisis triggers, unique calming techniques, and copingstrategies)—The initial evaluation only sometimes or never contained a written evaluation.

Risk assessment vs. no riskassessment

The risk of developing aggressive/violent or self-mutilating/suicidal behaviors was always systematicallyevaluated and documented, and care actions were initiated—These behaviors were only sometimes ornever evaluated and documented.

Mandatory review vs. no mandatoryreview

An audit, review, or similar systematic follow-up evaluation was conducted after all MR episodes—Theevaluation was only conducted after some or none of the MR episodes.

Staff training (> 50% vs. less) An average of 50% or more of the staff had participated in training regarding de-escalation, avoiding physicalattacks, careful restraining techniques, legislation, MR predictors/antecedents, and suicide prevention—Lessthan 50% of the staff had such training.

Positive staff attitudes vs. less positive All staff directed their attention toward the patient and possessed communicative skills, optimism, andempathy—Not all of the staff possessed these skills.

Patient–staff ratio (≥ 3 staff perpatient vs. fewer)

The staff-to-patient ratio was higher than or equal to three-to-one on average—This ratio was lower thanthree-to-one on average.

Staff education (≥ 3 years vs. fewer) The staff had received basic training for at least 3 years—The staff had less than 3 years of training.Staff experience (> 5 years vs. fewer) The experience of the staff in the psychiatric specialty was more than 5 years on average—Staff experience

was 5 years or fewer.Debriefing, defusing vs. none Follow-up evaluations were initiated after each MR episode for debriefing, defusing, or other crisis

interventions—No follow-up was initiated.Positive patient responses to rules vs.

negative responsesThe patients reacted positively to the rules, and limit-setting was motivational—The opposite occurred.

Activity-based nursing vs. noactivity-based nursing

Individual patient-activity plans were devised, and the patients participated in many or several physical orother-type activities—The patients had no activity plans, and they participated in few or no physical or otheractivities.

Acceptable work environment vs.poor work environment

Presence of a workplace assessment, no or only insignificant physical or mental work-environment problemsexisted, and the overall work environment was evaluated as acceptable—The work environment was notevaluated, more work-environment problems existed, and the overall work environment was evaluated aspoor.

No crowding vs. crowding Two of the following three factors were present for no crowding: only one bed in a patient’s room, more than25 m2 of all-day-accessible space per patient, and the perception of no crowding—Only one or none ofthese factors was present.

Substitute staff (< 50 vs. more) Substitute staff was used less than 50 times (8-hr shifts) over the year—These employees were used morethan or equal to 50 times.

Staff opposition to MR (many vs.some)

All staff had the attitude that “MR should be initiated rarely to protect the lives of the patient andothers”—Not all staff had this attitude.

Alarm systems (many andwell-trained responses vs. lesswell-trained responses)

More than three night-shift staff members responded to an alarm call, and they were all trained in thistask—Not all such staff were trained, or they were trained but failed to respond.

Communication (all staff havecommunicative skills vs. less thatall)

All staff had general communicative skills or special skills with regard to communicative de-escalation—Lessthan all the staff had these skills.

Well-maintained, clean, and tidysurroundings (always vs. seldom)

The unit was perceived as well maintained, clean, and tidy—The perceptions were otherwise.

Separation of acutely disturbedpatients vs. no separation

The unit was able to separate acutely disturbed patients from other patients—The unit was unable to do so.

MR, mechanical restraint.



than Danish/Norwegian (Knutzen, Sandvik, Hauff,Opjordsmoen, & Friis, 2007; Norredam, Garcia Lopez,Keiding, & Krasnik, 2010), and type of unit (De Benedictiset al., 2011). See the distributions in Table 3.

Other variables (e.g., the diagnostic distribution amongpatients, the unit’s population base, patient flow, age, medi-cation and types of therapy, or other treatment activities)could affect the outcome; however, these variables wereeither not the focus of this study (nonmedical factors,including medication and types of therapy) or “covered” inthe earlier mentioned covariates, as unit type and geo-graphic region acted as proxy covariates (e.g., diagnostic dis-tribution, patient flow, and age were included in unit type;country was included in geographic region; and the popula-tion base was included both in the unit type and the geo-graphic region).

Gender is most likely not associated with MR(Keski-Valkama et al., 2010; Knutzen et al., 2007); thus, wedid not use it as a covariate. The number of patients whoexperienced MR and time spent in MR was closely related tothe number of MR episodes; therefore, these factors were notconfounders.

Data Analyses

The data analyses were conducted in four steps. First, we ana-lyzed how the factors were individually associated with thenumber of MR episodes (Model 1). Second, we analyzed howthe factors were associated with the number of MR episodes,using a model with all factors included (Model 2). Third, toeliminate multicollinearity and the chance that importanteffects may hide each other from Model 2, a third model wasconstructed by a forward factor-selection procedure, wherethe most significant factors were included stepwise until thecorrected quasi-likelihood under independence model crite-rion value began to rise, denoting no further improvementfor the model (Model 3). All these models were additionallyadjusted for the potentially confounding variables. Addition-ally, we adjusted for nesting of units in departments—17departments contained two units each, and three depart-ments contained three units each—by a generalized estimat-ing equations technique. Finally, we analyzed the correlationbetween the factors by Pearson’s correlation coefficient,adjusted for department, to explore multicollinearity andjustify a reduction of Model 2.

Table 3. Potential Confounders

Potential Confounders Missing n

Number of MRs per Unit

Mean Range

Geography Helse Sør-Øst RHF 0 26 26 0–74Helse Vest RHF 6 31 4–64Helse Midt-Norge RHF 10 9 0–22Helse Nord RHF 6 11 2–21Region Hovedstaden 16 28 0–254Region Sjælland 6 39 2–105Region Syddanmark 8 94 38–184Region Midtjylland and Nordjylland 12 40 0–87

Unit size 0–9 beds 0 16 17 0–5010–19 beds 63 37 0–254More than 20 beds 11 26 0–108

Bed occupancya ≤ 100% occupancy rate 1 78 34 0–254> 100% occupancy rate 11 21 3–87

Type of bed Closed 0 61 34 0–254Shielded or integrated 16 42 0–184Open or other 13 16 2–64

Reason for the MRa Self-mutilation or suicidal behaviors 3 87 8 0–150Aggressive behaviors or “acting out” 87 25 0–170

Forensic provision ≤ 20.65% of patients with a forensic provision 3 64 32 0–184> 20.65% of patients with a forensic provision 23 34 2–254

Ethnicity ≤ 10.6% of patients with an ethnic backgroundother than Danish/Norwegian

7 51 32 0–184

> 10.6% of patients with an ethnic backgroundother than Danish/Norwegian

32 34 1–254

Type of unit Acute 0 55 40 0–254Forensic/secure 23 26 0–184Special/rehabilitation/long term 12 12 0–31

Note. aUsed as continuous variable in the further analyses. MR, mechanical restraint. RHF, regionale helseforetak.



In Models 1–3, linear regression was used to assess the asso-ciations between the number of MR episodes and the chosenfactors. To stabilize the variance, the outcome was logarithmi-cally transformed. Hence, the regression coefficients arereported back-transformed (exponentiated; exp(B)) and areto be interpreted multiplicatively. Hence, because the preven-tive factors were dichotomized, a regression coefficientexp(B) indicates the average number of times MR increasedfor units where this factor was present compared to unitswhere this factor was absent. Even though the outcome is acount, for which a Poisson regression may often be used, thelinear regression fit the data better, for example, the quasi-likelihood under independence model criterion value wasmore than twice as high using Poisson regression.

To account for multiple testing, a p value < .01 was chosento indicate significance. All analyses were performed usingIBM SPSS Statistics for Windows, Version 19.0 (Released2010. IBM Corp, Armonk, NY, USA).

Approval and Ethics

Questionnaires such as the one included in the present studydo not require approval from the scientific-ethical committeesystem in Denmark or Norway when the respondents are ableto provide informed consent or when sensitive information isnot collected. The Danish Data Protection Agency (Journalnumber: 2007-2058-0015, PSV-2009-2013) approved thisstudy,and we received permission from the Danish Health andMedicines Authorities and the Danish Psychiatric CentralResearch Register to use the register data on the frequency ofMR in Denmark (Munk-Jørgensen & Mortensen, 1997).

This study followed the Ethical Guidelines for NursingResearch in the Nordic Countries (Northern Nurse Federation,2003) and the Recommendations on Legal Protection of PersonsSuffering From Mental Disorders Especially Those Placed asInvoluntary Patients (Council of Europe, 2000).

Findings

Flow of Participating Study Units

Figure 2 displays a flowchart of the participating units.

Respondents

The responding and nonresponding units were equally dis-tributed across the countries (response rate = 49%). Clinicalnurse managers did not respond for the following reasons, asprovided by managers: “change of clinical nurse manager” (n= 9), “lack of personal resources for accomplishing the task”(n = 7), sick leave (n = 5), inability to remember/identify thenumber of MR episodes (n = 5), lack of belief that this type ofresearch was beneficial (n = 1). The remaining 63 did notprovide a reason.

Number of MR episodes

The respondents reported 2,361 MR episodes in Denmarkand 1,017 MR episodes in Norway during the examined timeperiods. Nine units accounted for 451 of the excluded MRepisodes (>25 per person), leaving 1,938 episodes inDenmark and 989 in Norway for analysis.

Estimating Associations

The outcome variable (the number of MR episodes during1 year per unit) varied between 0 and 254, with a mean valueof 32.5.Alternatively, the number of MR episodes per year perbed varied between 0 and 15.9, with a mean value of 2.5.

In Table 4, Model 1, one MR-preventive factor was signifi-cantly associated with a low frequency of MR episodes: man-datory review (exp(B) = .41, p < .01).

In Model 2, seven MR-preventive factors were significantlyassociated with a low frequency of MR episodes: patientinvolvement (exp(B) = .14, p < .01), staff education (exp(B) =.14, p < .01), mandatory review (exp(B) = .27, p < .01), positivepatient responses to rules (exp(B) = .34, p < .01), staff experi-ence (exp(B) = .40, p < .01), identification of the patient’scrisis triggers (exp(B) = .45, p < .01), and no crowding (exp(B)= .50, p < .01). One MR-preventive factor was significantlyassociated with a high frequency of MR episodes: staff train-ing (exp(B) = 3.60, p < .01).

In Model 3, three MR-preventive factors were significantlyassociated with a low frequency of MR episodes: mandatoryreview (exp(B) = .36, p < .01), patient involvement (exp(B) =.42, p < .01), and no crowding (exp(B) = .54, p < .01).

The Pearson’s correlation matrix showed one highly(Pearson’s r > .50) intercorrelated pair: communication andpositive staff attitudes (Pearson’s r = .77). Furthermore, sixpairs showed medium intercorrelation (Pearson’s r .30 ≤ .50):

Denmark(n = 87)

Excluded: special high-security forensic units

(n = 3)

Norway(n = 96)

Nonresponders/Refused/Missing

data (n = 41)

Included(n = 43)

Included(n = 47)

Nonresponders/Refused/Missing

data (n = 49)

Norway does not have special high-

security units(n = 0)

Participants from Denmark and Norway(n = 90)

Figure 2. Flowchart of the Participating Units



patient-centered care and cognitive milieu therapy (Pearson’sr = .49); communication and staff opposition to MR(Pearson’s r = .38); substitute staff and activity-based nurs-ing (Pearson’s r = .36); communication and risk assessment(Pearson’s r = .32); positive patient responses to rules andidentification of the patient’s crisis triggers (Pearson’s r =.31); and staff education and cognitive milieu therapy(Pearson’s r = .31). All correlations were significant (p < .01).

Patient involvement, mandatory review, and crowdingwere the only potential MR-preventive factors that had a sig-nificant effect (p < .01) in Model 3. The parameter estimates(exp(B) = .36, .42, and .54) are interpreted below.

Mandatory review: Units in which the staff audited,reviewed, or conducted a similar systematic follow-up evalu-ation after all MR episodes (to analyze and learn from the

episodes) experienced an average of .36 times as many (or64% fewer) MR episodes than did those where the evaluationwas conducted after some or none of the MR episodes.

Patient involvement: Units in which the patients partici-pated in ward rounds/conferences, influenced the rulesapplied to the unit, or influenced what was written in thenurse’s journal experienced an average of .42 times as many(or 58% fewer) MR episodes than those in units in which nosuch participation existed.

Crowding: Units in which two of the following three condi-tions were present experienced an average of .54 times (or46% fewer) MR episodes than those in units where only oneor none of the conditions was present: only one bed in apatient’s room, more than 25 m2 of all-day-accessible spaceper patient, and the perception of no crowding.

Table 4. Effects of Potential MR-Preventive Factors on the Annual Number of MR Episodes

MR-Preventive Factors

Model 1 Model 2 Model 3

exp(B)95% CIs ofexp(B) p exp(B)

95% CIs ofexp(B) p exp(B)

95% CIs ofexp(B) p

Cognitive milieu therapy vs. no cognitive milieu therapy 0.75 [0.26, 2.07] .57 0.67 [0.37, 1.18] .17 – – –Patient involvement vs. no patient involvement 0.51 [0.29, 0.88] .02 0.14* [0.07, 0.30] .00 0.42* [0.25, 0.73] .00Patient-centered care vs. no patient-centered care 0.79 [0.30, 2.06] .62 0.42 [0.17, 1.01] .05 – – –Identification of the patient’s crisis triggers vs. no

identification0.73 [0.46, 1.15] .17 0.45* [0.29, 0.72] .00 – – –

Risk assessment vs. no risk assessment 2.01 [0.84, 4.81] .12 2.91 [1.28, 6.60] .01 – – –Mandatory review vs. no mandatory review 0.41* [0.23, 0.73] .00 0.27* [0.18, 0.39] .00 0.36* [0.21, 0.63] .00Staff training (> 50% vs. less) 1.08 [0.69, 1.69] .74 3.60* [2.49, 5.20] .00 – – –Positive staff attitudes vs. less positive 1.15 [0.61, 2.16] .67 1.15 [0.58, 2.26] .69 – – –Patient-staff ratio (≥ 3 staff per patient vs. less) 0.72 [0.42, 1.24] .24 1.36 [0.83, 2.21] .22 – – –Staff education (≥ 3 years vs. less) 1.16 [0.39, 3.41] .79 0.14* [0.06, 0.32] .00 – – –Staff experience (> 5 years vs. less) 0.76 [0.47, 1.23] .27 0.40* [0.24, 0.67] .00 – – –Debriefing, defusing vs. none 1.20 [0.81, 1.70] .36 0.85 [0.59, 1.22] .39 – – –Positive patient responses to rules vs. negative responses 0.82 [0.46, 1.47] .50 0.34* [0.19, 0.62] .00 – – –Activity-based nursing vs. no activity-based nursing 0.93 [0.51, 1.70] .82 1.65 [1.08, 2.51] .02 – – –Acceptable work environment vs. poor work environment 0.54 [0.33, 0.90] .02 1.09 [0.76, 1.58] .64 – – –No crowding vs. crowding 0.56 [0.34, 0.92] .02 0.50* [0.31, 0.80] .00 0.54* [0.36, 0.81] .00Substitute staff (< 50 vs. more) 0.82 [0.51, 1.32] .42 0.57 [0.33, 1.01] .05 – – –Staff opposition to MR (many vs. some) 0.85 [0.54, 1.34] .49 1.12 [0.77, 1.62] .56 – – –Alarm systems (many and well-trained responses vs. less

well-trained responses)1.12 [0.42, 3.00] .82 2.31 [1.15, 4.63] .02 – – –

Communication (all staff have communicative skills vs.less that all)

1.08 [0.66, 1.75] .77 1.53 [0.84, 2.81] .17 – – –

Well-maintained, clean, and tidy surroundings (alwaysvs. seldom)

0.69 [0.41, 1.17] .17 1.08 [0.71, 1.66] .72 – – –

Separation of acutely disturbed patients vs. no separation 1.55 [0.95, 2.52] .08 1.14 [0.73, 1.78] .57 1.65 [1.10, 2.49] .02

Note. The exp(B) parameters were estimated using a linear regression, and all three models were adjusted for the background variables: geography, unitsize, bed occupancy, type of bed, reason for the MR, forensic provision, ethnicity, and type of unit. Additionally, the nesting of units in departmentswas adjusted for by generalized estimating equations techniques. In Model 1, the MR-preventive factors were analyzed alone. In Model 2, all theMR-preventive factors were analyzed together. In Model 3, a forward-selection procedure included the most significant MR-preventive factors until theCorrected quasi-likelihood under independence model criterion value began to rise. In all three models, the outcome variable (i.e., the number of MRepisodes) was log-transformed. Hence, the parameters should be interpreted multiplicatively (e.g., in Model 1, cognitive milieu therapy results in 0.75times the average number of MR incidents without cognitive milieu therapy or 25% fewer MR episodes occur with cognitive milieu therapy thanwithout). *p < .01. MR, mechanical restraint.



Discussion

Mandatory review was significant in all three models; corre-spondingly, no significant intercorrelations were found withthe other factors. Thus, no other factors greatly influenced theassociation between mandatory review and the number ofMR episodes. Previous research also reported that conductingmandatory reviews of MR episodes was associated with a lownumber of MR episodes (Donat, 1998, 2003; McCue, Urcuyo,Lilu, Tobias, & Chambers, 2004; Pollard, Yanasak, Rogers, &Tapp, 2007; Prescott, Madden, Dennis, Tisher, & Wingate,2007). The previous research has suggested that an optimalprocess should review all MR episodes within a few days andthat the review board should consist of an interdisciplinaryteam that includes a nurse specialist, a physician, a psycholo-gist, and the clinical department heads with the supervisoryauthority to change nursing and treatment practices. In addi-tion, the review process should require a protocol thatincludes a description of the episode, the environmental con-ditions that might have provoked the problem behavior, pos-sible “crisis triggers” and any unique calming techniquesidentified, the ways in which the patient is involved, theirmedications, any behavioral treatments and milieu therapies,and other items.

Patient involvement was significant in both Model 2 andModel 3; because patient involvement was not significant inModel 1, minor intercorrelations did occur (albeit notstrongly or significantly) with, for example, staff training andwell-maintained, clean, and tidy surroundings. Potentially,elements of patient involvement are included in some stafftraining, and patients who are more involved in their owncare may have a greater sense of responsibility toward theirsurroundings. Thus, the nonsignificant and highly complexstatistics in the correlation matrix with regard to patientinvolvement did not precisely explain which factors influ-enced the association between patient involvement and thenumber of MR episodes. Previous research also reported thatpatient involvement might reduce MR episodes (Borckardtet al., 2011; Thyrsting & Hall, 2008). In our study, patientinvolvement was measured using three items (whether thepatients participated in ward rounds/conferences, as well aswhether they influenced the rules and what was written innurses’ journals). Of course, patient involvement is a muchmore complex phenomenon that includes the concept thatthe patient is the expert on his or her life, and the staff is expertwith regard to their care or treatment; actions are negotiatedvia dialog. It is, however, reasonable to believe that units witha recovery-orientated or an empowering approach, that inte-grate patient involvement in the delivery of care and treat-ment, will include one of the aforementioned items (whetherthe patients participated in ward rounds/conferences, as wellas whether they influenced the rules and what was written innurses’ journals).

No crowding was significant in both Model 2 and Model 3;one minor but significant intercorrelation between no crowd-ing and patient/staff ratio existed (Pearson’s r, .24, p < .05).This intercorrelation is understandable, where more staffcreates a perception of more crowding. Previous research alsoreported that no crowding might influence the number ofMR episodes (Meehan, McIntosh, & Bergen, 2006; Vaaler,Morken, Fløvig, Iversen, & Linaker, 2006; van der Schaaf,Dusseldorp, Keuning, Janssen, & Noorthoorn, 2013;Virtanen et al., 2011).

Several MR-preventive factors had high parameter esti-mates that indicated an increase in the number of MR epi-sodes, although none was significant in Model 3. One of themost interesting was risk assessment; strong evidence sup-ported the role of this preventive factor in earlier studies,including at least two cluster randomized controlled trials(RCTs). One study showed a reduction in coercive measures(Abderhalden et al., 2008), and the other showed a reductionin aggressive incidents (van de Sande et al., 2011). The reasonwe did not find a significant relationship between risk assess-ment and MR reduction in our study might be that we com-bined the risk assessments for self-destructive/suicidalbehaviors and aggressive/violent behaviors. The difference inthe risk assessments of self-destructive behaviors is mostlikely impossible to measure because both Denmark andNorway have focused on this factor over the past several years;thus, almost all institutions have implemented this system ofassessment. Assessment of aggressive/violent behavior hasalso become more common. Fifty-five percent of the patientswere systematically assessed for the risk of developing aggres-sive or violent behaviors in the units; however, this assessmentsystem appears to have been introduced initially in units withhigh numbers of MR episodes. Therefore, risk assessmentmust be investigated by another means (e.g., single risk-assessment methods using a cluster-RCT design). The samepattern could be seen with staff training, which was the onlysignificant factor with high parameter estimates, though onlyin Model 2. Training in de-escalation, avoiding physicalattacks, careful restraining techniques, etc., were likely offeredfirst to staff in units with high numbers of MR episodes.

Overall, because the quality of the supporting researchdoes not establish a causal relationship between the numberof MR episodes and mandatory review, patient involvement,or no crowding, the findings of this study cannot be seen as arejection of hypotheses in practice; rather, they support andprovide a better understanding of the mechanisms involvedin MR prevention.

Limitations

Several limitations must be considered when examining theresults. First, some confounding covariates might be missingfrom the analyses. This basic limitation is embedded in the



study design, given that there was no controlled laboratoryenvironment and that it was not possible to control all poten-tial covariates (e.g., patient flow, diagnostic distribution, theunits’ different population bases, and age of the patients whoexperienced MR); however, either these covariates are notassociated with the number of MR episodes, or their effectdisappeared after geographic region and type of unit wereadded to the analyses of the data from the Danish NationalRegister on coercion, where they acted as proxy covariates.The two datasets (the questionnaire and the Danish NationalRegister on coercion) were obtained from different organiza-tional levels (units/wards) in the same year, which leads us tobelieve that the same effect might be found in our question-naire data (the covariates without confounding); however, wedid not have access to these data from the Norwegian sample;therefore, we do not know whether these conclusions apply tothe entire dataset.

Second, recall bias is possible, especially for the Danishdata; the clinical nurse manager had to recall all MR-preventive factors and the conditions and number of MR epi-sodes from 2 years prior. The outcome variable (i.e., thenumber of MR episodes) and the number of patients whoexperienced MR more than 25 times were validated or reex-amined (see the Data Collection section). This analysis didnot reveal significant recall bias with regard to the outcomevariable. Nurses’ recall is reliable over 4 weeks (Gerolamo,2008). In this study, the clinical nurse managers were able tosupport their memories by checking charts, protocols, etc., toanswer many of the questions. Additionally, recall bias couldaffect when the MR-preventive factors were implemented, forexample, if patient involvement only had been implementedfor the last 3 months and not over the whole year under inves-tigation. Thus, this factor would probably only increase theeffect on patient involvement. Nevertheless, we must considerthe possibility of recall bias with regard to other variables.

Third, three participants answered the questionnairestwice over a 3-month interval to check the test–retest reliabil-ity (see the Data Collection section). Although three partici-pants are too few to properly calculate reliability, the resultsindicate no substantial difference in the measured constructsbetween the two administrations.

Fourth, the nonresponders’ reasons for failure to replywere reviewed, and no systematic differences were foundbetween those who responded and those who did not.Reviewing the Danish National Register on coercion, wefound for the same population that 49% of responding unitswere “responsible” for 52% of the MR episodes, suggestingvery small differences in MR episodes between respondersand nonresponders; however, no proper dropout analysis wasconducted. Consequently, we cannot establish whether thehigh dropout rate created systematic flaws, but we believe thatthe most important preventive factors and background vari-ables were included in the study. Thus, the regression coeffi-

cients most likely can be generalized to the populations ofboth countries. Transferring the results to other countries,however, would require a thorough examination of their cul-tures, legislations, and other relevant factors.

Finally, when the data were reanalyzed including patientswho experienced MR more than 25 times, the result differedonly marginally. Some of the parameter estimates wereslightly higher and some slightly lower, and in Model 1,patient involvement became significant, but the effect inModel 3 was unchanged. Overall, exclusion of the patientswho experienced MR more than 25 times did not affectresults.

Implications for Nursing Practice and Research

Because no causal effect was established between the reduc-tion in MR episodes and mandatory review, patient involve-ment, or no crowding, we cannot claim that implementationof these potential preventive measures would reduce thenumber of MR episodes. More studies are needed to investi-gate effective procedures to reduce the number of MRepisodes; however, because mandatory review, patientinvolvement, and no crowding did not have any adverseeffects, they are recommended because they might contributepositively to existing MR-preventive management programs.

Acknowledgments

We thank the Health Science Research Foundation in theCapital Region of Denmark, the Mental Health Services in theCapital Region of Denmark, and the Mental Health CentreSankt Hans, Copenhagen University Hospital for funding thisproject.

We thank all of the clinical nurse managers of the “closed”psychiatric units in Norway and Denmark for participating inthis project.

The authors also thank Volkert Dirk Siersma, Statistician,PhD, the Research Unit for General Practice, Centre forHealth and Society, Copenhagen University, and the Depart-ment of Biostatistics at Copenhagen University for supervis-ing the data analyses.

References

Abderhalden, C., Needham, I., Dassen, T., Halfens, R., Haug, H. J.,& Fischer, J. E. (2008). Structured risk assessment and violencein acute psychiatric wards: Randomised controlled trial. BritishJournal of Psychiatry: The Journal of Mental Science, 193(1),44–50. doi:10.1192/bjp.bp.107.045534

Bak, J., & Aggernæs, H. (2012). Coercion within Danishpsychiatry compared with 10 other European countries.Nordic Journal of Psychiatry, 66(5), 297–302.doi:10.3109/08039488.2011.632645



Bak, J., Brandt-Christensen, M., Sestoft, D. M., & Zoffmann, V.(2012). Mechanical restraint—Which interventions preventepisodes of mechanical restraint?—A systematic review.Perspectives in Psychiatric Care, 48, 83–94.doi:10.1111/j.1744-6163.2011.00307.x

Bonner, G., Lowe, T., Rawcliffe, D., & Wellman, N. (2002). Traumafor all: A pilot study of the subjective experience of physicalrestraint for mental health inpatients and staff in the UK.Journal of Psychiatric and Mental Health Nursing, 9(4),465–473. doi:10.1046/j.1365-2850.2002.00504.x

Borckardt, J. J., Madan, A., Grubaugh, A. L., Danielson, C. K.,Pelic, C. G., Hardesty, S. J., . . . Fruen, B. C. (2011). Systematicinvestigation of initiatives to reduce seclusion and restraint in astate psychiatric hospital. Psychiatric Services (Washington,D.C.), 62(5), 477–483.

Bower, F. L., McCullough, C. S., & Timmons, M. E. (2003). Asynthesis of what we know about the use of physical restraintsand seclusion with patients in psychiatric and acute caresettings: 2003 update. Online Journal of Knowledge Synthesis forNursing, 10(1), 1–29. doi:10.1111/j.1524-475X.2003.00001.x

Bowers, L., Van Der Werf, B., Vokkolainen, A., Muir-Cochrane, E.,Allan, T., & Alexander, J. (2007). International variation incontainment measures for disturbed psychiatric inpatients: Acomparative questionnaire survey. International Journal ofNursing Studies, 44(3), 357–364.doi:10.1016/j.ijnurstu.2006.01.005

Bremmes, R., Hatling, T., & Bjørngaard, J. H. (2008). Bruk avtvangsmidler i psykisk helsevern i 2001, 2003, 2005 og 2007.Trondheim: SINTEF Helse.

Council of Europe. (2000). White Paper Regarding a DraftRecommendation on Legal Protection of Persons Suffering fromMental Disorder Especially Those Placed as Involuntary Patients.Strasbourg: Author.

Cusack, K. J., Frueh, B. C., Hiers, T., Suffoletta-Maierle, S., &Bennett, S. (2003). Trauma within the psychiatric setting: Apreliminary empirical report. Administration and Policy inMental Health, 30(5), 453–460. doi:10.1023/A:1024697715409

De Benedictis, L., Dumais, A., Sieu, N., Mailhot, M. P., Tran, M. A.M., Stikarovska, I., . . . Lesage, A. D. (2011). Staff perceptionsand organizational factors as predictors of seclusion andrestraint on psychiatric wards. Psychiatric Services (Washington,D.C.), 62(5), 484–491. doi:10.1176/appi.ps.62.5.48

Donat, D. C. (1998). Impact of a mandatory behavioralconsultation on seclusion/restraint utilization in a psychiatrichospital. Journal of Behavior Therapy and ExperimentalPsychiatry, 29(1), 13–19. doi:10.1016/S0005-7916(97)00039-6

Donat, D. C. (2003). An analysis of successful efforts to reduce theuse of seclusion and restraint at a public psychiatric hospital.Psychiatric Services (Washington, D.C.), 54, 1119–1123.doi:10.1176/appi.ps.54.8.1119

Frueh, B. C., Knapp, R. G., Cusack, K. J., Grubaugh, A. L.,Sauvageot, J. A., Cousins, V. C., . . . Hiers, T. G. (2005). Specialsection on seclusion and restraint: Patients’ reports oftraumatic or harmful experiences within the psychiatric

setting. Psychiatric Services (Washington, D.C.), 56(9),1123–1133. doi:10.1176/appi.ps.56.9.112

Gerolamo, A. M. (2008). Measuring adverse outcomes ininpatient psychiatry: The reliability of nurse recall.Archives of Psychiatric Nursing, 22(2), 95–103.doi:10.1016/j.apnu.2007.03.004

Hem, E., Steen, O., & Opjordsmoen, S. (2001). Thrombosisassociated with physical restraints. Acta PsychiatricaScandinavica, 103(1), 73–76.doi:10.1111/j.1600-0447.2001.00082.x

Hick, J. L., Smith, S. W., & Lynch, M. T. (1999). Metabolicacidosis in restraint associated cardiac arrest: a case series.Academic Emergency Medicine, 6(3), 239–243.doi:10.1111/j.1553-2712.1999.tb00164.x

Johnson, M. E. (2010). Violence and restraint reduction efforts oninpatient psychiatric units. Issues in Mental Health Nursing,31(3), 181–197. doi:10.3109/01612840903276704

Keski-Valkama, A., Sailas, E., Eronen, M., Koivisto, A. M.,Lönnqvist, J., & Kaltiala-Heino, R. (2007). A 15-year nationalfollow-up: Legislation is not enough to reduce the use ofseclusion and restraint. Social Psychiatry and PsychiatricEpidemiology, 42(9), 747–752. doi:10.1007/s00127-007-0219-7

Keski-Valkama, A., Sailas, E., Eronen, M., Koivisto, A. M.,Lönnqvist, J., & Kaltiala-Heino, R. (2010). Who are therestrained and secluded patients: A 15-year nationwide study.Social Psychiatry and Psychiatric Epidemiology, 45(11),1087–1093. doi:10.1007/s00127-009-0150-1

Knutzen, M., Sandvik, L., Hauff, E., Opjordsmoen, S., & Friis, S.(2007). Association between patients’ gender, age andimmigrant background and use of restraint-A 2-yearretrospective study at a department of emergency psychiatry.Nordic Journal of Psychiatry, 61(3), 201–206.doi:10.1080/08039480701352520

Laursen, S. B., Jensen, T. N., Bolwig, T., & Olsen, N. V. (2005).Deep venous thrombosis and pulmonary embolism followingphysical restraint. Acta Psychiatrica Scandinavica, 111(4),324–327. doi:10.1111/j.1600-0447.2004.00456.x

Lov nr. 62-§ 4-8 (1999). Lov om etablering og gjennomføring avpsykiak helsevern.

McCue, R. E., Urcuyo, L., Lilu, Y., Tobias, T., & Chambers, M. J.(2004). Reducing restraint use in a public psychiatric inpatientservice. Journal of Behavioral Health Services and Research,31(2), 217–224.

Meehan, T., McIntosh, W., & Bergen, H. (2006). Aggressivebehaviour in the high secure forensic setting: The perceptionsof patients. Journal of Psychiatric and Mental Health Nursing,13(1), 19–25. doi:10.1111/j.1365-2850.2006.00906.x

Morrison, A., & Sadler, D. (2001). Death of a psychiatric patientduring physical restraint. Excited delirium—A case report.Medicine, Science, and the Law, 41(1), 46–50.

Munk-Jørgensen, P., & Mortensen, P. B. (1997). The DanishPsychiatric Central Register. Danish Medical Bulletin, 44(1), 82.

The National Collaborating Centre for Nursing and SupportiveCare. (2005). Violence, the short-term management of



disturbed/violent behaviour in in-patient psychiatric settings andemergency departments (Clinical Guideline 25). London: TheNational Institute for Health and Clinical Excellence.

National Risk Management Programme. (2007). Best practice inmanaging risk. London: Department of Health.

Nielsen, A. S. (2005). Deep venous thrombosis and fatalpulmonary embolism in a physically restrained patient.Ugeskrift for laeger, 167(21), 2294.

Norredam, M., Garcia Lopez, A., Keiding, N., & Krasnik, A.(2010). Excess use of coercive measures in psychiatryamong migrants compared with native Danes.Acta Psychiatrica Scandinavica, 121(2), 143–151.doi:10.1111/j.1600-0447.2009.01418.x

Northern Nurse Federation. (2003). Ethical guidelines for nursingresearch in the Nordic countries. Vård i Norden, 23(4), 1–19.

Olsen, H. (2006). Guide til gode spørgeskemaer. København:Socialforskningsinstituttet.

Pollard, R., Yanasak, E. V., Rogers, S. A., & Tapp, A. (2007).Organizational and unit factors contributing to reduction inthe use of seclusion and restraint procedures on an acutepsychiatric inpatient unit. Psychiatric Quarterly, 78(1), 73–81.doi:10.1007/s11126-006-9028-5

Prescott, D. L., Madden, L. M., Dennis, M., Tisher, P., & Wingate,C. (2007). Reducing mechanical restraints in acute psychiatriccare settings using rapid response teams. Journal of BehavioralHealth Services and Research, 34(1), 96–105.

Psykiatriloven, LBK nr. 1729 (2010). Bekendtgørelse af lov omanvendelse af tvang i psykiatrien.

Scanlan, J. N. (2010). Interventions to reduce the use of seclusionand restraint in inpatient psychiatric settings: What we know sofar a review of the literature. International Journal of SocialPsychiatry, 56(4), 412–423. doi:10.1177/0020764009106630

Schaeffer, N. C., & Presser, S. (2003). The science of askingquestions. Annual Review of Sociology, 29, 65–89.doi:10.1146/annurev.soc.29.110702.110112

Steinert, T., Bergbauer, G., Schmid, P., & Gebhardt, R. P. (2007).Seclusion and restraint in patients with schizophrenia: Clinicaland biographical correlates. Journal of Nervous and MentalDisease, 195(6), 492–496. doi:10.1097/NMD.0b013e3180302af6

Steinert, T., Lepping, P., Bernhardsgrütter, R., Conca, A., Hatling,T., Janssen, W., . . . Whittington, R. (2010). Incidence ofseclusion and restraint in psychiatric hospitals: A literaturereview and survey of international trends. SocialPsychiatry and Psychiatric Epidemiology, 45, 889–897.doi:10.1007/s00127-009-0132-3. Retrieved from Springer.

Stewart, D., Van der Merve, M., Bowers, L., Simpson, A., & Jones,J. (2010). A review of interventions to reduce mechanicalrestraint and seclusion among adult psychiatric inpatients.Issues in Mental Health Nursing, 31(6), 413–424.doi:10.3109/01612840903484113

Strout, T. D. (2010). Perspectives on the experience of beingphysically restrained: An integrative review of the qualitativeliterature. International Journal of Mental Health Nursing,19(6), 416–427. doi:10.1111/j.1447-0349.2010.00694.x

Sundhedsstyrelsen. (2008). Nye tal fra sundhedsstyrelsen.Anvendelse af tvang i psykiatrien 2007. København: Author.

Sundhedsstyrelsen. (2009). Tvangsforanstaltninger ipsykiatrien—en udredning. København: Author.

Thyrsting, K., & Hall, E. O. C. (2008). Prevention of coercion inpsychiatric units: A secondary analysis of a practitionerresearch project. Klinisk Sygepleje, 22(2), 78–86.

Vaaler, A. E., Morken, G., Fløvig, J. C., Iversen, V. C., & Linaker, O.M. (2006). Effects of a psychiatric intensive care unit in anacute psychiatric department. Nordic Journal of Psychiatry,60(2), 144–149. doi:10.1080/08039480600583472

van de Sande, R., Nijman, H. L. I., Noorthoorn, E. O., Wierdsma,A. I., Hellendoorn, E., van der Staak, C., & Mulder, C. L. (2011).Aggression and seclusion on acute psychiatric wards: Effect ofshort-term risk assessment. British Journal of Psychiatry:The Journal of Mental Science, 199(6), 473–478.doi:10.1192/bjp.bp.111.09514

van der Schaaf, P. S., Dusseldorp, E., Keuning, F. M., Janssen, W.A., & Noorthoorn, E. O. (2013). Impact of the physicalenvironment of psychiatric wards on the use of seclusion.British Journal of Psychiatry: The Journal of Mental Science,202(2), 142–149. doi:10.1192/bjp.202.2.162

Virtanen, M., Vahtera, J., Batty, G. D., Tuisku, K., Pentti, J.,Oksanen, T., . . . Kivimäki, M. (2011). Overcrowding inpsychiatric wards and physical assaults on staff: Data-linkedlongitudinal study. British Journal of Psychiatry: The Journal ofMental Science, 198(2), 149–155. doi:10.1192/bjp.bp.110.08238

Watt, T., Rasmussen, Å., Groenvold, M., Bjorner, J. B., Watt, S. H.,Bonnema, S. J., & Feldt-Rasmussen, U. (2008). Improving anewly developed patient-reported outcome for thyroidpatients, using cognitive interviewing. Quality of Life Research,17(7), 1009–1017. doi:10.1007/s11136-008-9364-z

Willis, G. B. (2005). Cognitive interviewing. A tool for improvingquestionnaire design. London: Sage.

Willis, G. B., & Lessler, J. T. (1999). Question appraisal system(QAS-99). Atlanta: Research Triangle Institute.



Paper 4

Comparing the effect of non-medical mechanical restraint preventive factors

between psychiatric units in Denmark and Norway.

Comparing the effect of non-medical mechanical restraint preventive factors between psychiatric units in Denmark and Norway JESPER BAK , VIBEKE ZOFFMANN , DORTE MARIA SESTOFT , ROGER ALMVIK , VOLKERT DIRK SIERSMA , METTE BRANDT-CHRISTENSEN

Bak J, Zoffmann V, Sestoft DM, Almvik R, Siersma VD, Brandt-Christensen M. Comparing the effect of non-medical mechanical restraint preventive factors between psychiatric units in Denmark and Norway. Nord J Psychiatry 2015;68:1–11.

Background: The use of mechanical restraint (MR) is controversial, and large differences regarding the use of MR are often found among countries. In an earlier study, we observed that MR was used twice as frequently in Denmark than Norway. Aims: To examine how presumed MR preventive factors of non-medical origin may explain the differing number of MR episodes between Denmark and Norway. Methods: This study is a cross-sectional survey of psychiatric units. Linear regression was used to assess the confounding effects of the MR preventive factors, i.e. whether a difference in the impact of these factors is evident between Denmark and Norway. Results: Six MR preventive factors confounded [ Δ exp( B ) � 10%] the difference in MR use between Denmark and Norway, including staff education ( � 51%), substitute staff ( � 17%), acceptable work environment ( � 15%), separation of acutely disturbed patients (13%), patient – staff ratio ( � 11%), and the identifi cation of the patient ’ s crisis triggers ( � 10%). Conclusions: These six MR preventive factors might partially explain the difference in the frequency of MR episodes observed in the two countries, i.e. higher numbers in Denmark than Norway. One MR preventive factor was not supported by earlier research, the identifi cation of the patient ’ s crisis triggers; therefore, more research on the mechanisms involved is needed. Clinical implications: None of the six MR preventive factors presents any adverse effects; therefore, units in Denmark and Norway may consider investigating the effect of implementing, the identifi cation of the patient ’ s crisis triggers, an increased number of staff per patient, increased staff education, a better work environment and reduced use of substitute staff in practice.

• Coercion , Mechanical restraint , Mental health , Nursing , Psychiatry.

Jesper Bak, Research Institute, Mental Health Centre Sct. Hans, Boserupvej 2, 4000 Roskilde, Denmark, E-mail: [email protected]; Accepted 4 December 2014.

In Denmark and Norway, mechanical restraint (MR) can be legally performed when psychiatric inpatients

pose a risk to themselves or others (1, 2). The procedure is primarily performed by nursing staff, and involves the use of leather or fi bre belts to secure the patient to a bed. MR exposes patients to potential physical risks (3 – 7) and has negative psychological effects for many patients (8 – 12).

A Danish review (13) identifi ed numerous MR pre-ventive factors based on studies that target MR preven-tion. No MR preventive factors found in the review received the highest scientifi c recommendation grade. 1 Furthermore, several of these MR preventive factors

might be effective only in controlled research environ-ments.

In an earlier retrospective association study (14), we found three MR preventive factors associated with low numbers of MR episodes in Denmark and Norway (man-datory review, patient involvement and no crowding). In that study, we eliminated the differences between the coun-tries by making adjustments to make the fi ndings more universal. However, focusing on the differences between the countries is important when the aim is to develop effective MR reduction programs in Denmark and Norway.

In a comparative study on the use of coercion in 11 European countries, including Denmark and Norway

© 2015 Informa Healthcare DOI: 10.3109/08039488.2014.996600

Nor

d J

Psyc

hiat

ry D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y R

igsh

ospi

tale

t on

01/2

3/15

For

pers

onal

use

onl

y.

JBAK0003

Tekstboks

J BAK ET AL.

2 NORD J PSYCHIATRY·EARLY ONLINE·2015

access to the requested information. The units were the smallest organizational collection of beds led by a clini-cal nurse manager.

Time frame The Danish questionnaire was administered between 23 December 2009 and 28 May 2010, and the collected data pertained to 2008. The Norwegian questionnaire was administered between 29 October 2010 and 27 January 2011, and the collected data pertained to the previous 12 months, most of which occurred in 2010.

Included participants The study included all psychiatric hospital units in Den-mark ( n � 87) and Norway ( n � 96) that treated adult inpatients (18 – 65 years) and used MR for approximately 2 years prior to the year of investigation. MR did occur in Norway in settings other than hospitals; however, given that only 1.5% of recorded MR episodes in 2007 occurred in these settings (16) and that hospitals served as the pri-mary focus of this study, these units were not included.

Excluded participants The special high-security department in Denmark was excluded due to specifi c legislation concerning seclusion.

Excluded outcome data If MR was used more than 25 times on the same patient during the investigation period, then the episodes of MR � 25 were excluded. Chance determined the units to which these patients were admitted, but the presence of these patients could signifi cantly affect the units ’ MR frequencies, distorting the results.

Data collection: the questionnaire A questionnaire was developed based on the results from an earlier systematic literature review that focused on factors that might reduce the use of MR (13).

Each MR preventive factor was “ translated ” into ques-tions to identify the degree to which the factor was pres-ent in the units.

Six factors were not included in the present study. Combined intervention programmes, were omitted because it were too complex to measure. The following four factors were omitted because they were not consid-ered measured appropriately by the questionnaire: treat-ment of substance abuse and psychosis, patients perceived as a resource in their own treatment, pro necessitate [p.n.]/as-needed medication guided by individual patients, and care alliance. Lastly, music used as a diversion and for relaxation were omitted, because it were partially cov-ered by another factor, activity-based nursing.

Cognitive milieu therapy and patient involvement and patient-centred care and patient involvement, were sepa-

(15), Norway had the lowest rate of physical coercion (MR, seclusion and physical restraint/holding) with 29 inhabitants per 100,000 experiencing physical coercion. In comparison, Denmark had 122 episodes of MR per 100,000 inhabitants, and Norway had 87 episodes of MR per 100,000 inhabitants in 2007 (16, 17). The same ten-dency was observed in the subpopulation we investigated earlier (14); Denmark had 72 MR episodes per 100,000 inhabitants, and Norway had 42 MR episodes per 100,000 inhabitants. The lower frequency of MR epi-sodes in Norway might indicate that more effective MR preventive factors have been implemented in Norway compared with Denmark or that MR preventive factors exhibit differing effects between the countries.

Aim To examine how presumed MR preventive factors of non-medical origin may explain the differing number of MR episodes between Denmark and Norway.

Material and methods A comparative study was conducted to investigate the degree to which the non-medical MR preventive factors identifi ed in a previous review (13) were implemented in the psychiatric units and whether the chosen MR preven-tive factors association with the number of MR episodes differed between Denmark and Norway.

The following conditions must be present legally to initiate MR according to the Danish Mental Health Act (translated by the authors) (1):


1) exposing their body or health or the body or health of others to danger,

2) pursuing or in any other way grossly molesting fel-low patients, or

3) committing signifi cant acts of vandalism.

These conditions are consistent with Norwegian legis-lation (translated by the authors) (2):

Restraints may be used on a patient only when absolutely necessary to prevent him or her from damaging him or herself or others or to prevent damage to buildings, cloth-ing, inventory or other things.

This paper uses Bowers et al. ’ s (18) defi nition of MR: The use of restraining straps, belts or other equipment to restrict movement . Note that this defi nition refers only to the restraint of inpatients in psychiatric hospitals.

The MR preventive factors identifi ed by the literature review (13) were included in an online questionnaire and sent to the clinical nurse managers for the purpose of quantifying whether they were implemented in the units. These managers were chosen because they had exclusive

Nor

d J

Psyc

hiat

ry D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y R

igsh

ospi

tale

t on

01/2

3/15

For

pers

onal

use

onl

y.

JBAK0003

Tekstboks

PREVENTING MECHANICAL RESTRAINTS IN PSYCHIATRY

NORD J PSYCHIATRY·EARLY ONLINE·2015 3

Later, the outcome variable (i.e. the number of MR episodes) was re-examined. This procedure was neces-sary due to the confusing wording of a heading in the Danish questionnaire and the lack of precise numbers regarding the Norwegian patients who experienced MR more than 25 times (14 answer categories instead of the actual number were provided, which made the exclusion imprecise). The re-examination was accom-plished by contacting all of the clinical nurse managers in Denmark and four of the clinical nurse managers in Norway. The respondent rate was 94%, and the results lowered the overall frequency of MR by 3.2%. A high correlation coeffi cient ( t Ken,b � 0.89) was found between the two answers, revealing that the confusing headline and the categorization error did not have an important effect.

A description of how the 22 dichotomized MR pre-ventive factors were measured in the Supplementary Appendix to be found online at http://informa healthcare.com/doi/abs/10.3109/08039488.2014.996600.

Background variables Several conditions might have affected the outcome. These conditions included unit size, bed occupancy, type of bed, reason for the MR, forensic provision, ethnicity other than Danish/Norwegian and type of unit. A descrip-tion of the background variables included in the study is presented in Table 1.

Other variables, e.g. the unit ’ s population base, the country ’ s inhabitants and patient fl ow, could also affect the outcome; however, these variables were covered by the previously mentioned covariates. Unit type and unit size acted as proxy covariates [e.g. patient fl ow and pop-ulation base were included in unit type; inhabitants were included in unit size, where the number of beds per 100,000 inhabitants in each country was very similar (Denmark � 23.96 and Norway � 22.81 beds per 100,000 inhabitants)].

A patient ’ s gender is most likely not associated with MR (24, 25); thus, we did not use gender as a covariate.

Data analyses The data analyses were conducted in fi ve steps. First, we analysed the difference in the frequency of MR episodes between the countries. Second, we analysed how the fac-tors were individually associated with the number of MR episodes in Denmark and Norway separately. Third, we determined whether an interaction between the MR pre-ventive factors and the country (second order interaction, SOI) existed. Fourth, we analysed the difference in the prevalence of the MR predictive factors between the two countries using Pearson ’ s chi-squared test ( χ 2 ). Fifth, we analysed the confounding effects of the MR preventive factors, i.e. the difference in the effect of the country in

rated into three factors: cognitive milieu therapy, patient-centred care, and patient involvement. Although patient involvement is part of cognitive milieu therapy and patient-centred care, it might also occur by itself. Leav-ing, 22 separate MR preventive factors to be included in the present study.

The questionnaire was developed by conceptualizing and operationalizing the MR preventive factors. To increase validity and reliability, we followed the Ques-tion Appraisal System [QAS-99 (19)] that was developed to identify problems in questionnaires, including the cognitive processes inherent to the question-answering process.

The construction and pre-testing of the questionnaire involved an eight-step appraisal of each item (QAS-99): reading, instructions, clarity, assumption, knowledge/mem-ory, sensitivity/bias, response categories and others. 2 The pre-test was conducted by creating a cognitive-test proto-col that included probes for each item. The probes cov-ered the aforementioned eight steps and consisted of several types: general (e.g. How did you reach this answer?); specifi c (e.g. Why do you believe MR is an intrusive and coercive measure?); understanding/interpreta-tion (e.g. How do you defi ne physical activity?); safety (e.g. How sure are you of your answer?); paraphrasing (e.g. Can you repeat the question in your own words?); and memory (e.g. How did you know that you had 56 MR last year?).

The pre-test consisted of cognitive interviews (20) of four clinical nurse managers individually using concur-rent probing (21), in which verbal probes were asked immediately after a question. This protocol led to the revision of seven questions that were subsequently pre-tested in a group interview.

Data collection, administration, and processing, as well as error identifi cation and the creation of electronic data fi les (22, 23), were administered using the electronic web-based survey system Enalyzer (2010).

The questionnaire was checked for test – retest reliabil-ity by asking three clinical nurse managers to answer the questionnaire a second time after a 3-month interval (The respondents were instructed not to check their previous answers). The estimated mean correlation between all of the fi rst and second answers was reasonable (Kendall ’ s Tau b ( t Ken,b ) � 0.75). With regard to the Norwegian por-tion of the study, the questionnaire was translated and retranslated by JB and then checked by two healthcare professionals who were familiar with the subject. A pre-test was conducted to test face validity through cognitive interviews of seven clinical nurse managers. This test led to the revision of eight questions, mostly due to impre-cise translations. Moreover, one question was added to the questionnaire due to different staff competences: psy-chologists in Norway are partially responsible for their patients ’ treatment, which is not the case in Denmark.

Nor

d J

Psyc

hiat

ry D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y R

igsh

ospi

tale

t on

01/2

3/15

For

pers

onal

use

onl

y.

JBAK0003

Tekstboks

J BAK ET AL.


coeffi cient exp( B ), indicates the difference between the countries in an average number of MR episodes.

All models except the fourth set of analyses ( χ 2 -test) were additionally adjusted for background variables. Additionally, we adjusted for nesting of units in depart-ments (17 departments contained two units each and three departments contained three units each), using Gen-eralized Estimating Equations (GEE).

To account for multiple testing, a P -value � 0.01 was chosen to indicate signifi cance. All analyses were per-formed using IBM SPSS Statistics for Windows, Version 19.0 (Released 2010; IBM Corp., Armonk, NY).

Approval and ethics Questionnaires, such as the one included in the present study, do not require approval from the scientifi c ethical committee system in Denmark or Norway when the respondents are able to provide informed consent or when sensitive information is not collected. The Danish Data Protection Agency (Journal number: 2007-58-0015, PSV-

models with and without the corresponding MR preven-tive factor expressed as a percentage [ Δ exp( B )]; a per-centage of greater than 10% was chosen to denote a difference.

Linear regression was used to estimate the associa-tions among the frequency of MR episodes, the coun-tries, and the MR preventive factors. To stabilize the variance, the outcome was logarithmically transformed. Hence, the regression coeffi cients are back-transformed [exponentiated; exp( B )] and should be interpreted multi-plicatively. Because the preventive factors were dichoto-mized, a regression coeffi cient exp( B ) indicates the average number of times MR increased for units where a factor was present compared with units where this factor was absent (e.g. if the result was exp( B ) � 0.42, it should be interpreted as 0.42 times the average number of MR incidents without the MR preventive factor or 58% fewer MR episodes occur with the MR preventive factor than without). Except in the fi rst analysis, where the dichoto-mized variable was “ country ” , here the regression

Table 1 . Background variables.

Background variables Missing

Number of MR episodes per

unit in Denmark

Number of MR episodes per

unit in Norway

χ 2 : Difference in the prevalence of the background

variable between countries

n Mean n Mean P

Unit size0 – 9 beds 0 2 16 14 17 0.0010 – 19 beds 33 51 30 23More than 20 beds 8 28 3 21

Bed occupancy * � 100% occupancy rate 1 35 50 43 21 0.08 � 100% occupancy rate 8 22 3 20

Type of bedClosed 0 6 15 7 18 0.54Shielded or integrated 12 50 4 15Open or other 25 50 36 22

Reason for the MR * Self-mutilation or suicidal behaviours 2 41 10 47 4 0.48Aggressive behaviours or “ acting out ” 2 41 34 47 18 0.46

Forensic provision � 20.65% of patients with a forensic provision 3 25 45 39 23 0.03 � 20.65% of patients with a forensic provision 15 47 8 10

Ethnicity � 10.6% of patients with an ethnic background other than Danish/

Norwegian7 24 48 27 18 0.79

� 10.6% of patients with an ethnic background other than Danish/Norwegian

16 42 16 25

Type of unitAcute 0 6 15 7 18 0.05Forensic/secure 12 50 4 15Special/rehabilitation/long-term 25 50 36 22

MR, mechanical restraint. Missing, number of missing data from units within each background variable; n , number of units in each category. * Used as a continuous variable in further analyses. χ 2 , P -value of a chi square test for the difference in the prevalence of the background variables in the units between the countries.

Nor

d J

Psyc

hiat

ry D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y R

igsh

ospi

tale

t on

01/2

3/15

For

pers

onal

use

onl

y.

JBAK0003

Tekstboks



in Denmark. In Norway, three MR preventive factors were signifi cantly associated with a higher frequency of MR episodes: cognitive milieu therapy [exp( B ) � 7.46, P � 0.00], patient-centred care [exp( B ) � 5.01, P � 0.00] and alarm systems [exp( B ) � 3.72, P � 0.00]. Notably, for cognitive milieu therapy and patient-centred care, the confi dence intervals were large, which makes the inter-pretation less precise.

We identifi ed three incidents of interaction among the MR preventive factor, the countries and the frequency of MR episodes: cognitive milieu therapy (SOI, P � 0.00), patient-centred care (SOI, P � 0.01) and alarm systems (SOI, P � 0.00). Each of these factors exhibited a posi-tive effect in Denmark [exp( B ): 0.36, 0.41, and 0.18] and a negative effect in Norway [exp( B ): 7.46, 5.01, and 3.72].

Six MR preventive factors were signifi cantly different between the two countries: the identifi cation of a patient ’ s crisis triggers ( χ 2 , P � 0.01), patient – staff ratio ( χ 2 , P � 0.00), staff education ( χ 2 , P � 0.00), acceptable work environment ( χ 2 , P � 0.01), substitute staff ( χ 2 , P � 0.00) and separation of acutely disturbed patients ( χ 2 , P � 0.00).

Lastly, we identifi ed six MR preventive factors with confounding effects, i.e. the MR preventive factor acted as a confounder [ Δ exp( B ) � 10%]: identifi cation of the patients ’ crisis triggers [ Δ exp( B ) � � 10%], patient – staff ratio [ Δ exp( B ) � � 11%], staff education [ Δ exp( B ) � � 51%], acceptable work environment [ Δ exp( B ) � � 15%], substitute staff [ Δ exp( B ) � � 17%] and separation of acutely disturbed patients [ Δ exp( B ) � 13%]. In both countries, the following MR preventive factors was associated with lower numbers of MR episodes: identifi cation of patients ’ crisis triggers (Denmark, exp( B ) � 0.65; Norway, exp( B ) � 0.75), a higher patient – staff ratio (Denmark, exp( B ) � 0.96; Norway, exp( B ) � 0.62), higher levels of staff education (Denmark, exp( B ) � 0.34; Norway, exp( B ) � 0.76), better work envi-ronment (Denmark, exp( B ) � 0.52; Norway, exp( B ) � 0.55), reduced use of substitute staff (Denmark, exp( B ) � 0.59;

Denmark (n=87)

Excluded: special high-security forensic units

(n=3)

Norway(n=96)

Non-responders/ Refused/Missing

data (n=41)

Included (n=43, response rate 49%)

Included (n=47, response rate 49%)

Non-responders/Refused/Missing

data (n=49)

Participants from Denmark and Norway (n=90, response rate 49%)

Fig. 1 . Flowchart of the units participating in the survey.

Table 2 . Annual number of mechanical restraint (MR) episodes.

Denmark and Norway ( n � 90)

Denmark ( n � 43)

Norway ( n � 47)

No. Range No. Range No. Range

Annual number of MR episodes 2927 – 1938 – 989 – Mean annual number of MR episodes per unit 32.5 0 – 254 45.1 0 – 254 21.0 0 – 74Mean annual number of MR episodes per bed 2.5 0 – 15.9 3.2 0 – 15.9 2.0 0 – 7.1Mean annual number of MR episodes per 100,000 inhabitants 58 – 72 – 42 –

No., the number of MR episodes. Nine units accounted for 451 of the excluded MR episodes ( � 25 per person), thereby leaving 1938 episodes in Denmark and 989 in Norway for analysis. The number of MR episodes per 100,000 inhabitants in this sub-sample is estimated from the number of MR episodes, the response percentages and the number of inhabitants the actual year.

2009-13) approved the data collection, and we received permission from the Danish Health and Medicines Authorities and the Danish Psychiatric Central Research Register to use the register data on the frequency of MR in Denmark (26).

This study followed the Ethical Guidelines for Nurs-ing Research in the Nordic Countries (Northern Nurse Federation, 2003) and the Recommendations on Legal Protection of Persons Suffering from Mental Disorders Especially Those Placed as Involuntary Patients (27).

Results A fl owchart of the participating units can be found in Figure 1, and the annual number of MR episodes can be found in Table 2 .

A difference in MR episodes was observed between the countries in a model that included the background variables [exp( B ) � 1.92, P � 0.01] (Fig. 1, Table 2).

Table 3 indicates that a single MR preventive factor, staff education [exp( B ) � 0.34, P � 0.00], was signifi -cantly associated with a lower frequency of MR episodes

Nor

d J

Psyc

hiat

ry D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y R

igsh

ospi

tale

t on

01/2

3/15

For

pers

onal

use

onl

y.

JBAK0003

Tekstboks

J BAK ET AL.


Tabl

e 3 .

Dif

fere

nces

in

the

effe

cts

of p

oten

tial

mec

hani

cal

rest

rain

t (M

R)

prev

entiv

e fa

ctor

s on

the

ann

ual

num

ber

of M

R e

piso

des

in D

enm

ark

and

Nor

way

.

MR

pre

vent

ive

fact

ors

Den

mar

kN

orw

ay

SOI:

Int

erac

tion

betw

een

the

MR

pr

even

tive

fact

or

and

the

coun

try

χ 2 : D

iffe

renc

e in

the

pr

eval

ence

of

the

MR

pr

edic

tive

fac

tor

betw

een

coun

trie

s

Δ ex

p( B

) D

iffe

renc

e in

the

ef

fect

of

coun

try

No.

%ex

p( B

)95

% C

is o

f ex

p( B

) P

N

o.%

exp(

B )

95%

CI

of

exp(

B )

P

P

P

%

Cog

nitiv

e m

ilieu

the

rapy

vs.

no

cogn

itive

mili

eu t

hera

py5

130.

36[0

.11 –

1.16

]0.

091

37.

46[2

.94 –

18.8

9]0.

00 *

0.00

* 0.

096

Patie

nt i

nvol

vem

ent

vs. n

o pa

tient

inv

olve

men

t34

810.

66[0

.26 –

1.71

]0.

4037

860.

60[0

.27 –

1.31

]0.

200.

800.

53 �

2Pa

tient

-cen

tred

car

e vs

. no

patie

nt-c

entr

ed c

are

513

0.41

[0.1

2 – 1.

38]

0.15

410

5.01

[2.0

3 – 12

.36]

0.00

* 0.

00 *

0.72

3Id

entifi

cat

ion

of t

he p

atie

nt ’ s

cri

sis

trig

gers

vs.

no

iden

tifi c

atio

n29

690.

65[0

.38 –

1.10

]0.

1140

930.

75[0

.26 –

2.17

]0.

590.

670.

01 *

� 1

0 †

Ris

k as

sess

men

t vs

. alm

ost

none

or

no r

isk

asse

ssm

ent

921

2.85

[1.2

8 – 6.

36]

0.01

1737

1.39

[0.6

0 – 3.

26]

0.45

0.13

0.11

7M

anda

tory

rev

iew

vs.

no

man

dato

ry r

evie

w8

190.

38[0

.15 –

0.97

]0.

047

150.

53[0

.21 –

1.37

]0.

190.

970.

60 �

1St

aff

trai

ning

( �

50%

vs.

les

s)20

491.

26[0

.65 –

2.46

]0.

4921

480.

88[0

.39 –

2.01

]0.

760.

220.

920

Posi

tive

staf

f at

titud

es v

s. l

ess

posi

tive

410

1.30

[0.5

6 – 2.

99]

0.54

1328

1.28

[0.5

3 – 3.

06]

0.59

0.75

0.03

3Pa

tient

– sta

ff r

atio

( �

3 s

taff

per

pat

ient

vs.

les

s)2

50.

96[0

.45 –

2.06

]0.

9125

530.

62[0

.32 –

1.19

]0.

150.

560.

00 *

� 1

1 †

Staf

f ed

ucat

ion

( � 3

.5 y

ears

vs.

les

s)3

70.

34[0

.17 –

0.66

]0.

00 *

4393

0.76

[0.2

8 – 2.

04]

0.59

0.05

0.00

* �

51 †

St

aff

expe

rien

ce (

� 5

yea

rs v

s. l

ess)

3684

0.81

[0.4

5 – 1.

44]

0.47

4498

0.56

[0.1

4 – 2.

19]

0.41

0.84

0.02

� 1

Deb

riefi

ng,

def

usin

g vs

. non

e28

720.

79[0

.46 –

1.36

]0.

4024

652.

04[1

.07 –

3.89

]0.

030.

020.

52 �

3Po

sitiv

e pa

tient

res

pons

es t

o ru

les

vs. n

egat

ive

resp

onse

s32

780.

61[0

.29 –

1.29

]0.

2034

851.

92[0

.49 –

7.44

]0.

350.

800.

42 �

1A

ctiv

ity-b

ased

nur

sing

vs.

no

activ

ity-b

ased

nur

sing

1538

0.78

[0.3

4 – 1.

79]

0.56

2254

0.69

[0.2

5 – 1.

88]

0.47

0.53

0.14

� 6

Acc

epta

ble

wor

k en

viro

nmen

t vs

. poo

r w

ork

envi

ronm

ent

614

0.52

[0.2

2 – 1.

25]

0.14

1740

0.55

[0.3

0 – 1.

03]

0.06

0.93

0.01

* �

15 †

N

o cr

owdi

ng v

s. c

row

ding

1951

0.59

[0.3

1 – 1.

13]

0.11

2059

0.55

[0.2

7 – 1.

14]

0.11

0.22

0.53

6Su

bstit

ute

staf

f ( �

50

vs. m

ore)

819

0.59

[0.3

2 – 1.

10]

0.10

3373

0.54

[0.2

5 – 1.

18]

0.12

0.33

0.00

* �

17 †

St

aff

oppo

sitio

n to

MR

(m

any

vs. s

ome)

1127

1.09

[0.4

7 – 2.

54]

0.84

1739

0.70

[0.3

0 – 1.

63]

0.41

0.65

0.25

� 1

Ala

rm s

yste

ms

(man

y an

d w

ell-

trai

ned

resp

onse

s vs

. les

s w

ell-

trai

ned

resp

onse

s)3

70.

18[0

.03 –

1.17

]0.

075

113.

72[2

.16 –

6.41

]0.

00 *

0.00

* 0.

504

Com

mun

icat

ion

(all

staf

f po

sses

s co

mm

unic

ativ

e sk

ills

vs.

less

tha

n al

l)8

201.

56[0

.85 –

2.85

]0.

1517

361.

03[0

.51 –

2.10

]0.

930.

430.

083

Wel

l-m

aint

aine

d, c

lean

and

tid

y su

rrou

ndin

gs (

alw

ays

vs.

seld

om)

2355

0.89

[0.4

2 – 1.

87]

0.75

2963

1.03

[0.5

3 – 2.

04]

0.92

0.63

0.43

0

Sepa

ratio

n of

acu

tely

dis

turb

ed p

atie

nts

vs. n

o se

para

tion

2150

1.05

[0.4

6 – 2.

40]

0.90

3883

1.94

[0.9

2 – 4.

09]

0.08

0.28

0.00

* 13

†

No.

, th

e nu

mbe

r of

uni

ts w

here

the

fac

tor

was

pre

sent

; %

, th

e pe

rcen

tage

of

units

whe

re t

he f

acto

r w

as p

rese

nt;

exp(

B ),

the

par

amet

er e

stim

ate.

The

out

com

e va

riab

le (

i.e.

the

num

ber

of M

R e

piso

des)

w

as lo

g-tr

ansf

orm

ed. H

ence

, the

exp

( B )

shou

ld b

e in

terp

rete

d m

ultip

licat

ivel

y (e

.g. i

n D

enm

ark,

cog

nitiv

e m

ilieu

ther

apy

resu

lts in

0.3

6 tim

es th

e av

erag

e nu

mbe

r of

MR

inci

dent

s w

ithou

t cog

nitiv

e m

ilieu

th

erap

y (o

r 64

% f

ewer

MR

epi

sode

s oc

cur

with

cog

nitiv

e m

ilieu

the

rapy

tha

n w

ithou

t). 9

5% C

I of

exp

( B ),

95%

Wal

d co

nfi d

ence

int

erva

l fo

r ex

p( B

); P

, P -v

alue

; SO

I, s

econ

d or

der

inte

ract

ion,

P -v

alue

of

a te

st f

or th

e in

tera

ctio

n be

twee

n th

e M

R p

reve

ntiv

e fa

ctor

and

the

coun

try,

i.e.

whe

ther

the

effe

ct o

f th

e M

R p

reve

ntiv

e fa

ctor

dif

fers

bet

wee

n th

e co

untr

ies;

χ 2 ,

P -v

alue

of

a ch

i squ

are

test

for

the

diff

eren

ce

in t

he p

reva

lenc

e of

the

MR

pre

vent

ive

fact

or i

n th

e un

its b

etw

een

the

coun

trie

s; Δ

exp

( B ),

Dif

fere

nces

in

the

effe

ct o

f th

e co

untr

y in

mod

els

with

and

with

out

the

corr

espo

ndin

g M

R f

acto

r (c

onfo

undi

ng

effe

ct o

f th

e M

R p

reve

ntiv

e fa

ctor

) ex

pres

sed

as a

per

cent

age;

neg

ativ

e Δ

exp(

B )

valu

es s

houl

d be

inte

rpre

ted

as a

MR

pre

vent

ive

fact

or th

at m

inim

ises

the

diff

eren

ce b

etw

een

the

coun

trie

s. I

n th

e es

timat

ion

of e

xp( B

), S

OI,

and

Δ e

xp( B

), w

e us

ed l

inea

r re

gres

sion

and

adj

uste

d fo

r th

e fo

llow

ing

back

grou

nd v

aria

bles

: un

it si

ze,

bed

occu

panc

y, t

ype

of b

ed,

reas

on f

or t

he M

R,

fore

nsic

pro

visi

on,

ethn

icity

and

ty

pe o

f un

it. A

dditi

onal

ly, t

he n

estin

g of

uni

ts i

n de

part

men

ts w

as a

djus

ted

usin

g G

ener

aliz

ed E

stim

atin

g E

quat

ions

(G

EE

) te

chni

ques

. * P

� 0

.01.

† Δ e

xp( B

) � �

10%

.

Nor

d J

Psyc

hiat

ry D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y R

igsh

ospi

tale

t on

01/2

3/15

For

pers

onal

use

onl

y.

JBAK0003

Tekstboks



(RCT), Putkonen et al. (29) reported the same tendency, where seclusion and restraint time decreased by 51% after implementing six core strategies to prevent seclu-sion and restraint. In both studies, the effect was attrib-uted to variables other than the patients ’ crisis triggers. In a one-way case-crossover designed study, Fluttert et al. (30) reported that implementing the early recogni-tion method (ERM) decreased the number of seclusions by 54%. The ERM is a technique for patients and staff to work systematically with the patients ’ triggers (early warning signs), crisis intervention and coping strategies; this is the same identifi cation procedure we asked the clinical nurse managers about in the questionnaire. Over-all, the earlier research supports the fi nding that, identifi -cation of a patient ’ s crisis triggers, is part of explaining the difference in MR use between the two countries. The mechanism involved in this process is probably that patients and staff earlier recognize sign of a forthcoming crisis, and together develop counteractions to avoid esca-lation of the crisis, thereby reducing the risk of a MR episode. Early prevention could constitute one of the most effective strategies to reduce MR episodes.

Second, patient – staff ratio: units in which the staff-to-patient ratio was higher than or equal to three-to-one on average vs. those units where this ratio was lower than three-to-one. In Denmark, the effect was small, pos-sibly due to the small number of units where the staff-to-patient ratio was lower than three-to-one on average. The factor explains a large [ Δ exp( B ) � � 11%] portion of the difference in MR episodes between the countries. The effects of the staff-to-patient ratio were partially sup-ported by earlier research. Smith et al. (31) found a 66% decrease in restraint episodes. However, because several factors contributed to the effect, it is unclear what por-tion of the MR reduction was attributable to the 50% increase in staffi ng. Janssen et al. (32) found that seclu-sion was used more commonly when the number of patients per staff was greater in long-term wards. Donat (33) reported a signifi cant association between increased staffi ng levels and decreased MR and seclusion numbers. However, Owen et al. (34) noted that the relative risk of violence increased with more nursing staff. In addition, Bowers et al. (35) reported that increased levels of aggressive incidents were associated with increased staff-ing. Nonetheless, both studies with opposite outcomes did not precisely target MR or seclusion but instead targeted violence and aggression; the results could be infl uenced by staff employment policies, wherein staffi ng levels are increased in units with more aggressive incidents. Our study most likely eliminated this problem by introducing “ type of unit ” and “ type of bed ” as confounders. Overall, the earlier research specifi c dealing with the same condi-tions does support the fi nding that, the staff to patient ratio, is part of explaining the difference in MR use between the two countries. The mechanisms involved in

Norway, exp( B ) � 0.54), except more separation of acutely disturbed patients, which was associated with increased MR episodes (Denmark, exp( B ) � 1.05; Norway, exp( B ) � 1.94).

Discussion Three potential MR preventive factors displayed very dif-ferent effects in the two countries, which was demon-strated by signifi cant interactions among the MR preventive factor, the countries and the frequency of MR episodes. These preventative factors included cognitive milieu therapy (SOI, P � 0.00), patient-centred care (SOI, P � 0.01) and alarm systems (SOI, P � 0.00). None of the factors had a signifi cant difference in prevalence ( χ 2 , P � 0.09, 0.72, and 0.50). Therefore, it was diffi cult to interpret the differences in the effect based on the coun-try [ Δ exp( B )] due to the interaction. It was also diffi cult to determine why these factors exhibited a different effect in the two countries. One explanation of why cognitive milieu therapy and patient-centred care were effective in Denmark could be that these two factors were well described as methods and included a treatment philoso-phy in Denmark. Therefore, in Denmark, these descrip-tions of care and treatment actually covered phenomena such as patient involvement and cooperation; this was, to our knowledge, not the case in Norway in 2010. Another potential explanation is that the limited number of units in Norway that had implemented the three MR preven-tive factors could be special units with fewer MR epi-sodes. Although the three MR preventive factors were signifi cantly associated with an increased number of MR episodes in Norway, we would not recommend that Nor-way implement less of these three factors given the small number of units that implemented the factors. Further research is needed to explore the different effects of these three MR preventive factors.

Six MR preventive factors exhibited a signifi cant dif-ference between the countries with regard to the fre-quency of MR episodes [ Δ exp( B )].

First, identifi cation of a patient ’ s crisis triggers: units wherein the initial nurse evaluation of at-risk patients always contained a written evaluation of aggressive, self-mutilating or suicidal behaviours (e.g. crisis triggers, unique calming techniques, and coping strategies) vs. units where the initial evaluation only occasionally or never contained a written evaluation. This factor explains a large [ Δ exp( B ) � � 10%] portion of the difference in MR episodes between the countries. The identifi cation of a patient ’ s crisis triggers was supported by earlier research. In a well-designed before and after study, Joni-kas et al. (28) observed a signifi cant decrease (between 49% and 99%) in physical restraint episodes after patients were interviewed to determine their crisis trig-gers and personal crisis management strategies. In a rea-sonably designed cluster randomized controlled trial

Nor

d J

Psyc

hiat

ry D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y R

igsh

ospi

tale

t on

01/2

3/15

For

pers

onal

use

onl

y.

JBAK0003

Tekstboks

J BAK ET AL.


association between work-related factors and some types of violence. Some of the other examined factors are also part of the overall work environment, such as staff training, staff/patient ratio, etc. Overall, the research that supports the association between MR and work environment includes many parts of a healthy work environment, and it supports the fi nding that the work environment partly explains the difference in MR use between the two countries Whatever part of the environment that creates a positive work environment, the wellbeing of the staff is most certainly affecting the wellbeing of the patients, so when staff experiences a more positive work environment the patient probably experience a more positive environment.

Fifth, substitute staff: units in which substitute staff was used less than 50 times (8-h shifts) over the year vs. units where these employees were used greater than or equal to 50 times. The factor explains a large [ Δ exp( B ) � � 17%] portion of the difference in MR epi-sodes between the countries. The effects of substitute staff were supported by earlier research. Bonner et al. (8) found that both patients and staff discriminated between permanent and regular staff. Bowers et al. (44) reported that adverse incidents were more likely to occur during weeks where regular staff was absent. In another study, Bowers et al. (35) noted that increased levels of aggression were associated with an increased use of bank and agency staff. None of the studies was an RCT, but all the studies indicate similar results, supporting less use of substitute staff. The reason for this fi nding could be that substitute staff lacks knowledge of the unit struc-tures, processes and patients, thereby creating insecurity and disturbances in the everyday life of the unit among some of the most vulnerable patients. This could be a stressing factor for the patients creating more aggressive or self-destructive behaviour — hence, more MR episodes.

Sixth, one factor, separation of acutely disturbed patients, had different results: units that were able to separate acutely disturbed patients from other patients vs. units that were unable to separate patients. The factor explains a large [ Δ exp( B ) � 13%] portion of the differ-ence in MR episodes between the countries. This nega-tive association was not supported by earlier research; in fact, the opposite result was indicated. In a qualitative content analysis study, Meehan et al. (45) noted that patients proposed that the separation of acutely disturbed patients would reduce the number of aggressive inci-dents. Overall, this result is not supported by earlier research. Therefore, more research is needed to explore the mechanisms involved in placing one or few patients in a separate and isolated part of the unit.

Limitations Several limitations must be considered when examining the results.

higher staffi ng levels could be that the staff would be more able to observe, interact and corporate with the patients, probably reduce the number of stressed staff, thereby calming the environment, having more personal resources to positive reinforcement, support, and commu-nication, hopefully creating a more empowering and recovery orientated environment for the patients.

Third, staff education: units wherein the staff received basic training for at least 3 1/2 years vs. units where the staff had less than 3 1/2 years of training. The factor explains a large portion [ Δ exp( B ) � � 51%] of the differ-ence in MR episodes between the countries. Staff educa-tion was supported by earlier research. Janssen et al. (32) reported that seclusion was used more commonly in units with less educated staff in the long-term wards. Gim et al. (36) found a signifi cant difference between psychi-atric trained and non-psychiatric trained staff in identify-ing and managing precipitants of violence. In a large 20-year analysis of 1,565 mental health workers, Flan-nery et al. (37) found that staff who were victims of patients ’ aggressions were less formally educated and less trained. None of the studies was an RCT, but other-wise well designed, so it is reasonable to deduce, that earlier research supports the fi nding that, staff education, is part of explaining the difference in MR use between the two countries. Higher basic education among staff is probably leaving the staff with higher analytic and action competences, even in crisis situations, thereby reducing the number of MR episodes.

Fourth, acceptable work environment: units where a workplace environment assessment was present, no or only insignifi cant physical or mental work-environment problems existed, and the overall work environment was evaluated as acceptable vs. units where the workplace environment was not assessed, more work-environment problems existed, and the overall work environment was evaluated as poor. This factor explains a large [ Δ exp( B ) � � 15%] portion of the differences in MR episodes between the countries. An acceptable work environment is somewhat complex because the work environment is a mix of leadership, degree of order, organization, staffi ng, training, crowding, etc. De Benedictis et al. (38) reported that several organiza-tional factors, e.g. team climate, staff perception of aggression and interaction of team-members, were asso-ciated with increased use of seclusion and restraint. Hanrahan et al. (39) observed a relationship between psychiatric nurses ’ work environments and inpatients ’ environments. van der Schaff et al. (40) reported a number of building design features that had an effect on seclusion and restraint. Virtanen et al. (41) indicated that patient overcrowding was associated with violent assaults towards employees. Huckshorn (42) found that leaders played a critical role in reducing seclusion and restraint. Camerino et al. (43) found a signifi cant

Nor

d J

Psyc

hiat

ry D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y R

igsh

ospi

tale

t on

01/2

3/15

For

pers

onal

use

onl

y.

JBAK0003

Tekstboks



MR, but it is important to acknowledge that this type of comparative research does not allow for cause – effect conclusions. Thus, we cannot claim that the implementa-tion of these potential preventive measures in Denmark and Norway would reduce the number of MR episodes. More high-quality studies are needed to investigate the identifi cation of a patient ’ s crisis triggers, an increased number of staff per patient, increased staff education, better work environment, reduced use of substitute staff, and separation of acutely disturbed patients.

Notes

The review was a a systematic review combining qual-1. itative and quantitative research, so the highest recom-mendation grade could be: qualitative; at least one high-quality meta-synthesis, systematic review, or meta-summary, or a body of high-quality evidence, or quan-titative; at least one meta-analysis, systematic review or one or a body of high-quality RCT. STEP 1: READING: Determine if it is diffi cult for the 2. interviewers to read the question uniformly to all respondents. STEP 2: INSTRUCTIONS: Look for problems with any introductions, instructions or expla-nations from the respondent ’ s point of view. STEP 3: CLARITY: Identify problems related to communicat-ing the intent or meaning of the question to the respond-ent. STEP 4: ASSUMPTIONS: Determine whether there are problems with assumptions or the underlying logic. STEP 5: KNOWLEDGE/MEMORY: Check whether respondents are likely to not know or have trouble remembering information. STEP 6: SENSITIV-ITY/BIAS: Assess questions for sensitive nature or wording, and for bias. STEP 7: RESPONSE CATEGO-RIES: Assess the adequacy of the range of responses to be recorded. STEP 8: OTHER: Look for problems not identifi ed in Steps 1 – 7 (19).

Acknowledgements — We thank the Health Science Research Foundation in the Capital Region of Denmark, the Mental Health Services in the Capital Region of Denmark, and the Mental Health Centre Sct. Hans, Copenhagen University Hospital for funding this project. We thank all of the clinical nurse managers of the “ closed ” psychiatric units in Norway and Denmark for participating in this project.

Declaration of interest: The authors report no confl icts of interest. The authors alone are responsible for the content and writing of the paper.

References Indenrigs- og Sundhedsministeriet . Bekendtg ø relse af lov om 1. anvendelse af tvang i psykiatrien . Psykiatriloven, LBK nr . 1729 ; 2010 . Helse- og omsorgsdepartementet . Lov om etablering og gjennom-2. f ø ring av psykiak helsevern . Lov nr . 62 ; 1999 .

First, some confounding covariates might be missing from the analyses. We believe that we included the most important covariates or covered the effects of others by proxy (46), e.g. patients who experience MR could be more aggressive or self-mutilating in one of the coun-tries — hence, acting as a confounder. However, one ear-lier study that assessed aggressive behaviour (47) did not report large differences between Denmark and Norway. Additionally, some of the covariates (number of forensic patients, type of bed and the reason for the MR) could eliminate portions of the difference. Therefore, it is not likely that the potential differences in aggressive or self-mutilating patients would infl uence the results.

Second, recall bias is possible, especially for the Danish data; the clinical nurse manager had to recall all MR preventive factors as well as the conditions and number of MR episodes from 2 years ago. The outcome variable (i.e. the number of MR episodes) were validated or re-examined, and this did not reveal signifi cant recall bias with regard to the outcome variable. Additionally, recall bias could affect the results if the MR preventive factors had only been implemented for a portion of the year. Thus, this recall bias would most likely only increase the effect on patient involvement. Nevertheless, we must consider the possibility of recall bias with regard to some variables.

Third, three participants answered the questionnaires twice during a 3-month interval to check the test – retest reliability. Although three participants are insuffi cient to calculate reliability properly, the results indicate no sub-stantial difference in the measured constructs between the two administrations.

Fourth, selection bias — the non-responders ’ reasons for failure to reply were reviewed, and no systematic differences were observed between responders and non-responders. Reviewing the Danish National Register on coercion, we found that 49% of responding units were “ responsible ” for 52% of the MR episodes in the same population, suggesting minimal differences in MR epi-sodes between responders and non-responders; however, no proper drop-out analysis was conducted. Thus, the results most likely can be generalized to the population of both countries. Transferring the results to other coun-tries, however, would require a thorough examination of their cultures, legislation and other relevant factors.

Finally, when data were reanalysed to include patients who experienced MR more than 25 times, the results changed only marginally. Therefore, the exclusion of the patients who experienced MR more than 25 times did not affect results.

Conclusion The comparison of these MR preventive factors may contribute to better understanding of the phenomenon of

Nor

d J

Psyc

hiat

ry D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y R

igsh

ospi

tale

t on

01/2

3/15

For

pers

onal

use

onl

y.

JBAK0003

Tekstboks

J BAK ET AL.


Munk-J ø rgensen P , Mortensen PB . The Danish Psychiatric Central 26. Register . Dan Med Bull 1997 ; 44 : 82 . Council of Europe . White Paper Regarding a Draft Recommenda-27. tion on Legal Protection of Persons Suffering from Mental Disorder Especially Those Placed as Involuntary Patients . Strasbourg: Coun-cil of Europe ; 2000 . Jonikas JA , Cook JA , Rosen C , Laris A , Kim JB . A program to 28. reduce use of physical restraint in psychiatric inpatient facilities . Psychiatr Serv 2004 ; 55 : 818 – 20 . Putkonen A , Kuivalainen S , Louheranta O , Repo-Tiihonen E , 29. Ryyn ä nen O-P , Kautiainen H , et al . Cluster-randomized controlled trial of reducing seclusion and restraint in secured care of men with schizophrenia . Psychiatr Serv 2013 ; 64 : 850 – 5 . Fluttert FAJ , Van Meijel B , Nijman H , Bj ø rkly S , Grypdonck M . 30. Preventing aggressive incidents and seclusions in forensic care by means of the Early Recognition Method . J Clin Nurs 2010 ; 19 : 1529 – 37 . Smith GM , Davis RH , Bixler EO , Lin HM , Altenor A , Altenor RJ , 31. et al . Pennsylvania State Hospital System ’ s seclusion and restraint reduction program . Psychiatr Serv 2005 ; 56 : 1115 . Janssen W , Noorthoorn E , Linge R , Lendemeijer B . The infl uence 32. of staffi ng levels on the use of seclusion . Int J Law Psychiatry 2007 ; 30 : 118 – 26 . Donat DC . Impact of improved staffi ng on seclusion/restraint reli-33. ance in a public psychiatric hospital . Psychiatr Rehabil J 2002 ; 25 : 413 – 16 . Owen C , Tarantello C , Jones M , Tennant C . Violence and 34. aggression in psychiatric units . Psychiatr Serv 1998 ; 49 : 1452 – 57 . Bowers L , Allan T , Simpson A , Jones J , Van Der Merwe M , Jeffery 35. D . Identifying key factors associated with aggression on acute inpa-tient psychiatric wards . Issues Ment Health Nurs 2009 ; 30 : 260 – 71 . Gim YS , Cheng SOB , Hong TL , Chin MWB , Hen WH , Lee YJ . 36. Nurses experiences in identifying and managing violence precipi-tants in a psychiatric in-patient setting . Nurs J Singapore 1999 ; 26 : 21 – 4 . Flannery Jr RB, LeVitre V , Rego S , Walker AP . Characteristics of 37. staff victims of psychiatric patient assaults: 20-year analysis of the Assaulted Staff Action Program . Psychiatr Q 2011 ; 22 : 11 – 21 . De Benedictis L , Dumais A , Sieu N , Mailhot MP , Tran MAM , 38. Stikarovska I , et al . Staff perceptions and organizational factors as predictors of seclusion and restraint on psychiatric wards . Psychiatr Serv 2011 ; 62 : 484 – 91 . Hanrahan NP , Aiken LH , McClaine L , Hanlon AL . Relationship 39. between psychiatric nurse work environments and nurse burnout in acute care general hospitals . Issues Ment Health Nurs 2010 ; 31 : 198 – 207 . van der Schaaf PS , Dusseldorp E , Keuning FM , Janssen WA , 40. Noorthoorn EO . Impact of the physical environment of psychiatric wards on the use of seclusion . Br J Psychiatry 2013 ; 202 : 142 – 9 . Virtanen M , Vahtera J , Batty GD , Tuisku K , Pentti J , Oksanen T , 41. et al . Overcrowding in psychiatric wards and physical assaults on staff: Data-linked longitudinal study . Br J Psychiatry 2011 ; 198 : 149 – 55 . Huckshorn KA . Reducing seclusion and restraint use in mental 42. health settings: A phenomenological study of hospital leader and staff experiences . Minneapolis, MN: Capella University ; 2013 . Camerino D , Estryn-Behar M , Conway PM , van Der Heijden BIJM , 43. Hasselhorn HM . Work-related factors and violence among nursing staff in the European NEXT study: A longitudinal cohort study . Int J Nurs Stud 2008 ; 45 : 35 – 50 . Bowers L , Allan T , Simpson A , Nijman H , Warren J . Adverse inci-44. dents, patient fl ow and nursing workforce variables on acute psychi-atric wards: The Tompkins Acute Ward Study . Int J Soc Psychiatry 2007 ; 53 : 75 – 84 . Meehan T , McIntosh W , Bergen H . Aggressive behaviour in the high 45. secure forensic setting: The perceptions of patients . J Psychiatr Ment Health Nurs 2006 ; 13 : 19 – 25 . Janssen WA , van de Sande R , Noorthoorn EO , Nijman HLI , 46. Bowers L , Mulder CL , et al . Methodological issues in monitoring the use of coercive measures . Int J Law Psychiatry 2011 ; 34 : 429 – 38 . Nijman HLI , Palmstierna T , Almvik R , Stolker JJ . Fifteen years of 47. research with the Staff Observation Aggression Scale: A review . Acta Psychiatr Scand 2005 ; 111 : 12 – 21 .

Hem E , Steen O , Opjordsmoen S . Thrombosis associated with 3. physical restraints . Acta Psychiatr Scand 2001 ; 103 : 73 – 6 . Hick JL , Smith SW , Lynch MT . Metabolic Acidosis in Restraint 4. associated Cardiac ArrestA Case Series . Acad Emerg Med 1999 ; 6 : 239 – 43 . Laursen SB , Jensen TN , Bolwig T , Olsen NV . Deep venous 5. thrombosis and pulmonary embolism following physical restraint . Acta Psychiatr Scand 2005 ; 111 : 324 – 7 . Morrison A , Sadler D . Death of a psychiatric patient during phys-6. ical restraint. Excited delirium — A case report . Med Sci Law 2001 ; 41 : 46 – 50 . Nielsen AS . Deep venous thrombosis and fatal pulmonary embo-7. lism in a physically restrained patient . Ugeskr Laeger 2005 ; 167 : 2294 . Bonner G , Lowe T , Rawcliffe D , Wellman N . Trauma for all: A 8. pilot study of the subjective experience of physical restraint for mental health inpatients and staff in the UK . J Psychiatr Ment Health Nurs 2002 ; 9 : 465 – 73 . Bower FL , McCullough CS , Timmons ME . A synthesis of what 9. we know about the use of physical restraints and seclusion with patients in psychiatric and acute care settings: 2003 update . J Knowl Synth Nurs 2003 ; 10 : 1 – 29 . Cusack KJ , Frueh BC , Hiers T , Suffoletta-Maierle S , Bennett S . 10. Trauma within the psychiatric setting: A preliminary empirical report . Adm Policy Ment Health 2003 ; 30 : 453 – 60 . Frueh BC , Knapp RG , Cusack KJ , Grubaugh AL , Sauvageot JA , 11. Cousins VC , et al . Special section on seclusion and restraint: Patients ’ reports of traumatic or harmful experiences within the psy-chiatric setting . Psychiatr Serv 2005 ; 56 : 1123 – 33 . Steinert T , Bergbauer G , Schmid P , Gebhardt RP . Seclusion and 12. restraint in patients with schizophrenia: Clinical and biographical correlates . J Nerv Ment Dis 2007 ; 195 : 492 – 6 . Bak J , Brandt-Christensen M , Sestoft DM , Zoffmann V . 13. Mechanical restraint — Which interventions prevent episodes of mechanical restraint? A systematic review . Perspect Psychiatr Care 2011 ; 48 : 83 – 94 . Bak J , Zoffmann V , Sestoft DM , Almvik R , Brandt-Christensen M . 14. Mechanical restraint in psychiatry: Preventive factors in theory and practice. A Danish – Norwegian Association study . Perspect Psychi-atr Care 2013 ; 50 : 155 – 66 . Bak J , Aggern æ s H . Coercion within Danish psychiatry compared 15. with 10 other European countries . Nord J Psychiatry 2011 ; 66 : 297 – 302 . Bremmes R , Hatling T , Bj ø rngaard JH . Bruk av tvangsmidler i 16. psykisk helsevern i 20 01, 2003, 2005 og 2007. Trondheim: SINTEF Helse ; 2008 . Sundhedsstyrelsen . Nye tal fra sundhedsstyrelsen. Anvendelse af 17. tvang i psykiatrien 2007 . K ø benhavn: Sundhedsstyrelsen; 2008 . Bowers L , Van Der Werf B , Vokkolainen A , Muir-Cochrane E , 18. Allan T , Alexander J . International variation in containment meas-ures for disturbed psychiatric inpatients: A comparative question-naire survey . Int J Nurs Stud 2007 ; 44 : 357 – 64 . Willis GB , Lessler JT . Question Appraisal System (QAS-99) . 19. Atlanta, GA: Research Triangle Institute ; 1999 . Willis GB . Cognitive interviewing. A tool for improving question-20. naire design . London: Sage ; 2005 . Watt T , Rasmussen Å , Groenvold M , Bjorner JB , Watt SH , 21. Bonnema SJ , et al . Improving a newly developed patient-reported outcome for thyroid patients, using cognitive interviewing . Qual Life Res 2008 ; 17 : 1009 – 17 . Olsen H . Guide til gode sp ø rgeskemaer . K ø benhavn: Socialforskn-22. ingsinstituttet ; 2006 . Schaeffer NC , Presser S . The science of asking questions . 23. Annu Rev Sociol 2003 ; 29 : 65 – 89 . Keski-Valkama A , Sailas E , Eronen M , Koivisto AM , L ö nnqvist J , 24. Kaltiala-Heino R . Who are the restrained and secluded patients: A 15-year nationwide study . Soc Psychiatry Psychiatr Epidemiol 2010 ; 45 : 1087 – 93 . Knutzen M , Sandvik L , Hauff E , Opjordsmoen S , Friis S . 25. Association between patients ’ gender, age and immigrant background and use of restraint — A 2-year retrospective study at a department of emergency psychiatry . Nord J Psychiatry 2007 ; 61 : 201 – 6 .

Nor

d J

Psyc

hiat

ry D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y R

igsh

ospi

tale

t on

01/2

3/15

For

pers

onal

use

onl

y.

http://informahealthcare.com/action/showLinks?crossref=10.1111%2Fj.1553-2712.1999.tb00164.x&isi=000079862700015

http://informahealthcare.com/action/showLinks?crossref=10.1007%2Fs11136-008-9364-z&isi=000258657900005

http://informahealthcare.com/action/showLinks?crossref=10.1111%2Fj.1744-6163.2011.00307.x&isi=000302060300005


http://informahealthcare.com/action/showLinks?crossref=10.1007%2Fs11126-010-9153-z

http://informahealthcare.com/action/showLinks?crossref=10.1176%2Fappi.ps.201200393&isi=000327277600007

http://informahealthcare.com/action/showLinks?crossref=10.1111%2Fppc.12036&isi=000339274900003



http://informahealthcare.com/action/showLinks?crossref=10.1016%2Fj.ijlp.2011.10.008&isi=000298905900008

http://informahealthcare.com/action/showLinks?crossref=10.1176%2Fps.62.5.pss6205_0484&isi=000290028700007

http://informahealthcare.com/action/showLinks?isi=000168249000009

http://informahealthcare.com/action/showLinks?crossref=10.1176%2Fappi.ps.56.9.1115&isi=000231718300013

http://informahealthcare.com/action/showLinks?crossref=10.1146%2Fannurev.soc.29.110702.110112&isi=000185094800004

http://informahealthcare.com/action/showLinks?system=10.3109%2F08039488.2011.632645&isi=000309427200001


http://informahealthcare.com/action/showLinks?system=10.3109%2F01612840903200068

http://informahealthcare.com/action/showLinks?crossref=10.1192%2Fbjp.bp.112.118422&isi=000314861100015

http://informahealthcare.com/action/showLinks?crossref=10.1016%2Fj.ijlp.2006.04.005&isi=000244993100003

http://informahealthcare.com/action/showLinks?crossref=10.1007%2Fs00127-009-0150-1&isi=000283117900008

http://informahealthcare.com/action/showLinks?crossref=10.1192%2Fbjp.bp.110.082388&isi=000287173400013

http://informahealthcare.com/action/showLinks?crossref=10.1046%2Fj.1365-2850.2002.00504.x

http://informahealthcare.com/action/showLinks?crossref=10.1037%2Fh0094994&isi=000175516100013

http://informahealthcare.com/action/showLinks?system=10.1080%2F08039480701352520&isi=000246808400007

http://informahealthcare.com/action/showLinks?crossref=10.1023%2FA%3A1024697715409

http://informahealthcare.com/action/showLinks?crossref=10.1016%2Fj.ijnurstu.2006.01.005&isi=000245500600006

http://informahealthcare.com/action/showLinks?crossref=10.1176%2Fps.49.11.1452&isi=000076905200011


http://informahealthcare.com/action/showLinks?crossref=10.1016%2Fj.ijnurstu.2007.01.013&isi=000252517000005

http://informahealthcare.com/action/showLinks?system=10.1080%2F01612840802710829

http://informahealthcare.com/action/showLinks?isi=A1997WL42900007

http://informahealthcare.com/action/showLinks?crossref=10.1097%2FNMD.0b013e3180302af6&isi=000247128000005

http://informahealthcare.com/action/showLinks?crossref=10.1177%2F0020764007075011&isi=000244389000007


JBAK0003

Tekstboks



Roger Almvik, R.N., R.M.N., D.Phil., Research Director, St. Olavs University Hospital, Forensic Department Br ö set, Centre for Research & Education in Forensic Psychiatry, Norway. E-mail: [email protected]. Volkert Dirk Siersma, Ph.D., Statistician, The Research Unit for General Practice and Section of General Practice, Department of Public Health, University of Copenhagen, Copenhagen, Denmark, E-mail: [email protected]. Mette Brandt-Christensen, M.D., Ph.D., Head of Department, Forensic Psychiatry, Mental Health Centre Glostrup, Copenhagen University Hospital, Denmark, E-mail: [email protected].

Jesper Bak, R.N., S.D., M.P.H., Ph.D. student, Head nurse, Mental Health Centre Sct. Hans, Roskilde, Denmark. E-mail: [email protected]. Vibeke Zoffmann, R.N., M.P.H., Ph.D., Head of Research, Women and Children ’ s Health, Juliane Marie Centre, Copenhagen University Hospital, Rigshospitalet, Copenhagen, Denmark. E-mail: [email protected]. Dorte Maria Sestoft, M.D., Ph.D., Head of Department, Ministry of Justice, Clinic of Forensic Psychiatry, Copenhagen, Denmark, E-mail: [email protected].

Supplementary material available online Appendix: A description of mechanical restraint (MR) preventive factor measurements

Nor

d J

Psyc

hiat

ry D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y R

igsh

ospi

tale

t on

01/2

3/15

For

pers

onal

use

onl

y.

JBAK0003

Tekstboks

Corrigendum (private)

Correction in the abstract:

Conclusions: These six MR preventive factors might partially explain the difference in the

frequency of MR episodes observed in the two countries, i.e. higher numbers in Denmark than

Norway. One MR preventive factor was not supported by earlier research, the identification of

the patient’s crisis triggers; therefore, more research on the mechanisms involved is needed.

Should be corrected to:

Conclusions: These six MR preventive factors might partially explain the difference in the

frequency of MR episodes observed in the two countries, i.e. higher numbers in Denmark than

Norway. One MR preventive factor was not supported by earlier research, separation of acutely

disturbed patients; therefore, more research on the mechanisms involved is needed.

Correction on page seven:

Second, patient-staff ratio: units in which the staff-to-patient ratio was higher than or equal to

three-to-one on average vs. those units where this ratio was lower than three-to-one. In

Denmark, the effect was small, possibly due to the small number of units where the staff-to-

patient ratio was lower than three-to-one on average.

Should be corrected to:

Second, patient-staff ratio: units in which the staff-to-patient ratio was higher than or equal to

three-to-one on average vs. those units where this ratio was lower than three-to-one. In

Denmark, the effect was small, possibly due to the small number of units where the staff-to-

patient ratio was higher than three-to-one on average.


Supplementary material for Bak J, et al., Comparing the effect of non-medical mechanical restraint preventive factors between psychiatric units in Denmark and Norway, Nordic Journal of Psychiatry, 2015; doi: 10.3109/08039488.2014.996600.

Supplementary Appendix: A description of mechanical restraint (MR) preventive factor measurements.

MR preventive factors Measurement description

Cognitive milieu therapy vs. no cognitive milieu therapy

The clinical nurse manager described the care in his or her unit as “ cognitive behaviour therapy ” or “ cognitive milieu therapy ” , and the staff had participated in 40 h of training or more in this care method — The care method was not described.

Patient involvement vs. no patient involvement One of the following factors was present: the patients participated in ward rounds/conferences or the patients infl uenced the rules imposed by the unit or infl uenced what was written in the nurse ’ s journal — Patients did not participate in these areas.

Patient-centred care vs. no patient-centred care The clinical nurse manger described the care in his or her unit using the words “ individual ” , “ cooperative ” or “ patient-centred ” , and the staff participated in 40 h of training or more in this care method — These words were not used to describe this care method.

Identifi cation of the patient ́ s crisis triggers vs. no identifi cation

The initial nurse evaluation of at-risk patients always contained a written evaluation of aggressive, self-mutilating or suicidal behaviours (e.g. crisis triggers, unique calming techniques and coping strategies) — The initial evaluation only occasionally or never contained a written evaluation.

Risk assessment vs. almost none or no risk assessment

The risk of developing aggressive/violent or self-mutilating/suicidal behaviours was always or almost always systematically evaluated and documented, and care actions were initiated — These behaviours were only occasionally or never evaluated and documented.

Mandatory review vs. no mandatory review An audit, review or similar systematic follow-up evaluation was conducted after all MR episodes — The evaluation was only conducted after some or none of the MR episodes.

Staff training ( � 50% vs. less) An average of 50% or more of the staff participated in training regarding de-escalation, avoiding physical attacks, careful restraining techniques, legislation, MR predictors/antecedents and suicide prevention — Less than 50% of the staff experienced such training.

Positive staff attitudes vs. less positive All staff directed their attention towards the patient and possessed communicative skills, optimism and empathy — Not all of the staff possessed these skills.

Patient-staff ratio ( � 3 staff per patient vs. fewer) The staff-to-patient ratio was higher than or equal to three-to-one on average — This ratio was lower than three-to-one on average.

Staff education ( � 3.5 years vs. fewer) The staff received basic training for at least 3 ½ years — The staff had less than 3 ½ years of training.

Staff experience ( � 5 years vs. fewer) The staff had an average of more than 5 years of experience in the psychiatric speciality was more than 5 years on average — Staff experience was 5 years or fewer.

Debriefi ng and defusing vs. none Follow-up evaluations were initiated after each MR episode for debriefi ng, defusing or other crisis interventions — No follow-up was initiated.

Positive patient responses to rules vs. negative responses

The patients reacted positively to the rules, and limit-setting was motivational — The opposite occurred.

Activity-based nursing vs. no activity-based nursing Individual patient-activity plans were devised, and the patients participated in many or several physical or other types of activities — The patients had no activity plans, and they participated in few or no physical or other activities.

Acceptable work environment vs. poor work environment

The unit had the following features: a workplace environment assessment was present, no or only insignifi cant physical or mental work-environment problems existed, and the overall work environment was evaluated as acceptable — The workplace environment was not assessed, more work-environment problems existed and the overall work environment was evaluated as poor.

No crowding vs. crowding Two of the following three factors were evident for no crowding: only one bed in a patient ’ s room, more than 25 m 2 of all day-accessible space per patient and the perception of no crowding — Only one or none of these factors was evident.

Substitute staff ( � 50 vs. more) Substitute staff was used less than 50 times (8-h shifts) over the year — These employees were used greater than or equal to 50 times.

Staff opposition to MR (many vs. some) All staff had the attitude that “ MR should be initiated rarely to protect the lives of the patient and others ” — Not all staff had this attitude.

Alarm systems (numerous well trained responses vs. less well trained responses)

More than three night-shift staff members responded to an alarm call, and they were all trained in this task — Not all of the night-shift staff members were trained, or they were trained but failed to respond.

Communication (all staff possess communicative skills vs. less than all)

All staff had general communicative skills or special skills with regard to communicative de-escalation — Less than all the staff had these skills.

Well maintained, clean and tidy surroundings (always vs. seldom)

The unit was perceived as well maintained, clean and tidy — The perceptions were otherwise.

Separation of acutely disturbed patients vs. no separation

The unit was able to separate acutely disturbed patients from other patients — The unit was unable to do so.

phd thesis - region hovedstadens psykiatri · ken, b: kendall’s taub soi: second order...

Documents