the oral health of critically ill children -...
TRANSCRIPT
THE ORAL HEALTH OF
CRITICALLY ILL CHILDREN
Amanda Ullman
B.N., Grad. Cert. PICU
n5669405
A thesis submitted as partial requirement of fulfilment of a
MASTERS OF APPLIED SCIENCE (RESEARCH)
School of Nursing and Midwifery
QUEENSLAND UNIVERSITY OF TECHNOLOGY
2009
i
ii
ABSTRACT
Introduction. In adults, oral health has been shown to worsen during critical
illness as well as influence systemic health. There is a paucity of paediatric critical
care research in the area of oral health; hence the purpose of the Critically ill
Children’s Oral Health (CCOH) study is to describe the status of oral health of
critically ill children over time spent in the paediatric intensive care unit (PICU). The
study will also examine the relationship between poor oral health and a variety of
patient characteristics and PICU therapies and explore the relationship between
dysfunctional oral health and PICU related Healthcare-Associated Infections (HAI).
Method. An observational study was undertaken at a single tertiary-referral
PICU. Oral health was measured using the Oral Assessment Scale (OAS) and
culturing oropharyngeal flora. Information was also collected surrounding the use of
supportive therapies, clinical characteristics of the children and the occurrence of
PICU related HAI.
Results. Forty-six participants were consecutively recruited to the CCOH
study. Of the participants 63% (n=32) had oral dysfunction while 41% (n=19)
demonstrated pathogenic oropharyngeal colonisation during their critical illness. The
potential systemic pathogens isolated from the oropharynx and included Candida sp.,
Staphylococcus aureus, Haemophilus influenzae, Enterococcus sp. and Pseudomonas
aeruginosa. The severity of critical illness had a significant positive relationship
(p=0.046) with pathogenic and absent colonisation of the oropharynx. Sixty-three
iii
percent of PICU-related HAI involved the preceding or simultaneous colonisation of
the oropharynx by the causative pathogen.
Conclusion. Given the prevalence of poor oral health during childhood critical
illness and the subsequent potential systemic consequences, evidence based oral
hygiene practices should be developed and validated to guide clinicians when nursing
critically ill children.
KEYWORDS
Oral health, oropharyngeal colonisation, oral hygiene, paediatrics, critical illness,
paediatric intensive care, healthcare-associated infection.
FUNDING ATTRACTED BY THE RESEARCH PROGRAMME
Nursing Research Projects Grant, Royal Children’s Hospital Foundation
$23,500.
Novice Researcher Grant, Australian College of Critical Care Nurses $5,000.
iv
TABLE OF CONTENTS
ABSTRACT .............................................................................................................. II
KEYWORDS ............................................................................................................ III
FUNDING ATTRACTED BY THE RESEARCH PROGRAMME ....................... III
TABLE OF CONTENTS ........................................................................................ IV
STATEMENT OF ORIGINAL AUTHORSHIP .................................................. VIII
LIST OF ABBREVIATIONS .................................................................................. IX
ACKNOWLEDGEMENTS ..................................................................................... XI
CHAPTER 1 INTRODUCTION ................................................................................. 1
1.1 INTRODUCTION .................................................................................................... 1
1.2 PURPOSE OF THE RESEARCH ................................................................................ 4
1.3 SIGNIFICANCE OF STUDY........................................................................................5
1.4 SUMMARY ............................................................................................................. 5
CHAPTER 2 LITERATURE REVIEW ...................................................................... 6
2.1 INTRODUCTION .................................................................................................... 6
2.2 LITERATURE SEARCH STRATEGIES ....................................................................... 6
2.3 ORAL HEALTH ...................................................................................................... 7
2.3.1 Physiology of the oral cavity ......................................................................... 7
2.3.2 The ‘dysfunctional’ oral cavity ................................................................... 10
2.4 CRITICAL ILLNESS IN CHILDHOOD ..................................................................... 12
2.4.1 PICU population ......................................................................................... 12
2.4.2 Measuring critical illness in childhood ...................................................... 14
2.4.3 The effect of PICU therapies on oral health ............................................... 16
2.4.4 The oral health of the critically ill .............................................................. 17
v
2.4.5 Current oral hygiene practices ................................................................... 21
2.5 HEALTHCARE-ASSOCIATED INFECTION (HAI) ................................................... 25
2.5.1 Pneumonia .................................................................................................. 25
2.5.2 Bacteraemia and blood-stream infections .................................................. 27
2.6 CURRENT LIMITATIONS IN LITERATURE .............................................................. 29
2.7 SUMMARY: WHAT THE LITERATURE SUGGESTS .................................................. 31
CHAPTER 3 METHODS ......................................................................................... 32
3.1 INTRODUCTION ................................................................................................... 32
3.2 RESEARCH DESIGN .............................................................................................. 32
3.3 RESEARCH QUESTIONS ........................................................................................ 33
3.4 STUDY SETTING AND POPULATION ...................................................................... 34
3.5 SAMPLING STRATEGY AND SIZE ........................................................................... 35
3.6 DATA COLLECTION ............................................................................................. 36
3.6.1 Oropharyngeal cultures .............................................................................. 36
3.6.1.1 Saliva sampling .................................................................................... 37
3.6.1.2 Laboratory protocol ............................................................................. 39
3.6.2 Instruments ................................................................................................ 41
3.6.2.1 Oral Assessment Scale (OAS) .............................................................. 41
3.6.2.2 Pediatric Logistic Organ Dysfunction score (PELOD) and Paediatric
Indicator of Mortality 2 (PIM2) ....................................................................... 45
3.6.2.3 Clinical Pulmonary Infection Score (CPIS) ........................................ 46
3.6.2.4 Data extraction tool ............................................................................. 47
3.8 DATA ANALYSIS ................................................................................................... 48
3.9 ETHICAL CONSIDERATIONS ................................................................................. 52
vi
3.9.1 Informed Consent ........................................................................................ 52
3.9.2 Risk management procedures ..................................................................... 53
3.9.3 Confidentiality and security ........................................................................ 54
CHAPTER 4 RESULTS ........................................................................................... 55
4.1 INTRODUCTION .................................................................................................. 55
4.2 CHARACTERISTICS OF PARTICIPANTS ................................................................. 55
4.3 DATA ANALYSIS .................................................................................................. 58
4.4.1 What is the status of oral health in critically ill children during admission
to a PICU? ........................................................................................................... 61
4.4.2 How does the oral health of critically ill children change during their
admission to PICU? ............................................................................................. 65
4.4.3 Is the oral health of critically ill children affected by patient characteristics
or PICU therapies? .............................................................................................. 68
4.4.4 Is there a relationship between dysfunctional oral health in critically ill
children and PICU-related healthcare-associated infections (HAI)? ................. 69
CHAPTER 5 DISCUSSION ..................................................................................... 75
5.1 INTRODUCTION .................................................................................................. 75
5.2 KEY FINDINGS .................................................................................................... 76
5.2.1 What is the status of oral health in critically ill children during admission
to a PICU? ........................................................................................................... 76
5.2.2 How does the oral health of critically ill children change during their
admission to PICU? ............................................................................................. 79
5.2.3 Is the oral health of critically ill children affected by patient characteristics
or PICU therapies? .............................................................................................. 81
vii
5.2.4 Is there a relationship between dysfunctional oral health in critically ill
children and PICU-related healthcare-associated infections (HAI)? ................. 83
5.3 STRENGTHS AND LIMITATIONS OF THE STUDY ................................................... 86
5.5 SUMMARY .......................................................... ERROR! BOOKMARK NOT DEFINED.
CHAPTER 6 CONCLUSION .................................................................................. 90
6.1 INTRODUCTION ................................................................................................... 90
6.4 CONCLUSION ....................................................................................................... 94
REFERENCES ........................................................................................................... 95
APPENDICES .......................................................................................................... 101
ORAL ASSESSMENT SCALE (OAS) .......................................................................... 101
PEDIATRIC LOGISTIC ORGAN DYSFUNCTION SCORE (PELOD) .............................. 102
PAEDIATRIC INDEX OF MORTALITY 2 (PIM2) ........................................................ 103
CLINICAL PULMONARY INFECTION SCORE (CPIS) ................................................ 104
DATA EXTRACTION TOOL ....................................................................................... 105
ETHICAL APPROVAL FROM ROYAL CHILDREN’S HOSPITAL, BRISBANE HREC ...... 107
ETHICAL APPROVAL FROM THE QUEENSLAND UNIVERSITY OF TECHNOLOGY HREC
................................................................................................................................ 108
PARENT / GUARDIAN INFORMATION AND CONSENT FORMS ..................................... 110
YOUTH ASSENT INFORMATION AND CONSENT FORMS ............................................. 114
viii
STATEMENT OF ORIGINAL AUTHORSHIP
The work contained in this thesis has not been previously submitted to meet
requirements for an award at this or any other higher education institution. To the best
of my knowledge and belief, the thesis contains no material previously published or
written by another person except where due reference is made.
_______________________________________
Amanda Ullman
_______________________________________
Date
ix
LIST OF ABBREVIATIONS
ACCCN = Australian College of Critical Care Nurses
ANOVA = Analysis of Variance
ANZPICR = Australia and New Zealand Paediatric Intensive Care Registry
ARDS = Acute Respiratory Distress Syndrome
CCOH = Critically ill Children’s Oral Health
CDC = Centers for Disease Control and Prevention
CP = Cerebral Palsy
CFU = Colony Forming Units
CINAHL = Cochrane Library and Cumulative Index to Nursing and Allied Health
CO2 =Carbon Dioxide
CPIS = Clinical Pulmonary Infection Score
CVVHD = Continuous Veno-Venous Haemo-Dialysis
ECMO = ExtraCorporeal Membrane Oxygenation
ETT = Endo-Tracheal Tube
FiO2=Fraction of Inspired Oxygen
HAI = Healthcare-associated infection
HBA = Horse Blood Agar
HREC = Human Research Ethics Committee
ICU = Intensive Care Unit
IgA = Immunoglobulin A
IgM = Immunoglobulin M
INR=International Normalised Ratio
kPa = Kilopascal
LVAD = Left Ventricular Assist Device
x
MAC = MacConkey’s Agar
mmHg = Millimetres of mercury
mmol/l = Millimols per litre
MSA = Mannitol Salt Agar
O2 = Oxygen
OAS = Oral Assessment Scale
PaCO2=Arterial Carbon Dioxide Pressure
PaO2=Arterial Oxygen Pressure
PELOD = Pediatric Logistic Organ Dysfunction Score
PICU = Paediatric Intensive Care Unit
PIM2 = Paediatric Index of Mortality
RBWH = Royal Brisbane and Women’s Hospital
RCH = Royal Children’s Hospital, Brisbane
SAB = Sabouraud Dextrose Agar
SBP = Systolic Blood Pressure
SD = Standard Deviations
SPSS = Statistical Package for Social Sciences
VAP = Ventilator associated pneumonia
xi
ACKNOWLEDGEMENTS
Like many people completing post-graduate work, this Master’s thesis has been a
labour of love.
Thank you to the Royal Children’s Hospital Foundation and the Australian College of
Critical Care Nurses (ACCCN) for the financial support to complete this project.
Thank you to the wonderful nurses at the Royal Children’s Hospital, Paediatric
Intensive Care Unit (PICU), for letting me add to your already busy days and the
support of my Nurse Unit Managers. It is great to work with a group of professionals
who are focused on providing quality, evidence-based, holistic care. A big thanks to
Debbie for pushing me in this path, one simple literature review just keeps getting
bigger.
Thank you to my family for their support - my husband, for letting me off house-
painting duties, and my dad for making it seem achievable and sensible.
Thank you also to my supervisory team, Dr Peter Lewis and Prof. Glenn Gardner.
1
CHAPTER ONE - INTRODUCTION
1.1 INTRODUCTION
Critically ill children experience physiological changes which result in instability and
acute crises, requiring intensive nursing care to support single organ or systemic
dysfunction. The treatment modalities used to support children experiencing critical
illness and the progression of critical illness may result in dysfunction within the oral
cavity. In a healthy child, the oral cavity harbours over 250 strains of commensal
bacteria (O'Reilly, 2003) which change in response to numerous factors including the
child’s general health and well-being. The changes generated by critical illness may
produce an imbalance of commensal oral bacteria, which allows the oral cavity to
become a haven for potential systemic pathogens (Fourrier, Duvivier, Boutigny,
Roussel-Delvallez, & Chopin, 1998). Due to the anatomical connection between the
oral cavity and the respiratory and circulatory systems, pathogens colonising the
oropharynx could potentially transfer to cause systemic infections (Munro, Grap,
Elswick et al., 2006).
In the paediatric intensive care environment, respiratory and blood-stream infections
caused by fungal and bacterial pathogens are associated with substantial financial,
morbidity and mortality costs (Inwald, Tasker, Peters, & Nadel, 2009; Rubenstein,
Kabat, Shulman, & Yogev, 1992; Safdar, Dezfulian, Collard, & Saint, 2005; Singhi,
Raman Rao, & Chakrabarti, 2008; Suljagic et al., 2005; Thorburn et al., 2009; Turton,
2008). However, research describing the prevalence of fungal and bacterial pathogens
in the oropharynx and their systemic effects during critical illness in children is
2
scarce. In comparison, adult critical care research has recognised the impact that poor
oral health may have on the morbidity and mortality outcomes of these patients
(Abele-Horn et al., 1997; A. Berry & Davidson, 2006; A. Berry, Davidson, Masters,
& Rolls, 2007; Binkley, Furr, Carrico, & McCurren, 2004; Brennan et al., 2004;
Chan, Ruest, O'Meade, & Cook, 2007; Cutler, 2005; Ewig et al., 1999; Fitch, Munro,
Glass, & Pellegrini, 1999; Fourrier et al., 1998; Garrouste-Orgeas et al., 1997; Grap,
Munro, Ashtiani, & Bryant, 2003; D. Jones & Munro, 2008; H. Jones, 2005; Kite &
Pearson, 1995; McNeill, 2000; Munro & Grap, 2004; Munro, Grap, Elswick et al.,
2006; O'Reilly, 2003; Pugin, Auckenthaler, Lew, & Suter, 1991; Ross & Crumpler,
2007; Scannapieco, Stewart, & Mylotte, 1992; Somerville, 1999; Stiefel, Damron,
Sowers, & Velez, 2000). This research is now focussing on treatment strategies which
can be used to improve oral health during critical illness. Interest in the systemic
effects of oral health during paediatric critical illness is growing, with a recent
publication by Thorburn et al., (2009) which examined the carriage of abnormal
bacterial flora and antibiotic resistant flora in the pharynx and gut of children with
cerebral palsy requiring mechanical ventilation, and their associated infection rates.
The study also found that in 65% of children with cerebral palsy requiring mechanical
ventilation who developed an infection, the infecting pathogen was carried in the
patients’ pharynx on admission or in their gut flora, and concluded that early targeted
antibiotic therapy may be beneficial.
Within paediatric and adult critical care practice, pneumonia is acknowledged as
being a major threat to mechanically ventilated patients (Chan et al., 2007; Safdar et
al., 2005; Schleder, 2003; Turton, 2008) . The relationship between poor oral health,
in the form of pathogenic oropharyngeal colonisation, and hospital-acquired
3
pneumonia, especially ventilator-associated pneumonia (VAP), has been well
documented in robust adult critical care clinical research (Bonten et al., 1994; Ewig et
al., 1999; Fourrier et al., 1998; Garrouste-Orgeas et al., 1997; Munro & Grap, 2004;
Pugin, Auckenthaler, Lew et al., 1991). Pathogenic microflora that have been isolated
in the dental and oropharyngeal flora of critically ill adults, that are also potential
microbial causative agents of pneumonia include - Staphylococcus aureus,
Streptococcus pneumoniae, Haemophilus influenzae, Pseudomonas aeruginosa and
Acinetobacter baumannii (Ewig et al., 1999; Fourrier et al., 1998; Munro & Grap,
2004). Aspiration of pathogenic microorganisms from the oropharynx whilst
intubated is a substantial risk factor and contributes to the development of VAP
(Bonten et al., 1994; Ewig et al., 1999; Fourrier et al., 1998; Garrouste-Orgeas et al.,
1997; Munro & Grap, 2004; Pugin, Auckenthaler, Lew et al., 1991). The risk factors,
progression and consequences of pathogenic colonisation of the oropharynx during
childhood critical illness have not been thoroughly described in previous research.
Hence the purpose of the Critically ill Children’s Oral Health (CCOH) study is to
describe the status of pathogenic colonisation of the oropharynx, within the context of
oral health, in critically ill children.
Traditionally, oral health and oral hygiene have been given low priority in the nursing
care of a critically ill child. Oral hygiene is often neglected or performed inadequately
by swabbing the children’s mouths for comfort. Currently, within the Paediatric
Intensive Care Unit (PICU) at the Royal Children’s Hospital (RCH) no oral hygiene
protocol exists to guide nursing practice. Previous researchers in critical care have
suggested that nursing practice surrounding oral hygiene is often based on tradition,
individual preferences, availability of products, anecdotal or subjective evaluation (A.
4
Berry et al., 2007; Gibson & Nelson, 2000; McNeill, 2000), rather than evidence-
based protocols. No evidence-based oral hygiene protocol has been researched,
validated and made available to guide clinicians when nursing critically ill children.
The aim of this study therefore is to contribute to the body of knowledge available
surrounding the oral health of critically ill children, so that this information can be
used in further research to develop evidence-based guidelines.
1.2 PURPOSE OF THE RESEARCH
The purpose of the CCOH study is to describe the status of oral health in critically ill
children. In addition, the study will examine the influence of time in the PICU, patient
characteristics and PICU therapies on the status of oral health. It will also explore the
link between poor oral health and PICU-acquired Healthcare-Associated Infections
(HAI). To meet this purpose, the study will answer the following questions:
1) What is the status of oral health in critically ill children during admission to a
PICU?
2) How does the oral health of critically ill children change during their admission to
PICU?
3) How is the oral health of critically ill children affected by patient characteristics or
PICU therapies?
4) What is the relationship between dysfunctional oral health in critically ill children
and PICU-related HAI?
5
1.3 SIGNIFICANCE OF THE STUDY
Previous research has determined that adults who are critically ill are at an increased
risk for developing systemic infections when they have poor oral health (Fourrier et
al., 1998; Munro, Grap, Elswick et al., 2006; Scannapieco et al., 1992). Since this risk
has been established, oral hygiene protocols have been developed to improve the oral
health of critically ill adults, endeavouring to decrease the risk of developing systemic
infections. However, current literature surrounding the oral health of critically ill
children is scarce. Before developing strategies to promote and improve oral health
during a child’s critical illness, baseline research needs to be undertaken to explore
the current state of oral health, the risk factors for developing poor oral health and the
subsequent effect on systemic health. The purpose of this study is to provide this
baseline research which may then be used to develop strategies which focus on
children with an increased risk for poor oral health and potential systemic
consequences. This study is significant as evidence-based guidelines can then be
developed and tested for efficiency and effectiveness in improving the oral and
systemic health of critically ill children.
1.4 SUMMARY
This thesis provides details of the study carried out at the PICU at the RCH, Brisbane,
which aimed to describe the oral health of critically ill children. It is structured over
six chapters to provide information surrounding the literature review and
methodological approach to support the study and the results, conclusions and clinical
implications of the study.
6
CHAPTER 2 - LITERATURE REVIEW
2.1 INTRODUCTION
This chapter reviews the existing literature surrounding the oral health of critically ill
children. It discusses components of oral health, critical illness in childhood, and
Healthcare-Associated Infections (HAI).
2.2 LITERATURE SEARCH STRATEGIES
A literature search was undertaken using the electronic databases of PubMed,
Medline, the Cochrane Library and the Cumulative Index to Nursing and Allied
Health (CINAHL) using MeSH headings and free text words. Articles that resulted
from the search were screened with the inclusion criteria of (1) oral health in critical
illness (2) oropharyngeal colonisation (3) human subjects (4) adult or child subjects
and (5) publication in English. The following terms were used individually and in
various combinations in the search process - oral health, oropharyngeal colonisation,
critical illness, critical care, intensive care, children, pediatrics, paediatrics, oral
hygiene, nosocomial, healthcare associated infection. The reference lists of published
materials were also hand searched for additional information. The aim was to identify
all relevant randomised controlled trials, observational studies as well as discussion
documents. Due to a lack of available research surrounding the oral health of critically
ill children, the search was widened to incorporate adult literature and literature dating
before 1999.
7
2.3 ORAL HEALTH
2.3.1 Physiology of the oral cavity
The oral cavity works in a state of equilibrium, functioning to aid in digestion,
communication and respiration and aims to protect the body from exogenous sources.
Structurally it is primarily made up of mucosal membranes - lips, gums, tongue,
buccal cheek, roof, and saliva, with teeth developing in early childhood (Munro,
Grap, Jablonski, & Boyle, 2006). The oropharynx of a healthy person is colonised by
greater than 250 different groups of microorganisms (O'Reilly, 2003) with the
majority of microorganisms living in symbiosis with the host, and are referred to as
commensal flora.
The acquisition of commensal oropharyngeal flora occurs within the first 18 hours
after birth (Kite & Pearson, 1995), dominated by viridans streptococci such as
Streptococcus oralis, Streptococcus mitis, and Streptococcus salivarius (Jenkinson &
Lamont, 2005; Kononen, 2000; Law, Seow, & Townsend, 2007; Marsh & Percival,
2006). The diversity of the oropharyngeal flora increases over time and between one
to seven months of age, Gram-negative anaerobes establish, with the most common
being Prevotella melaninogenica, Fusobacterium nucleatum and Veillonella spp
8
(Kononen, 2000; Kononen, Kanervo, Takala, Asikainen, & Jousimies-Somer, 1999;
Marsh & Percival, 2006).
Oropharyngeal flora change composition throughout childhood, with the eruption of
primary teeth having a great influence of the oral environment by providing suitable
niches for colonisation (Kononen, Asikainen, Karjalainen, & Jousimies-Somer, 1994).
This process leads to an increase in the number and complexity of the commensal
microflora in the oral cavity. Bacteria are the predominant commensal
microorganisms found in the mouth, and a diverse range of gram positive and gram
negative species can be isolated including Streptococcus sanguis, Staphylococcus sp.,
Veillonella sp., Neisseria sp., Actinomyces sp. and Lactobacillus sp. (Law et al.,
2007). In addition, yeasts, mycoplasmas and protozoa can occasionally be isolated
(Marsh & Percival, 2006). The commensal flora act directly as an important
component of the host defences by being a significant barrier to exogenous
populations (Marsh & Percival, 2006). The oropharyngeal microflora are not static,
but change in composition over age and in response to the individual’s physical state,
including their health (Roberts, 1998).
Saliva production has an important role in maintaining health and stability within the
oral cavity (Munro, Grap, Jablonski et al., 2006) as it contains several components
that are important in suppressing bacterial and fungal colonisation, serving both anti-
plaque producing and antibacterial roles (Kite & Pearson, 1995; Stiefel et al., 2000).
Saliva coats the mucous membranes, aids in digestion, regulates the pH of the mouth
and contains calcium and phosphate to maintain dental integrity (O'Reilly, 2003). In a
clean, healthy mouth, salivary proteins and glycoproteins form a film (pellicle) which
9
acts as a protectant for the teeth and contains immune components including
immunoglobulin A (IgA), immunoglobulin M (IgM) and lactoferrin, which protect the
oral cavity from potential pathogens (Munro & Grap, 2004; Munro, Grap, Jablonski et
al., 2006; O'Reilly, 2003). Saliva also provides mechanical removal of plaque and
microorganisms as it circulates the oral cavity (Munro, Grap, Jablonski et al., 2006).
During the day, salivary flow ranges from 0.25 to 0.35 millilitres per minute when
unstimulated, to four to six millilitres per minute when stimulated (A. Berry &
Davidson, 2006). When saliva production is reduced, its multiple protectant functions
are limited.
The commensal microflora of the mouth has a beneficial role as a barrier against the
colonisation of potentially pathogenic flora. However, some of the bacteria present in
the oral cavity can cause localised infection. The most common bacterial infections in
the mouth are dental caries (Kononen, 2000). Dental caries are primarily caused by
the species Streptococcus mutans and Streptococcus sobrinus. S. mutans has been
identified as the principal cariogenic bacterium for caries initiation while S. sobrinus
is thought to enhance progression and development of caries (Law et al., 2007).
Dental plaque is a biofilm found on tooth surfaces that provides an environment for
commensal and exogenous microorganisms. The combination of bacterial colonies,
polysaccharides and salivary proteins form the mixture of polymers known as plaque
(Kite & Pearson, 1995). Plaque provides opportunity for adherence of the organisms
to either the tooth surface or other microorganisms. Calculus occurs when minerals
are deposited intracellularly and extracellularly in dental plaque (Munro & Grap,
2004). Dental plaque, calculus and dental caries are oral conditions that affect healthy
10
children and adults, however in the critically ill, may potentially have systemic
consequences.
2.3.2 The ‘dysfunctional’ oral cavity
The oral cavity works in a fragile state of stability and when exogenous sources are
introduced, or the general health of the person changes, this balanced situation is
disturbed. Xerostomia is the severe reduction of saliva, manifested as a dry mouth and
can be caused by a variety of medications and Paediatric Intensive Care Unit (PICU)
therapies. Severe xerostomia is defined as an unstimulated salivary flow of less than
0.1 millilitres per minute (A. Berry & Davidson, 2006). When a patient is xerostomic,
increased levels of proteases in the mouth cause the removal of fibronectin from the
epithelial cell surface of the mucosal membranes (A. Berry et al., 2007). Fibronectin
is present on cell surfaces and acts as a reticuloendothelial-mediated host-defence
mechanism (A. Berry & Davidson, 2006). The depletion of fibronectin exposes oral
receptors and facilitates attachment of exogenous microorganisms, such as
Staphylococcus aureus, to mucosal membranes within the oropharynx. A build up of
dental plaque caused by a lack of oral hygiene can also serve as a reservoir for
pathogenic microorganisms in patients with poor oral hygiene (Fourrier et al., 1998;
Scannapieco et al., 1992). Oral mucositis - inflammation of the oral mucosa - is
characterised by erythema, oedema, bleeding, ulceration and pseudomembrane
formation within the mouth and is a painful and incapacitating side-effect of
chemotherapy and other medical conditions prevalent in critically ill children
(Donnelly, Bellm, Epstein, Sonis, & Symonds, 2003; Napenas, Brennan, Bahrani-
Mougeot, Fox, & Lockhart, 2007).
11
Quantifying ‘poor’ oral health is a challenge for clinicians and researchers. Munro et
al., (2006) propose that the measurement of saliva volume, dental plaque and oral
microbial flora provide a baseline for oral health measurement. However, the
logistical issues surrounding nursing assessment of stimulated and unstimulated saliva
volumes and dental plaque make it difficult to use this as a practical oral health
measurement.
Oral assessment tools have been developed and validated for use in multiple clinical
practice areas. Landmark work regarding the assessment of oral health by nurses was
primarily accomplished by Beck (1979). The oral assessment guide was originally
developed by Eilers et al. (1988) for clinical assessment of changes in the oral cavity
of individuals receiving cancer treatment and has established considerable evidence
for its validity and reliability in other clinical settings including the PICU (Andersson,
Persson, Hallberg, & Renvert, 1999; Barnason et al., 1998; Eilers, Berger, & Peterson,
1988; Ferozali, Johnson, & Cavangnaro, 2007; Holmes & Mountain, 1993; Jiggins &
Talbot, 1999). It involves assessment of five categories: lips, tongue, saliva, mucous
membranes/gingivae and teeth.
However oral assessment tools, including those by Beck (1979) and Eilers et al.,
(1988) do not incorporate the microbial changes occurring in the mouth. Considering
the relationship between microbial flora changes and oral health, this is an important
element of oral health to measure in research.
12
2.4 CRITICAL ILLNESS IN CHILDHOOD
2.4.1 PICU population
Children admitted to a PICU are a heterogeneous population of patients (Franklin,
Senior, James, & Roberts, 2000). Many children are pre-morbidly healthy, while
others have an acute episode related to a pre-existing condition. Critically ill children
sometimes require rescue-therapy in the form of complicated and invasive respiratory,
cardiovascular and renal support, while some simply require post-operative
monitoring.
The diversity in critically ill children is illustrated by the manner in which they are
admitted to the PICU and the conditions which require admission. Data retrieved from
the Australia and New Zealand Paediatric Intensive Care Registry (2008) reports that
nearly 44% of all PICU admissions in 2006 in Australia and New Zealand were
elective, that is the admission followed elective surgery, elective PICU procedure or
was required for elective monitoring (Alexander, Tregea, & Slater, 2008). The most
common diagnostic group was post operative (non-cardiac), with 24.3% requiring
admission to the PICU, followed by cardiovascular (23.3%), respiratory (22.7%),
miscellaneous (12.9%), neurological (7.6%), injury (7.3%), and gastrointestinal or
renal (1.9%) (Alexander et al., 2008). Depending on the PICU site, the percentage of
ventilated patients ranged from 4.8% to 85% (Alexander et al., 2008). The majority of
13
admissions to PICU were for children less than five years of age (65%), with infants
less than 12 months making up more than half this group (38.8% of all PICU
admissions) (Alexander et al., 2008). This diversity in age, severity, cause and
treatment of paediatric critical illness, may also be reflected in the oral health of
critically ill children.
Both the mechanism of critical illness and the baseline physiological function and
health of critically ill children are very different to critically ill adults. This is
demonstrated in their oral health. As described previously, oral health changes over
age, especially with the genesis of teeth and the resultant change in the microbial
environment. Within critical care, two examples of differences between adult and
paediatric patients are crude mortality rates and length of Intensive Care Unit (ICU)
stay. The length of stay for children admitted to ICU’s is varied, but universally
shorter than adults (ANZICS: Adult patient database, 2008), with the majority of
children staying less than two days (Alexander et al., 2008). The crude mortality rate
for children admitted to Australian and New Zealand PICUs in 2006 was 2.9%, with
neonates recording the highest age specific mortality (4.2%) (Alexander et al., 2008).
Adult ICUs have varying crude mortality rates, however they are collectively higher
than PICUs (ANZICS: Adult patient database, 2008; Fitch et al., 1999). Health
professionals and researchers agree that the universal application of adult research to
the paediatric population is not ideal given the differences between these two
populations. However, despite numerous differences between adults and children and
a consequential understanding that a majority of adult research should not be
transposed to a paediatric setting, adult research may well contribute to ideas,
thoughts and research methodologies with a paediatric focus. Additionally, while
14
adult research has the capacity to influence paediatric research within the area of oral
health research, blanket application of adult research to the paediatric population
should be regarded with caution.
2.4.2 Measuring critical illness in childhood
Children admitted to a PICU have widely varying levels of critical illness, and the
challenge lies with researchers to find a tool to quantify these levels for use in
research. Measuring critical illness effectively and reliably is important when
researching a diverse group, such as critically ill children. Critical illness scores allow
researchers to examine critical illness as a potential factor contributing to changing
oral health.
The Pediatric Logistic Organ Dysfunction Score (PELOD) (Lacroix & Cotting, 2005;
Leteurtre et al., 1999; Leteurtre et al., 2003; Santanae, Leite, de Carvalho, & Lopes,
2009; Thukral, Kohli, Lodha, Kabra, & Kabra, 2007; Yung, Wilkins, Norton, &
Slater, 2008) is a measure of severity of multiple organ dysfunction syndrome. It is
calculated for each patient by adding the scores for six individual organ systems based
on the recorded levels of the variables included in the systems, after stratifying into
age categories. The PELOD has been validated for repeated measures (Leteurtre et
al., 2003) showing the changing severity of organ dysfunction during an individuals’
critical illness.
While frequently used in other paediatric critical care studies (Lacroix & Cotting,
2005; Leteurtre et al., 1999; Leteurtre et al., 2003; Santanae et al., 2009; Thukral et
15
al., 2007; Yung et al., 2008), a major criticism of the PELOD score is that a majority
of the elements do not take into consideration the contribution of PICU therapies (e.g.
renal replacement) to improve scores. This means the PELOD score may be unfairly
low for patients who have early interventional therapies implemented by the PICU
treating team.
The Paediatric Index of Mortality 2 (PIM2) is a regression model that uses admission
data to predict intensive care outcome for children (Baghurst, Norton, & Slater, 2008;
Eulmesekian, Perez, Minces, & Ferrero, 2007; Inwald et al., 2009; Shann, Pearson,
Slater, & Wilkinson, 1997; Slater, Shann, & Pearson, 2002; Thukral, Lodha, Irshad, &
Arora, 2006; Wolfer, Silvani, Musicco, & Salvo, 2007). It was initially designed to
compare the standard of care between PICUs and within PICUs over time (Slater et
al., 2002), but is also commonly used to estimate mortality risk within groups of
patients in research (Slater et al., 2002). The PIM2 is a revision of the original PIM
(Shann et al., 1997) using expanded and updated data from 14 intensive care units:
eight in Australia, four in the UK and two in New Zealand.
The PIM2 model shows a Hosmer-Lemeshow goodness-of-fit of 8.14, with 8 df,
p=0.420 (Slater et al., 2002). The performance for the 14 ICUs showed the area under
the receiver operating characteristic plot ranged from 0.78 to 0.95 (Slater et al., 2002).
The ten variables in the PIM2 model were included only because they improved the
discrimination or calibration of the model (Slater et al., 2002). The model uses
physiological data, including systolic blood pressure and pupillary responses and
patient data, including admission to PICU, for the purposes of estimating recovery
post procedure, and presence of high risk/low risk diagnosis, to generate an overall
16
risk-score. However, the PIM2 is reflective of the patients’ mortality risk on
admission to PICU only, and is not designed to be used to describe individual
patients.
2.4.3 The effect of PICU therapies on oral health
Critical illness and admission to PICU require life-saving supportive therapies which
sometimes have an adverse effect on components of oral health. Most critically ill
children are restricted in their fluid intake to maximise respiratory, renal and cardiac
function (Jiggins & Talbot, 1999; Kite & Pearson, 1995), which then reduces salivary
flow. Clinical conditions which cause dehydration, such as fever, diarrhoea, and
burns, also cause a reduction of salivary volume (O'Reilly, 2003). Several
medications which are commonly used in PICU can also lead to diminished salivary
production including anticholinergics, diuretics, antiemetics, opioid analgesics and
retinoids (Kite & Pearson, 1995). Any interruption in saliva production due to
dehydration or drug therapy, can lead to infection of the salivary glands by the
oropharyngeal flora (Kite & Pearson, 1995) and alter the ability of the oral cavity to
respond to the introduction of exogenous, and potentially pathogenic microorganisms
(Munro, Grap, Jablonski et al., 2006).
Other drugs are also associated with alterations in the physiology of the mouth.
Phenytoin, a common anti-seizure medication, is associated with swelling and
hypertrophy of the gingivae (Kite & Pearson, 1995). Antibiotics are associated with a
suppression in commensal bacterial numbers in the mouth (Sixou, Medeiros-Batista,
& Bonnaure-Mallet, 1996), allowing for an overgrowth by previously minor drug-
17
resistant components of the microflora, or colonisation by exogenously-acquired (and
often pathogenic) micro-organisms (Marsh & Percival, 2006). Patients receiving
radiotherapy, chemotherapy or high doses of antineoplastic drugs are at very high risk
of oral mucositis which manifests as ulceration, xerostomia and secondary infection
of the oral cavity (White, 2000).
An unconscious or intubated child is unable to speak, eat or drink, which limits the
production and movement of saliva around the mouth (Kite & Pearson, 1995). The
endotracheal tube (ETT) presents a source of pressure for the oral cavity (Barnason et
al., 1998), especially in non-dentate patients where the tube rests on the gingivae
rather than on the teeth (D. Jones & Munro, 2008). The orally intubated patient is
forced to keep their mouth open for extended periods. Even when not intubated, the
patients are often exposed to high flow facial oxygen and oral suctioning causing
drying of the mucosa. Critically ill children regularly receive a combination of these
therapies and develop these clinical conditions, which may potentially have an effect
on their oral health.
2.4.4 The oral health of the critically ill
Critically ill children are at high risk of developing poor oral health because of their
clinical condition and the therapies instituted or continued within the PICU. However,
studies describing the oral health of this group are scarce. Franklin et al. (2000)
conducted an observational cohort study examining the oral health status of children
in the PICU and measured oral health by plaque accumulation, gingival inflammation
and gingival bleeding. The study involved 54 participants, however excluded all non-
18
dentate patients (mean age 4.8 years ± 4.3) which, as previously described, excludes
the majority of patients admitted to a PICU. Measurements of oral health were only
conducted on admission and discharge, and the outcomes were primarily focused on
investigating the efficacy of the oral care that they received while admitted to PICU,
rather than the participants’ oral health throughout their critical illness. They found a
statistically significant increase in mean plaque scores during PICU stay (admission =
22.5; discharge 25.8; p=0.001) and gingival inflammation (admission = 4.1; discharge
5.5; p=0.006). However, they did not examine oropharyngeal colonisation during
PICU admission or the relationship between the children’s oral health and their
clinical condition or PICU therapies.
Previous research authored by Jiggins and Talbot (1999) involved the implementation
a mouth care guideline within their PICU, however they did not carry out any baseline
or post implementation oral health measurement. Within their prospective descriptive
study to evaluate the frequency and route of endotracheal colonisation of intubated
children, Rubenstein et al (1992) described the colonisation of the buccal mucosa
(cheek). Only recruiting patients who were to be intubated for greater than five days,
and excluding patients less than one month of age, restricted their sample size to 19
patients. The pathogens that they cultured from the buccal mucosa included Candida
albicans and other Candida species, Enterococcus sp., Staphylococcus aureus,
Pseudomonas sp., Acinetobacter sp., Klebsiella sp., Haemophilus influenzae,
Enterobacter sp. and Escherichia coli. Patients who were colonised with Candida sp.
were sicker (Pediatric Risk of Mortality: 12.9 ± 2.8 vs. 4.9 ± 1.0; p=0.01). The study
did not examine the oral health of the children within their critical illness, and was
limited due to its poor sample size and the inclusion of patients who only received
19
prolonged intubation and ventilation, which is not descriptive of the entire PICU
population.
Recently, Thorburn et al. (2009) compared the prevalence of ‘abnormal’ and
antibiotic-resistant bacterial flora in children with cerebral palsy (CP) (n=53)
requiring mechanical ventilation, to critically ill children without CP (n=257). Within
their prospective, cohort study, they found that 89% of children with CP, carried
abnormal bacterial flora or potential pathogens, compared to 55% of children without
CP (RR=2.09 (CI 1.76-2.48)). Pseudomonas, Klebsiella, Enterobacter and
Citrobacter species were the most common potential pathogens. Other oral health
measurements were not collected and an analysis of the effect of other clinical
characteristics and PICU therapies was not carried out.
The majority of existing studies available describing the oral health of critically ill
patients are based on adult critical care. These studies provide paediatric critical care
providers with information to develop research and guide clinical practice where
paediatric research is not available. For example, landmark work was accomplished
by Scannapeico et al. (1992) when they conducted a case-control study describing the
prevalence of dental plaque and colonisation of that dental plaque by potential
respiratory pathogens. Thirty-four critically ill adults were compared to 25 healthy
patients who attended a preventative dentistry clinic. Scannapeico et al. (1992) found
a statistically significant higher mean plaque score for critically ill participants (1.9 ±
0.2) than the control group (1.4 ± 0.1; p=<0.005). They also found that 64.7% of ICU
patients were colonised in their teeth or oral mucosa by respiratory pathogens in
comparison to 16% of the control (p= <0.005). The respiratory pathogens colonised
20
included Staphylococcus aureus, Pseudomonas aeruginosa, Klebsiella pneumoniae,
Serratia marcescens and Escherichia coli.
These findings have been confirmed in other studies. Fourrier et al. (1998) examined
the dental status, prevalence of pathogenic colonisation in dental plaque and dental
plaque as a source of Healthcare-Associated Infection (HAI) in adult ICU patients. In
their prospective observational cohort study of 57 participants, they found a
statistically significant increase in dental plaque on patients remaining in ICU for five
days or greater (day 0: 1.1 ± 0.7; day 5: 1.6 ± 0.7; p<0.05). Dental colonisation by
respiratory pathogens showed an increasing trend with length of admission, however
due to the small sample size it did not reach statistical significance. Fourrier et al.
(1991) cultured similar pathogens to the study carried out by Scannapeico et al.
(1992), but also discovered Candida albicans. A positive bacterial plaque sample on
admission to the ICU showed a sensitivity of 0.47 and specificity of 0.92 (p<0.005) of
developing a HAI.
In an observational cohort study, Munro et al., (2006) examined the oral health status
of 66 critically ill adults through assessment of the oral cavity using a visual analogue
scale, culture of an oral specimen, measurement of salivary volume and analysis of
salivary immune components; IgA and lactoferrin. The association between oral
health and Ventilator-Associated Pneumonia (VAP) was also determined using the
Clinical Pulmonary Indicator Score (CPIS). They found a trend of worsening oral
health over time in ICU, evidenced by an increasing amount of plaque and a reduction
in salivary flow, however it did not reach statistical significance (Munro, Grap,
Elswick et al., 2006). Salivary volume did decrease in volume over length of ICU stay
21
(p=0.003). There was no discussion on specific organisms present on oral or tracheal
cultures (Munro et al. 2006).
The change in oropharyngeal colonisation over time spent in adult ICU was also
described by Garrouste-Orgeas et al., (1997). Their descriptive study found acquired
colonisation of the oropharynx by systemic bacterial pathogens such as Acinetobacter
baumannii, Klebsiella Pneumoniae, Enterobacteriaceae, Pseudomonadaceae, S.
aureus, Enterococcus spp. in 44 of 86 (51%) patients. The median time for
oropharyngeal colonisation was seven days.
Baseline oral health research is necessary in paediatric critical care in order to inform
practice surrounding the provision of oral hygiene. Despite the dearth of paediatric
literature in this area, to rely on adult research to inform paediatric practice would be
problematic. Differences between adult and paediatric populations lie not only in
their oral health, but also the nature and duration of their critical illness.
2.4.5 Current oral hygiene practices
Critically ill children are dependent on healthcare workers to provide all aspects of
their oral care. However, a number of authors suggest that oral care regimens are
often based on tradition, individual preferences, availability of products, anecdotal
and subjective evaluation rather than evidence-based protocols (A. Berry et al., 2007;
Gibson & Nelson, 2000; McNeill, 2000). Within the PICU at the Royal Children’s
Hospital (RCH), Brisbane, there is no standard or protocol to guide nursing practice
for oral care. In recent years, researchers have focused on developing evidence-based
22
protocols for the oral care of critically ill adult patients (Cutler, 2005; Ferozali et al.,
2007; Ross & Crumpler, 2007; Stiefel et al., 2000), however few have been proven to
be effective (Chan et al., 2007).
Oral health is assisted by mechanical and chemical cleansing. The use of a toothbrush
to clean the teeth and gums has been widely accepted as a simple and efficient method
of mechanically removing plaque and debris (Kite & Pearson, 1995; O'Reilly, 2003).
Despite these recommendations, many dentate patients in PICUs have their oral care
carried out with the use of foam or cotton swabs alone, rather than with toothbrushes
(Binkley et al., 2004; Costello & Coyne, 2008; Rello et al., 2007a). A large
percentage of critically ill children are non-dentate and protocols incorporating this
into guidelines are scarce. The physical design of cotton swabs suggests that they may
break-down while in the mouth, leaving cotton filaments which may then be micro-
aspirated.
There is a large variety of chemicals used in the prevention and treatment of poor oral
health. Chlorhexidine gluconate mouthwash is an anti-plaque agent which maintains
oral health through its ability to suppress overgrowth with Gram-positive and Gram-
negative bacteria as well as yeasts, making it useful in the prevention of the
development of plaque (A. Berry et al., 2007; Kite & Pearson, 1995; O'Reilly, 2003).
Chlorhexidine fixes to the surface of oral structures and is slowly released for up to 24
hours, requiring infrequent applications. A large volume of data is available
surrounding the use of chlorhexidine for oral decontamination, which has been
summarised in a recent meta-analysis by Chan (2007). Seven randomised controlled
trials totalling 2144 patients, using chlorhexidine in varying forms (0.12%, 0.2%, 2%)
23
showed a significant reduction in oropharyngeal colonisation (RR 0.46, 0.39 to 0.81;
p=0.002; I2=48.2%), however a majority of these trials have taken place in adult
cardiothoracic Intensive Care Units, so the application of such research to general
critical care (Chan et al., 2007), and specifically the PICU, is questionable.
There is scarce literature surrounding the safety and effectiveness of other solutions
including sodium bicarbonate, toothpaste, normal saline, tap water, and sterile water
(A. Berry et al., 2007) as oral cleansers. An early study examining hydrogen peroxide
found it was effective in improving the condition of oral mucosa (Passos & Brand,
1966), however subsequent research found it to be associated with mucosal
abnormalities and poor subjective reports by participants (Tombes & Galluci, 1993).
Lemon and glycerol has been associated with xerostomia and erosion of the tooth
enamel (A. Berry et al., 2007; Stiefel et al., 2000). The optimal frequency of oral
hygiene provision is also in dispute, with recommended frequencies ranging from
hourly, to twice daily (A. Berry et al., 2007; Chan et al., 2007; Grap, Munro,
Elswick, Sessler, & Ward, 2004; Tombes & Galluci, 1993). Small sample sizes,
subjective measurements, lack of adequate controls and poor interventional control is
present in the majority of research surrounding the provision of oral hygiene, which
makes it difficult to confidently apply these in the clinical setting.
The only published study available regarding the institution of an oral health guideline
on PICU patients was conducted by Jiggins and Talbot (1999). They developed a
guideline consisting of regular oral assessment which directed the use of different
products including toothbrush and toothpaste, Nystatin (an oral anti-fungal
medication), 3% hydrogen peroxide diluted 50:50 with water, and an artificial saliva
24
spray. They stated that there was some controversy with the choice of solutions and
implements, but did not provide any information regarding the effectiveness of the
protocol. Considering the lack of research to support elements within the guidelines
and the effectiveness of the guideline as a whole, it is not appropriate to institute this
as an evidence-based protocol.
The application of oral health research requires the cooperation and adoption by
clinical staff. Descriptive studies conducted by Fitch et al., (1999), Rello et al., (2007)
and Binkley et al., (2004) surveyed intensive care nurses throughout America and
Europe. They identified that while nurses recognise the importance of oral hygiene
practices, they may be hesitant to provide oral care to patients who are intubated
because ETTs may limit access to the oral cavity and they fear dislodging or
displacing the tube. The adult critical care nurses did not rate it as a high priority or as
essential when taking care of critically ill patients. The majority of those surveyed did
not currently follow an evidence-based pathway for oral care. However, none of these
studies surveyed PICU nurses, so it is not necessarily reflective of current PICU
practice or PICU nursing attitudes.
Although the provision of oral hygiene in the intensive care has been recognised as
being a priority by international health-care bodies, including the Centers for Disease
Control and Prevention (CDC) (Tablan, Anderson, Besser, Bridges, & Hajjeh, 2003),
there is a paucity of literature surrounding the effectiveness and appropriateness of
oral hygiene practices in adult and paediatric intensive care.
25
2.5 HEALTHCARE ASSOCIATED INFECTION (HAI)
HAI’s are a significant cause of mortality and morbidity for critically ill children
(Inwald et al., 2009; Rubenstein et al., 1992; Safdar et al., 2005; Singhi et al., 2008;
Suljagic et al., 2005; Thorburn et al., 2009; Turton, 2008), and were previously
known as ‘nosocomial infections’ (Garner, Jarvis, Emori, Horan, & Hughes, 1996).
As part of the National Healthcare Safety Network for the Centers for Disease Control
and Prevention (CDC), Horan, Andrus, & Dudeck (2008, pg. 309) define HAI’s as “a
localised or systemic condition resulting from an adverse reaction to the presence of
an infectious agent(s) or its toxin(s) . There must be no evidence that the infection was
present or incubating at the time of admission to the acute care setting”.
Within the scope of this study, interest is in infections acquired within the PICU,
which may be inferred as being a result of a patient’s clinical condition or their care
during admission in the PICU. A HAI is considered acquired in the PICU if the
clinical condition fulfilling the criteria develops after the patient was in the PICU for
two days and within two days after discharge (Horan et al., 2008). The most common
and clinically significant HAI within the PICU population are pneumonia (Schleder,
2003; Turton, 2008) and blood-stream infection (D. Jones & Munro, 2008) .
2.5.1 Pneumonia
26
Within adult critical care research, the relationship between poor oral health in the
form of pathogenic oropharyngeal colonisation, and hospital-acquired pneumonia,
especially VAP, has been well documented in robust clinical research (Bonten et al.,
1994; Ewig et al., 1999; Fourrier et al., 1998; Garrouste-Orgeas et al., 1997; Munro &
Grap, 2004; Pugin, Auckenthaler, Lew et al., 1991). The pathogenic microflora that
have been isolated in the dental and oropharyngeal flora of critically ill adults, that are
also potential microbial causative agents of VAP include Staphylococcus aureus,
Streptococcus pneumoniae, Hemophilus influenzae, Pseudomonas aeruginosa and
Acinetobacter baumanni (Ewig et al., 1999; Fourrier et al., 1998; Munro & Grap,
2004).
Pneumonia and VAP remain leading causes of morbidity and mortality among
mechanically ventilated adults (Schleder, 2003; Turton, 2008), with an incidence
ranging from 9-27%, crude mortality that may exceed 50% and estimated costs of
greater than $10,000-$20,000 per incident (Chan et al., 2007; Safdar et al., 2005).
Ventilator-associated pneumonia (VAP) has been difficult to define, however there is
general consensus on defining it as a pneumonia that develops 48 hours or more after
intubation with an endotracheal or tracheostomy tube, and was not present before
intubation (Feider, 2007; Garner et al., 1996; Munro, Grap, Elswick et al., 2006;
Pugin, Auckenthaler, Miller et al., 1991; Safdar et al., 2005). VAP is a multifaceted
pathophysiological phenomenon that develops when normal pulmonary or host
defence mechanisms are either overwhelmed or impaired by endotracheal tube
insertion and mechanical ventilation (Feider, 2007; Horan et al., 2008; Rello et al.,
2007a). Pneumonia is an inflammatory response to the presence and multiplication of
microorganisms that have invaded the lower respiratory tract (Feider, 2007). The
27
Clinical Pulmonary Infection Score (CPIS) has been used by multiple researchers
(Aydogdu & Gursel, 2008; McClure et al., 2009; Munro, Grap, Elswick et al., 2006;
Sakaquicki, Shime, S., & Hashimoto, 2008; Singh, Gayowski, Wagener, & Marino,
1999) to measure clinically evident pneumonia and VAP. The clinical diagnosis
combines body temperature, white blood cell count, appearance of tracheal secretions,
oxygenation requirements, infiltrates on chest x-ray and cultures of tracheal aspirate
to give a cumulative score. Using a CPIS of greater than six as clinical definition of
pulmonary infection in ventilated patients has shown a strong correlation (r = 0.84; p
<0.001) (Pugin, Auckenthaler, Miller et al., 1991) with the ‘gold standard’ of
quantitative culture of either bronchoscopic or non-bronchoscopic broncho-alveolar
lavage fluid, with sensitivity of 73-93% (A'Court et al., 1993) and a specificity of 84-
100% (A'Court et al., 1993).
In PICU research, there is a paucity of research available regarding VAP and its
relationship with poor oral health and pathogenic oropharyngeal colonisation.
Rubenstein et al., (1992) established a link between pathogenic colonisation of the
buccal mucosa and colonisation of the ETT, however they did not measure the
incidence of pneumonia within the sample. A recent literature review compiled by
Turton (2008) regarding VAP in PICU, comprised primarily adult critical care data.
Their extrapolation suggests the institution of adult-designed interventions, however
baseline data examining the prevalence of VAP and its relationship to poor oral health
does not exist in paediatrics.
2.5.2 Bacteraemia and blood-stream infections
28
In the highly vascular environment of the oral cavity, tissue destruction, such as
ulceration, can provide direct communication of pathogenic microorganisms to the
circulatory system (Kite & Pearson, 1995; Stiefel et al., 2000). There is substantial
evidence that bacteraemia occurs in healthy populations with manipulation of the oral
mucosa, such as with tooth brushing (D. Jones & Munro, 2008). Procedures that
manipulate the oral mucosa in critically ill children are frequent, including intubation,
oral suctioning and oral hygiene practices. Bloodstream infection represents the
failure of the immune system to contain infection at a focal site and the consequent
occurrence of disseminated disease.
Bacteraemia has been defined as the presence of viable bacteria within the circulating
blood (Daly, Mitchell, Grossberg, Highfield, & Stewart, 1997; D. Jones & Munro,
2008) and can result in blood stream infection. Blood-stream infection is a frequent
and challenging HAI and accounts for 15% (D. Jones & Munro, 2008) of HAI’s. In
the adult population, the crude mortality rate of hospital-acquired blood stream
infections is approximately 35%, with a range from 12% to 80% (D. Jones & Munro,
2008). Blood-stream infections have been shown to increase ICU and hospital stay,
the use of healthcare resources and significantly increase morbidity and mortality (D.
Jones & Munro, 2008).
In accordance with the CDC (Horan et al., 2008), within this study the diagnosis of
blood-stream infection is established with presence of the following criteria: a
recognised pathogen cultured from one or more blood cultures, where the organism
cultured from the blood is not related to an infection at another site (Horan et al.,
2008). If the pathogen is a common skin contaminant (e.g. S. epidermis), the
29
participant must have at least one of the following signs or symptoms; temperature
>38oC or <36oC (if less than one year old), hypotension, chills or bradycardia or
apnoea (if greater than one year old).
Little research exists around the prevalence of bacteraemia or blood-stream infections
caused by microorganisms originating in the oropharynx in paediatrics. In adults,
there have been increasing rates of bloodstream infections related to organisms such
as Staphylococcus aureus, Enterococcus sp., coagulase-negative Staphylococci and
Candida sp. (Hugonnet, Sax, Eggimann, Chevrolet, & Pittet, 2004; D. Jones &
Munro, 2008; Suljagic et al., 2005; Zolldann, Thiex, Waitschies, Lutticken, &
Lemmen, 2005), all of which are known to colonise the oropharynx in times of critical
illness.
2.6 CURRENT LIMITATIONS IN THE LITERATURE
Research describing the oral health of critically ill children is limited to a small group
of paediatric studies and extrapolations from adult literature. A study by Franklin et
al., (2000), within the PICU, excluded all non-dentate patients - who make up a large
percentage of critically ill children - and did not describe oropharyngeal colonisation.
They also did not examine the diversity of oral dysfunction caused by variations in
patient characteristics or PICU therapies. Jiggins and Talbot’s (1999) study did not
measure oral health, but discussed it as being poor and requiring an evidence-based
guideline for intervention. The study by Rubenstein et al., (1992) describing the
relationship between ETT colonisation and buccal mucosa flora was limited to a small
sample (n=19), all of whom were intubated for greater than five days, which is not
30
reflective of all critically ill children. They also did not measure oral health as a
whole. Thorburn et al’s (1999) study did not collect information regarding oral health
measurement outside of oropharyngeal flora, and only examined children who were
mechanically ventilated. Additionally, extrapolation from adult literature, while useful
to develop practice and research methodologies, should be universally applied with
caution considering both the differences in baseline oral health and the mechanism of
critical illness between the two groups.
There is a lack of research regarding the prevalence and consequences of VAP and its
relationship with poor oral health and pathogenic oropharyngeal flora in critically ill
children. There is also a paucity of research regarding the prevalence and
consequences of bacteraemia and blood-stream infections caused by microorganisms
originating in the oropharynx in paediatrics. Considering the mortality, morbidity and
economic consequences of VAP and blood-stream infection in the adult ICU, this
needs to be addressed.
There is also a lack of information regarding current PICU practice surrounding the
completion of oral hygiene. Adult literature has reported a gap between current
practice and evidence-based guidelines, however there is no information available
regarding paediatrics.
Research regarding all elements of the oral health of critically ill children, systemic
consequences of poor oral health in PICU and current oral hygiene practices in PICU
are scarce and poorly completed. Considering the significance and consequences of
31
the studies completed in adult ICU on the same topic, research in the oral health of
critically ill children would appear necessary.
2.7 SUMMARY: WHAT THE LITERATURE SUGGESTS
In summary, the literature provides clear direction for future PICU oral health
research.
The physiology of the oral cavity functions in fragile stability and disturbances
by exogenous or internal sources may cause imbalances in the commensal
flora.
Critically ill children receive therapies and have clinical conditions which may
alter the health of the oral cavity.
The scant literature addressing paediatric oral health suggests paediatric oral
health may worsen or change over admission to a PICU.
Adult literature has established that during critical illness oral health worsens
and the commensal oropharyngeal flora can change to potentially pathogenic
flora.
Within adults, pathogenic oropharyngeal and dental flora is significantly
associated with HAI including pneumonia.
VAP and blood-stream infections are associated with significant economic
costs and mortality and morbidity outcomes.
There are no published evidence-based guidelines for the provision of oral
hygiene in PICU.
In the adult ICU, nurses do not rate oral hygiene as being a critical element of
nursing care.
32
CHAPTER 3 - METHODS
3.1 INTRODUCTION
This study was designed to describe the status of oral health in critically ill children
and to examine the influence of time in the Paediatric Intensive Care Unit (PICU),
patient characteristics and PICU therapies on the status of oral health. It also explored
the link between poor oral health and PICU-acquired Healthcare-Associated
Infections (HAI). To fulfil these aims a prospective observational cohort design was
used. This chapter describes the research design, research questions, population and
sample, instruments, data analysis and ethical issues in the study.
3.2 RESEARCH DESIGN
The Critically ill Children’s Oral Health (CCOH) study used a prospective
observational cohort design (Bhopal, 2008; Friis & Sellers, 2009). The study’s
purpose was to describe the status of oral health in critically ill children which was
supported by the cohort design. In addition, the study examined the influence of time
in the PICU, patient characteristics and PICU therapies on the status of oral health. It
also explored the link between poor oral health and PICU-acquired HAI. The
prospective cohort design was chosen to support these research questions as it
provides strong evidence of association in comparison to other observational designs
(Bhopal, 2008).
33
An observational cohort design is a epidemiological study design commonly used
throughout paediatric (Bracken et al., 2009; Edwards, Powell, Mason, & Oliver, 2009;
Franklin et al., 2000; Sullivan et al., 2005; Veldman, Trautschold, Weiss, Fischer, &
Bauer, 2006), critical care (Edwards et al., 2009; Fourrier et al., 1998; Franklin et al.,
2000; Munro, Grap, Elswick et al., 2006; Veldman et al., 2006) and oral health
(Fourrier et al., 1998; Franklin et al., 2000; Munro, Grap, Elswick et al., 2006)
research. The goal of analytical observational study designs is to add to a body of
knowledge, test hypotheses, generate new hypotheses and suggest mechanisms of
causation (Friis & Sellers, 2009). The prospective cohort design supports the research
objectives within this study as it facilitates exploration, description and causation
(Aschengrau & Seage, 2008; Friis & Sellers, 2009). It also provides information about
the characteristics of a study population over a period of time.
3.3 RESEARCH QUESTIONS
1) What is the status of oral health in critically ill children during admission to a
PICU?
2) How does the oral health of critically ill children change during their admission to
PICU?
3) How is the oral health of critically ill children affected by patient characteristics or
PICU therapies?
4) What is the relationship between dysfunctional oral health in critically ill children
and PICU-related HAI?
34
3.4 STUDY SETTING AND POPULATION
The study was conducted at the PICU at the Royal Children’s Hospital (RCH),
Brisbane in Australia. This unit provides an eight bed tertiary level PICU for children
of Queensland, Northern New South Wales and adjacent states and countries. There
are 600-700 admissions per year, of which 40% require mechanical ventilation
(Alexander et al., 2008). All modes of intensive supportive care are provided
including nitric oxide therapy, continuous veno-venous haemo-dialysis (CVVHD),
high frequency oscillation ventilation but excluding extracorporeal membrane
oxygenation (ECMO), left ventricular assist device (LVAD), and cardiac bypass
surgery. A tertiary PICU is a specialist referral centre for children needing intensive
care and provides complex, multi-system life support for an indefinite period for
children less than 16 years of age (Joint Faculty of Intensive Care Medicine, 2003).
The population under investigation was critically ill children admitted to the PICU at
the RCH. Children admitted to the PICU are a heterogeneous population of patients
(Franklin et al., 2000) experiencing physiological instability requiring intensive
support. They are referred from many specialities, in particular respiratory medicine,
neurosurgery, orthopaedic, surgery and oncology/haematology (Franklin et al., 2000).
Admissions are sourced from multiple locations; specifically the RCH Department of
Emergency Medicine, RCH inpatient acute-care wards, operating theatres and outside
hospital referrals via the RCH retrieval service. Diagnoses include respiratory
illnesses such as bronchiolitis, multi-traumas including severe head injury, and
oncological conditions requiring multi-system support.
35
3.5 SAMPLING STRATEGY AND SIZE
All patients admitted to the PICU at the RCH, Brisbane within a seven month period
were screened for potential participation by the principal investigator or trained PICU
personnel. After satisfying the inclusion/exclusion criteria, informed consent was
obtained by the principal investigator or trained PICU personnel. A consecutive
sample was then recruited to the study.
Inclusion Criteria:
Admission to the PICU at the RCH, Brisbane,
Clinical condition suggesting a PICU stay greater than or equal to 48 hours
(Fourrier et al., 1998),
Recruitment, first oropharyngeal saliva sampling and oral assessment within
12 hours of admission to PICU (Fourrier et al., 1998; Scannapieco et al.,
1992),
All ages, that is both dentate and non-dentate,
Parental or guardian consent and youth assent where required.
Exclusion Criteria:
Patients who had undergone oral surgery or had an oral condition that required
specialised oral care (Fitch et al., 1999) for example; cleft palate repair or
oropharyngeal abscess,
Already participated in the CCOH study on a prior admission to PICU,
Parents or guardian unavailable or unable to give consent due to legal
concerns that were under the care of Child and Family Services, or lack of
English.
36
The sampling framework consisted of the time period (seven months) and the
inclusion/exclusion criteria. The sample size was constrained due to the scope of a
master’s study and budget.
3.6 DATA COLLECTION
In order to answer the previously described research questions, data were collected on
the dependent and independent variables. The dependent variables used to describe
oral health within this study were pathogenic colonisation of oropharyngeal flora and
the Oral Assessment Scale (OAS) (Barnason et al., 1998; Eilers et al., 1988; Gibson &
Nelson, 2000; Jiggins & Talbot, 1999; Ross & Crumpler, 2007) The independent
variables collected were demographic, clinical and condition characteristics. A
number of instruments were used to describe the status of oral health, clinical and
condition characteristics and to extract this information.
3.6.1 Oropharyngeal cultures
The definition of ‘pathogenic’ colonisation of the oropharynx used in the CCOH
study excludes bacteria which are considered commensal in a paediatric mouth. As
with previous studies (Fourrier et al., 1998; Rubenstein et al., 1992; Scannapieco et
al., 1992; Thorburn et al., 2009) conducted within paediatrics and adults, ‘pathogens’
are listed as:
Acinetobacter spp.
Aspergillus spp.
37
Candida spp.
Citrobacter diversus
Enterobacteriaceae spp.
Escherichia coli
Haemophilus influenzae
Klebsiella pneumoniae
Pseudomonas spp.
Serratia marcescens
Staphylococcus aureus
Stenotrophomonas maltophilia
Streptococcus pneumoniae
These are microorganisms that commonly cause infection or disease in critically ill
children. The micro-organisms cultured from the oropharyngeal saliva swabs were
categorised into ‘nil flora’, ‘commensal flora’ or ‘pathogenic flora’ as per these
definitions.
Saliva sampling
Participants had samples of oropharyngeal saliva collected for bacterial and fungal
culture completed within 12 hours of admission to PICU and then second daily for the
course of their admission to PICU. Fourrier et al (1998) established that salivary,
dental plaque and tracheal aspirate colonisations are closely linked, and as the
sampling was to be performed by nurses, this was judged to be the least invasive, but
most accurate method of sampling. Oropharyngeal saliva sampling was collected
38
within 12 hours of admission to PICU as base-line data. A similar restriction has
previously been used by several authors (Fourrier et al., 1998; Munro, Grap, Elswick
et al., 2006; Scannapieco et al., 1992).
Other methodologically similar studies completed in an adult critical care setting
involved oropharyngeal, or equivalent, sampling on admission and then daily
(Rubenstein et al., 1992), second daily (Pugin, Auckenthaler, Lew et al., 1991), third
daily (Munro, Grap, Elswick et al., 2006; Scannapieco et al., 1992), every five days
(Fourrier et al., 1998) and weekly (Sixou, Medeiros-Batista, Ganddemer, &
Bonnaure-Mallet, 1998). The Centers for Disease Control and Prevention (CDC)
suggest that ‘most’ bacterial HAI’s become evident 48 hours or more after admission,
this being their typical incubation period (Garner et al., 1996; Horan et al., 2008).
Therefore, in order to collect the most accurate information regarding HAI,
oropharyngeal saliva sampling was collected every 48 hours.
Oropharyngeal salivary samples were obtained using a TRANSWAB® by the
participants’ PICU nurse. Nurses were trained individually by the principal
investigator and information was placed at the bedside regarding the saliva sampling
protocol. The TRANSWAB® was placed in the patient’s mouth for a period of at
least 30 seconds to allow for saliva absorption, as per the manufacturer’s instructions.
Swabs were collected at approximately the same time (1000hrs), at least two hours
after oral hygiene care or oral intake (feeds, medications, diet) (Sixou et al., 1998).
Samples were then transported to the Queensland Health Pathology Services
Laboratory within one hour of collection, for semi-quantitative analysis.
39
Laboratory protocol
The oropharyngeal saliva swabs were placed into five millilitres of 0.85% saline and
vortexed for 30 seconds. Both the undiluted extract and a 106 dilution were cultured.
1uL of the undiluted extract was then plated onto the following media:
Horse Blood Agar (HBA) incubated at 35oC in Carbon Dioxide (CO2) for 48
hours
MacConkey’s Agar (MAC) incubated at 35oC in Oxygen (O2) for 48 hours
Chocolate Bacitracin Agar incubated at 35oC in CO2 for 48 hours
Sabouraud Dextrose Agar (SAB) incubated at 35oC in O2 for 48 hours
Mannitol Salt Agar (MSA) incubated at 35oC in O2 for 48 hours
These plates were chosen by a panel of microbiologists, including the Director of
Microbiology at the Royal Brisbane and Women’s Hospital (RBWH), in order to
confirm the presence of the previously identified ‘pathogenic’ organisms.
Representative colonies of all morphologic types present on the plates up to 48 hours
after inoculation were subcultured and identified by standard methods (Scannapieco et
al., 1992). Staphylococci were differentiated based on colony characteristics, mannitol
fermentation and the coagulase reaction (Scannapieco et al., 1992). No attempts were
made to identify other bacteria, legionella, mycoplasma or viruses.
40
The following procedure was used to formulate the 106 dilution from the previously
vortexed material:
50 microlitres into five millilitres of saline (102)
50 microlitres of 102 into five millilitres of saline (104), then 10 uL of this 104
dilution were plated onto the same media as described above.
Cultures were then examined at 24 and 48 hours incubation. The commensal or
pathogenic flora were identified and semi-quantitated using the count table provided
(see Table 3.1).
Table 3.1 Count table for semi-quantitated analysis
Screen Colony forming units on any medium (cfu)
Count reported
Undiluted screen
>100
11-99
1-10
>105
count x 103
count x 103
Diluted screen
>100
10-100
1-10
>109
count x 106
count x 106
The diagnosis of colonisation by pathogenic organisms is based on the positive
culture of the oropharyngeal saliva swab when greater than or equal to 106 colony
forming units (cfu) per litre without signs of clinical infection (Fourrier et al., 1998).
41
3.6.2 Instruments
Oral Assessment Scale (OAS)
In addition to oropharyngeal colonisation, the status of oral health during critical
illness in childhood was described using an oral assessment instrument. Few oral
assessment instruments have been specifically designed or adapted for intubated,
incommunicative or paediatric patients. Landmark work regarding the assessment of
oral health was accomplished by Beck (1979) in developing an oral assessment tool
which has been modified and improved by multiple authors in order to validate it for
different adult and paediatric population groups (Andersson et al., 1999; Barnason et
al., 1998; Eilers et al., 1988; Ferozali et al., 2007; Gibson & Nelson, 2000; Holmes &
Mountain, 1993; Jiggins & Talbot, 1999; Paulsson, Wardh, Andersson, & Ohrn, 2008;
Ross & Crumpler, 2007). The modifications and validations of the OAS are
summarised in Table 3.2.
42
Table 3.2 Modifications and validations of the Oral Assessment Scale (OAS)
Authors Population Modifications Validity & Reliability testing Comments
Andersson, P., Persson, L., Hallberg, I., & Renvert, S. (1999)
Adult Oncology 7 categories (saliva, swallow, voice, gums, teeth/dentures, lips, mucous membranes)
Internal consistency: cronbach’s alpha = 0.87
Inter-rater reliability between r=0.65-0.92
Recommended to not use voice component; not a useful measurement within their study
Barnason, S., Graham, J., Wild, M., Jenson, L., Rasmussen, D., Shulz, P., et al. (1998)
Adult Critical Care
6 categories (saliva, gums, teeth/dentures, lips, mucous membranes, tongue)
Inter-rater reliability: r=0.92 Removed elements that are not assessable
Beck (1979) Adult oncology 15 categories (voice (four assessments), swallow (two assessments), saliva (two assessments), lips, tongue, mucous membranes, gingival, teeth / dentures, plaque, diet)
Nil Original
Eilers, J., Berger, A., & Peterson, M. (1988)
Adult oncology 8 categories (voice, swallow, lips, tongue, saliva, mucous membranes, gingival & teeth / dentures)
Content validity: literature review / panel of experts
Inter-rater reliability: r=0.912
High level of clinical usability, comprehensiveness, consistency
Ferozali, F., Johnson, G., & Cavangnaro, A. (2007)
Adult long-term care
4 categories (lips, oral mucosa, gingival tissues, and plaque)
Inter-rater reliability: r=>0.85
43
Authors Population Modifications Validity & Reliability testing Comments
Gibson, F., & Nelson, W. (2000)
Paediatric Oncology
8 categories (voice, swallow, lips, tongue, saliva, mucous membranes, gingival & teeth / dentures) Non-dentate adaption (score 1=healthy for teeth category)
Nil
Holmes, S., & Mountain, E. (1993)
Adult Palliative Oncology
Tested Eilers (Eilers et al., 1988) & Beck (Beck, 1979) OAS
Eilers inter-rater reliability: correlation coefficient r=0.73
Beck inter-rater reliability: correlation coefficient r=0.84
Content validity by panel of experts believed Eilers’ was insufficient in detail & Becks’ was too detailed & lengthy
Jiggins, M., & Talbot, J. (1999)
Paediatric ICU 5 categories (saliva, gums, teeth/dentures, lips, mucous membranes)
Inter-rater reliability testing discussed but results not provided – reported to ‘improve’ after education
No discussion re. non-dentate assessment
Ross, A., & Crumpler, J. (2007)
Adult Critical Care
5 categories (saliva, gingiva, teeth/dentures, lips, mucous membranes)
Nil Simplification for ventilated patients
44
The oral health of critically ill children was measured throughout the study by the
Oral Assessment Scale (OAS) (see Appendix A). The adaptations eliminating the
elements of voice and swallow, validated by Jiggins and Talbot (1999) and Barnason
et al (1998) were incorporated, as well as the simplification integrated by Ross and
Crumpler (2007). During critical illness, the child is frequently unconscious, unable to
communicate or developmentally immature therefore assessment procedures that
require active participation of the patient are unusable (Munro, Grap, Jablonski et al.,
2006). However, in order to ensure validity, the assessment needs to accurately
capture all of the components which are needed (Munro, Grap, Jablonski et al.,
2006), without increasing complexity.
The OAS involves assessment over five categories: lips, tongue, saliva, mucous
membranes/gingival and teeth. Three levels of descriptors are identified for each
category and the overall oral assessment score is the sum of the subscale scores. A
normal finding for a category is given a rating of one. If there is a barrier breakdown
or loss of function, the category is given a three. Accordingly, the rating of two is
used for not normal, but without barrier breakdown or loss of function. The scores
from the five categories are calculated with a normal mouth given a score of five, and
the highest possible score being 15.
The OAS is primarily an objective tool and multiple authors (Andersson et al., 1999;
Barnason et al., 1998; Eilers et al., 1988; Ferozali et al., 2007; Holmes & Mountain,
1993; Jiggins & Talbot, 1999) have previously undertaken inter-rater reliability
testing, therefore testing was not repeated. OAS was carried out by the bed-side nurse
every twelve hours, within their initial patient assessment at the commencement of
45
their shift. Education was carried out by the principal investigator to the bed-side
nurses in multiple formats, describing both the OAS and the procedure for
assessment. These formats included one-on-one education sessions, presentations in
group sessions at ward meetings and handovers, posters throughout the PICU, as well
as summary information attached to the computer monitor at the bed-side and in the
participants’ bed-side folder.
Pediatric Logistic Organ Dysfunction score (PELOD) and Paediatric Indicator of
Mortality 2 (PIM2)
Two methods to measure critical illness were used in the study - the Pediatric Logistic
Organ Dysfunction Score (PELOD) (see appendix B) and the Paediatric Indicator of
Mortality 2 (PIM2) (see appendix C). This was done because of the strengths and
limitations of each instrument. Each scoring system produces composite scores, made
up of a group of variables consisting of clinical and laboratory data (Lacroix &
Cotting, 2005).
The PELOD score (Lacroix & Cotting, 2005; Leteurtre et al., 1999; Leteurtre et al.,
2003; Santanae et al., 2009; Thukral et al., 2007; Yung et al., 2008) is a measure of
the severity of multiple organ dysfunction syndrome in the PICU. It is calculated for
each patient by adding the scores for individual organ systems based on recorded
levels of variables included in the systems. Six organ systems are included:
neurological, cardiovascular, renal, respiratory, haematological and hepatic, each
containing multiple variables, totalling 12. The physiologic variables are stratified
into four age groups: neonates (less than seven days or one month of age), infants
46
(one to 12 months of age), children (12 to 144 months of age) and adolescents (greater
than 144 months of age). The PELOD has been validated (Leteurtre et al., 2003) for
repeated measures showing the changing severity of organ dysfunction during an
individuals’ critical illness.
The PIM2 (Baghurst et al., 2008; Eulmesekian et al., 2007; Inwald et al., 2009; Shann
et al., 1997; Slater et al., 2002; Thukral et al., 2006; Wolfer et al., 2007) is a
regression model that uses admission data to predict intensive care outcomes for
children. The score is designed to be generated using physiological and patient data
available within the first hour of the patients’ admission, to generate an overall risk-
score. Within this study, the PIM2 is being used to describe mortality risk, and
therefore critical illness of the patient groups.
Clinical Pulmonary Infection Score (CPIS)
The Clinical Pulmonary Infection Score (CPIS) (A'Court et al., 1993; Pugin,
Auckenthaler, Miller et al., 1991)(see appendix D) has been commonly used to
measure pneumonia (McClure et al., 2009; Sakaquicki et al., 2008; Singh et al.,
1999), and VAP (Aydogdu & Gursel, 2008; Munro, Grap, Elswick et al., 2006; Pugin,
Auckenthaler, Lew et al., 1991; Sakaquicki et al., 2008), in research. The clinical
diagnosis combines body temperature, white blood cell counts, appearance of tracheal
secretions, oxygenation requirements, infiltrates on chest x-ray, and cultures of
tracheal aspirate to give a cumulative score. Within this study, the CPIS has been
included to examine the link between pneumonia, measured as CPIS, and pathogenic
colonisation of oropharyngeal flora and poor oral health of critically ill children.
47
3. 7.4 Data extraction tool
In order to describe the various influences on oral health during critical illness,
information was collected by the principal investigator and summarised using a data
extraction tool (see appendix E & F). Demographic data, information regarding the
participants’ critical illness, oral hygiene care practices and PICU therapies were
obtained from CareVue - a clinical information system. The incidence of HAI was
collected, using the previously defined diagnostic criteria (Horan et al., 2008) with
laboratory data available on AusLab® - an electronic database. Diagnostic samples
for HAI, such as tracheal aspirates and blood, were only obtained when clinically
indicated. The researchers were not involved in the decision-making regarding the
acquiring of diagnostic samples for HAI. Diagnostic samples were processed using
standard microbiological methods and techniques at the Queensland Health Pathology
Services Laboratory.
3. 8 DATA MANAGEMENT
Data were entered regularly throughout the study into the Statistical Package for
Social Sciences (SPSS) program, version 14 (SPSS, Chicago, IL). In order to
minimise potential data entry error, the same person entered the data. To check for
accuracy of data entry, verification for each item on the OAS, oropharyngeal
colonisation, and PIM2 calculations were conducted throughout the study. The
principal investigator was responsible for all data entry. All data collected were kept
in a locked filing cabinet accessible only to the research team, or kept under password
if stored electronically. All data will be maintained and kept for a period of five years
48
following the completion of the study. After this time it will be destroyed as per RCH
ethics policy.
3. 9 DATA ANALYSIS
Statistical methods were chosen to answer the previously defined research questions,
within the confines of an observational cohort design, and are defined below.
Statistical analysis was performed using SPSS version 14.0 (SPSS, Chicago, IL).
Statistical significance was set at p≤0.05.
Research Question One:
What is the status of oral health in critically ill children during admission to a PICU?
Descriptive statistics were used to describe oral health using the OAS and
oropharyngeal colonisation. Descriptive statistics were also used describe the
frequency and type of oral hygiene provided, demographic information and the main
clinical characteristics of the participants including age, length of PICU stay,
admission source, primary diagnosis, dentate status, and critical illness severity.
Counts and percentages were calculated for categorical variables, means and standard
deviations (SD) for normally distributed continuous variables and medians and
minimum and maximum values for abnormally distributed continuous variables.
49
Research Question Two:
How does the oral health of critically ill children change during their admission to
PICU?
A time series analyses allowed participants to be grouped according to PICU length of
stay in two day intervals (group A <48 hours; group B 48-96 hours; group C 96-144
hours; group D 144-192 hours; group E >192 hours). Categorical and continuous
values for oropharyngeal colonisation and OAS were then described over length of
time in PICU within these interval groups and the sample as a whole.
Research Question Three:
How is the oral health of critically ill children affected by patient characteristics or
PICU therapies?
Oral health, as described by the OAS and oropharyngeal colonisation, was examined
for relationships with multiple patient characteristics and PICU therapies using an
Analysis of Variance (ANOVA). The main analysis was undertaken using data
collected on day two of the participants’ admission to PICU for two reasons: a) to
analyse data that were a result of PICU condition and care (≥48 hours) in comparison
to the oral health of participants on admission, and b) after day two the sample size
reduced quickly due to discharge from PICU or death. A separate analysis was
undertaken to describe the effect the source of admission to PICU had on
oropharyngeal colonisation on the day of admission to PICU.
The ANOVA was used to examine the influence of the independent variables on the
variation of results in the dependent variables (Bhopal, 2008). Different types of
ANOVAs were used depending on the characteristics of the dependent and
50
independent variables. Oropharyngeal colonisation was again categorised into ‘nil
flora’, ‘commensal flora’ or ‘pathogenic flora’ as per previous definitions. The OAS
was analysed as a continuous variable and divided into a categorical variable: a score
less than five - no dysfunction; six to ten - moderate dysfunction; greater than or equal
to 11 - severe dysfunction. The data were first checked for normality of distribution to
determine whether parametric or nonparametric statistical tests of variance would be
used. The different ANOVA methods used for the dependent and independent
variables are explained in Table 3.3.
51
Table 3.3
Statistical tests for dependent and independent variables
Dependent variables Independent variables Statistical tests
OAS
Oropharyngeal
colonisation (OPC)
Critical illness scores
PIM2
PELOD
Age
Continuous OAS
Spearman’s rho
Categorical OAS & OPC
Kruskal Wallis
OAS
Oropharyngeal
colonisation
ETT present
Dentate status
Primary diagnosis
Oncological condition
Neutropenia
Clinical Pulmonary Infection
Score (CPIS)
Antibiotic therapy & oral anti-
fungal therapy -oropharyngeal
colonisation only
Continuous OAS
Mann-Whitney (2
independents)
Kruskal Wallis (3 or more
independents)
Categorical OAS & OPC
Fishers exact test
Oropharyngeal
colonisation (Day 0)
Admission source Fishers exact test
Research Question Four:
What is the relationship between dysfunctional oral health in critically ill children and
PICU-related HAI?
The data were first checked for normality of distribution to determine whether
parametric or nonparametric statistical tests would be used. The descriptive statistics
52
surrounding the clinical characteristics of the participants who developed PICU-
related HAI were summarised using frequency count, percentage, median, minimum
and maximum values. Non-parametric tests of significance - Kruskal, Wallis and
Fishers exact test - were undertaken to compare the demographic, clinical and oral
health characteristics between the two groups (those who developed PICU-related
HAI and those who did not). Cases of HAI and the corresponding oropharyngeal
colonisation results were outlined in a table format.
3.9 ETHICAL CONSIDERATIONS
Ethical approval to conduct the study was obtained from the Royal Children’s
Hospital, Brisbane and the Queensland University of Technology Human Research
Ethics Committees (see appendices G & H). The key elements of the ethical
considerations were informed consent, risk management and confidentiality and
security.
3.9.1 Informed consent
Informed consent for participation was obtained by the principal investigator or
trained PICU personnel. As the population studied were critically ill children, consent
for participation was gained from the child’s parents or legal guardians. The
participants and parents/guardians were provided with oral and written information
about the study and written consent was obtained for each participant prior to
oropharyngeal sampling (see appendix I & J). If the potential participant was greater
than 12 years of age, conscious, and could speak, read and write English, a youth
53
assent form was completed (see appendix K). Patients were not eligible for
recruitment if they had no parents or legal guardians present, were under the care of
Child and Family Services, or if the parents or guardians were unable to speak, read
or write English. Methodologically, approach for inclusion in the study was necessary
within 12 hours of PICU admission, however tact was used when choosing the time to
recruit, depending on the clinical condition of the child and the emotional state of the
parent or guardian.
Information was provided verbally and in written form. The information cover
sheet outlined that: 1) participation in the study is voluntary; 2) subjects are free
to withdraw before or during the study without comment or penalty; and 3)
subjects will not be prejudiced as a result of their participation. In addition,
parents, guardians and participants (where applicable) were made aware of the
purpose, nature and risks/benefits of the research. Parent/guardian and youth
assent information sheets and consent forms are shown in appendices 8 and 9.
3.9.2 Risk management procedures
There were no known risks associated with subjects participating in this study, as no
new therapeutic interventions were used and assessments were reasonably non-
invasive. As mentioned previously, consent to participate in the study was entirely
voluntary and participants were free to withdraw from the study at any time without
penalty. As the researcher is a senior nurse on the unit, participants may have thought
that the decision whether or not to participate in the study may threaten provision of
54
their future care. However, during the consenting process, participants were informed
that participation or non-participation would not change the care that their child
received in the PICU or the RCH. Nurses were encouraged to group oropharyngeal
swabs within the normal process of care, in order to minimise the disturbance of the
children. The anticipated benefits of participation were the early detection of oral
infections and notification of severe dysfunction to the Children’s Oral Health Service
located in the RCH.
3.9.3 Confidentiality and security
The confidentiality and anonymity of all participants and family members was
assured throughout the study. Coding mechanisms were applied to the data extraction
tool and information collected was not disclosed to anyone other than members of the
research team. No child or family members were identifiable within the final report or
subsequent publications.
All data collected were kept in a locked filing cabinet accessible only to the research
team, or kept under password if stored electronically. All data will be maintained and
kept for a period of five years following the completion of the study and after this
time destroyed, as per RCH ethics protocol.
55
Chapter 4 - RESULTS
4.1 INTRODUCTION
This chapter presents the results of the Critically ill Children’s Oral Health (CCOH)
study describing the oral health of critically ill children. The findings of the data
analysis, the characteristics of the study participants, and the data related to each
research question are presented.
4.2 CHARACTERISTICS OF PARTICIPANTS
Participants were recruited to the CCOH study between the 2nd of March 2008 and the
10th of December 2008. Within that period there were 414 admissions to the PICU at
the RCH and all were screened for potential study recruitment. The participants were
recruited following the previously explained inclusion/exclusion criteria. The sample
tree below (see Figure 1), depicts the recruitment of participants process. A total of 46
participants were recruited to the study.
56
Figure 1. Participant recruitment sample tree
The demographic characteristics of the CCOH study participants are presented in
Table 4.1. The participants had a wide variety of age, length of PICU stay, dentate
status, primary diagnosis, severity of critical illness and admission sources. In order
to estimate the study’s generalisability, comparisons were made between the CCOH
study population and information made available by the Australia and New Zealand
Paediatric Intensive Care Registry (ANZPICR) (2008) (Alexander et al., 2008).
414 admissions to RCH PICU
300 length of PICU stay < 48 hours
31 not recruited within 12 hours of admission
13 parents/care-givers not present within 12 hours
12 previously participated in study
6 under the care of child and family services
4 parents/care-givers unable to understand English
Informed consent
0 refused
46 participants
57
Table 4.1
Demographic characteristics and comparisons of CCOH study participants (n=46) and Australia and New Zealand Paediatric Intensive Care Registry (ANZPICR)
Demographics CCOH f ANZPICR
Age (months) Median
Minimum
Maximum
11.5
0.1
168
24
0.1
348
Length of PICU stay (hours)
Median
Minimum
Maximum
107.5
21
977
28.1
4
>4320
Dentate status Non-dentate 22 47.8% N/A
Dentate 24 52.2%
Participants receiving mechanical ventilation 35 76.0% 54.7%
Non-survivors of PICU 3 6.5% 2.9%
Primary Diagnosis Trauma 4 8.7% 7.3%
Respiratory failure 13 28.3% 22.7%
Post-operative 8 17.4% 24.3%
Neurology 4 8.7% 7.6%
Haematology/ Oncology 7 15.2% N/A
Sepsis 5 10.9% N/A
Other (liver etc) 5 10.9% 14.8%
Cardiovascular 0 0% 23.3%
Admission source Hospital ward 5 10.9% 19.1%
Hospital department of emergency medicine
9 19.9% 16.1%
Outside retrieval 17 32.6% 21.2%
Operating Theatre 15 37.0% 41.3%
PIM2 Median 1.79 1.08
Minimum
Maximum
0.23
31.58 0.02
99.59
PELOD (day 2)
Median 1.00 N/A
Minimum
Maximum
0.00
61.00
Pneumonia (day 2) 8 17.4% N/A
58
4.3 DATA ANALYSIS
Checks were conducted to reveal outliers on demographic data - age, length of
Paediatric Intensive Care Unit (PICU) stay, and for scores - the Oral Assessment
Scale (OAS), the Paediatric Index of Mortality (PIM2), the Pediatric Logistic Organ
Dysfunction Score (PELOD) and the Clinical Pulmonary Infection Score (CPIS).
Any discrepancies identified were checked and corrected where necessary.
The data cleaning process identified missing data within the oropharyngeal
colonisation results. Of the 46 study participants, seven (15.2%) did not have
oropharyngeal colonisation sampling completed on one of the required days within
their stay (two participants did not have swabs completed within 12 hours of
admission, seven on day two, four on day four, and one on day six or eight) The
demographics of the participants who had missing values are summarised in Table
4.2. As the missing values are widely spread among the participants’ demographic
groups and range, all available results were included in the final analyses.
59
Table 4.2
Summary demographic characteristics of participants with missing oropharyngeal colonisation results (n=7)
Age
(months)
Length of stay
(hours)
Dentate status Primary diagnosis
Median 56 Median 112 Dentate 4 (57.1%) Trauma 1 (14.3%)
Min. 1.5 Min. 109 Non-dentate
3 (42.9%) Neurology 1 (14.3%)
Max. 120 Max. 528 Respiratory failure
3 (42.9%)
Post-operative 1 (14.3%)
Haemotology/ Oncology
1 (14.3%)
Prior to undertaking the statistical analysis, standardisation of the OAS was necessary
to allow direct comparison between the non-dentate and dentate groups. One of the
characteristics of the OAS was assessment of the participants’ teeth, which had an
effect on the total score participants were able to achieve. There were two techniques
available for standardisation: increasing the non-dentate score by one; and factoring
the non-dentate scores by 15 (total available score for dentate group) divided by 12
(total available score for non-dentate group). Previous studies have utilised both
techniques (Gibson & Nelson, 2000; Jiggins & Talbot, 1999), and analyses showed
there were no significant differences in the median and mean results using the two
different methods. Following the rule of parsimony, the simpler method, increasing
the non-dentate score by one, was used for further analysis.
60
Oropharyngeal colonisation was categorised into ‘nil flora’, ‘commensal flora’ or
‘pathogenic flora’ as per previous definitions, and analysed as a categorical variable.
The OAS was analysed as a continuous and categorical variable: a score equal to five
indicated no dysfunction; six to 10 indicated moderate dysfunction; greater than or
equal to 11 indicated severe dysfunction.
Checks for normality in distribution were performed on the continuous data. The
OAS, PIM2, PELOD, age and length of stay were positively skewed and three
standard deviations above and below the means did not approximate the minimum
and maximum values. Therefore medians, minimum and maximum values were
utilised for descriptive statistics and non-parametric analysis was undertaken. CPIS
had a normal distribution, however was dichotomised into a categorical variable
(Pneumonia: Yes / No) as per previous definitions.
61
4.4 RESULTS OF RESEARCH QUESTIONS
4.4.1 Research Question One - What is the status of oral health in critically ill
children during admission to a PICU?
Within the study, no participants had severe oral dysfunction during their critical
illness, 32 (62.6%) had moderate oral dysfunction and 14 (37.4%) had no oral
dysfunction during the course of their critical illness. Twenty-two participants
(47.8%) had moderate oral dysfunction within 12 hours of their admission to PICU;
and excluding the day of admission, 21 participants (53.8%) had oral dysfunction
during their critical illness.
Nineteen participants (41.3%) had pathogenic oropharyngeal colonisation during their
critical illness. Ten participants (21.7%) had pathogenic oropharyngeal colonisation
within 12 hours of admission to PICU; and excluding day of admission, 15
participants (32.6%) had pathogenic oropharyngeal colonisation during their critical
illness. Half of the participants (n=23; 50.0%) had absent oropharyngeal flora during
their critical illness.
One hundred and fifty-two oropharyngeal swabs were taken during the course of the
study. Table 4.3 summarises the frequency of commensal, pathogenic and
combinations of oropharyngeal flora isolated.
62
Table 4.3
Frequency of commensal, pathogenic and combinations of oropharyngeal flora isolated
Commensal flora (n=152) Frequency Percentage
Alpha-haemolytic streptococcus
Coagulase-negative staphylococcus
Non-haemolytic streptococcus
Neisseria sp.
Corynebacterium spp.
65
49
22
19
8
42.76%
32.34%
14.47%
12.50%
5.26%
Pathogenic flora (n=65)
Candida sp.
Staphylococcus aureus
Haemophilus influenzae
Enterococcus sp.
Pseudomonas aeruginosa
Escherichia coli
Acinetobacter sp.
Stenotrophomonas maltophilia
Klebsiella pneumoniae
Enterobacter cloacae
Serratia marcescens
30
11
6
3
3
2
2
2
2
2
2
46.15%
16.92%
9.23%
4.61%
4.61%
3.08%
3.08%
3.08%
3.08%
3.08%
3.08%
Combinations of flora (n=152)
Nil
Commensal only
Pathogenic & commensal
Pathogenic only
45
50
35
22
29.60%
32.89%
23.03%
14.47%
63
The most common commensal bacteria isolated from the oropharynx of critically ill
children during this study were the alpha-haemolytic streptococcus (42.76%), with a
wide range of other species colonising less frequently. Candida sp. were the most
common pathogenic organisms to colonise the oropharynx (46.15%), with
Staphylococcus aureus present frequently (16.92%) and a range of gram positive and
gram negative bacteria less frequently. Some participants were colonised with several
pathogenic and/or commensal bacteria at the one time. Commensal colonisation is
most common (32.89%), however an absence of oropharyngeal flora is frequent
(29.60%) and colonisation with both pathogenic and commensal flora is possible
(23.03%).
The oral hygiene care that the participants received during the course of their critical
illness varied widely, as summarised in Table 4.4. The table shows the most frequent
oral cleansing solution used on day two of the participants admission to PICU was
water (77.50%), the most frequent oral cleansing implement used was a foam swab
(62.50%), and the most common frequency of the provision of oral hygiene was every
six hours (40.00%).
64
Table 4.4:
Oral hygiene received by participants
Oral cleansing solution used (day two) Frequency Percentage
Nil
Water
Chlorhexidine mouthwash
Toothpaste
Sodium bicarbonate mouthwash
5
31
1
2
1
12.50%
77.50%
2.50%
5.00%
2.50%
Oral cleansing implement used (day two)
Nil
Foam swab
Cotton swab
Toothbrush
5
28
4
3
12.50%
62.50%
10.00%
7.50%
Frequency of oral hygiene (day two)
Nil
Every 4 hours
Every 6 hours
Every 12 hours
Once daily
5
9
16
7
3
12.50%
22.50%
40.00%
17.50%
7.50%
Received oral or systemic antifungal therapy
during PICU admission
6 13.04%
Dentate patients who had their teeth brushed
within 48 hours of PICU admission
2 8.33%
Participants who received oral care using a
cotton swab during PICU admission
16 34.78%
65
4.4.2 Research Question Two - How does the oral health of critically ill children
change during their admission to PICU?
In order to examine the change of oral health of the participant’s critical illness and
admission to PICU, two analyses were undertaken. Firstly, the participants were
divided into groups using PICU length of stay to delineate (group A <48 hours; group
B 48-96 hours; group C 97-144 hours; group D 145-192 hours; group E >192 hours).
Table 4.5 presents colonisation results and median and categorical Oral Assessment
Scores over two day intervals in PICU. There was neither upward nor downward
trend in the incidence of pathogenic colonisation or oral dysfunction, nor a change in
median OAS over increasing length of stay in PICU.
66
a Frequency and percentage. b Median, minimum and maximum value.
Table 4.5 Colonisation and median and categorical OAS over two day intervals in PICU for subgroups
Colonisation:a Categorical OAS:a
Subgroup Day Nil Commensal Pathogen Normal Moderate Severe Median OAS b
A (n= 9)
<48 hours
0 1 (11%) 6 (55%) 2 (22%) 5 (56%) 4(44%) 0 5.0 (5-8)
B (n = 12) 0 3 (25%) 6 (50%) 3 (25%) 6 (50%) 6 (50%) 0 5.0 (5-9)
48-96 hours 2 1 (11%) 6 (66%) 2 (22%) 7 (58 %) 5 (42%) 0 5.0 (5-8)
C (n=11) 0 5 (50%) 5 (50%) 0 7 (64%) 4 (37%) 0 5.0 (5-8)
97-144 hours
2 6 (55%) 3 (33%) 1 (11%) 8 (73 %) 3 (27%) 0 5.0 (5-8)
4 2 (17%) 2 (33%) 3 (50%) 7 (64%) 4 (36%) 0 5.0 (5-8)
D (n=5) 0 2 (40%) 3 (60%) 0 2 (40%) 3 (60%) 0 6.0 (6-8)
145-192 hours
2 2 (40%) 0 3 (60%) 3 (60%) 2 (40%) 0 5.0 (5-8)
4 1 (20%) 2 (40%) 2 (40%) 5 (100%) 0 0 5.0 (5-5)
6 1 (50%) 1 (50%) 0 3 (60%) 2 (40%) 0 5.0 (5-6)
E (n=10) 0 0 4 (45%) 5 (55%) 4 (40%) 6 (60%) 0 6.0 (5-7)
>192 hours 2 0 4 (57%) 3 (43%) 7 (70%) 3 (30%) 0 5.0 (5-6)
4 3 (33%) 1 (11%) 5 (55%) 5 (50%) 5 (50%) 0 5.5 (5-10)
6 5 (55%) 1 (11%) 4 (45%) 3 (30%) 7 (70%) 0 6.0 (5-10)
8 3 (33%) 3 (33%) 4 (45%) 8 (80%) 2 (20%) 0 5.0 (5-9)
67
The second analysis was undertaken using the sample as a whole, rather than subgroups.
Figure 2 presents the colonisation percentages over two day intervals in PICU for the
entire group.
Figure 2. Colonisation percentages over two day intervals
Figure three displays the median OAS for the entire group over two day intervals.
4.5
5
5.5
6
Day 0 Day 2 Day 4 Day 6 Day 8
Days in PICU
Med
ian O
AS
Median OAS
Figure 3. Median OAS over two day intervals
0%
10%
20%
30%
40%
50%
60%
Day 0 Day 2 Day 4 Day 6 Day 8
Days in PICU
Nil Flora Commensal FloraPathogen
Per
cen
tage
of
Col
onis
atio
n
68
Figure four displays OAS percentages for the whole group over two day intervals.
Figure 4. Categorical OAS percentages over two day intervals
These figures support previous results showing no upward or downward trend in oral
health over length of stay in PICU.
4.4.3 Research Question Three – How is the oral health of critically ill children affected
by patient characteristics or PICU therapies?
The main analysis undertaken used data collected on day two of the participants’
admission to PICU to examine the effect of patient characteristics and PICU therapies on
the participants’ oropharyngeal colonisation and OAS. The OAS was analysed as both a
continuous (range - five to 15) and categorical (normal, moderate dysfunction or severe
dysfunction) variable, the oropharyngeal colonisation only as a categorical variable.
Oropharyngeal flora was analysed primarily with the categories of nil, commensal,
0%10%20%30%40%50%60%70%80%90%
Day 0 Day 2 Day 4 Day 6 Day 8
Days in PICU
Normal Moderate Severe dysfunction
OA
S P
erce
nta
ge
69
pathogenic, but some analysis was performed using the dichotomy of – pathogenic;
yes/no.
The patient characteristics examined were critical illness scores (PIM2 and PELOD), age,
dentate status, primary diagnosis, oncological condition, neutropenia and presence of
pneumonia as measured by CPIS. The PICU therapies examined were presence of an oral
or nasal endotracheal tube (ETT), antibiotic therapy and oral anti-fungal therapy. A
secondary analysis was undertaken to describe the effect admission source to PICU had
on oropharyngeal colonisation on day zero of PICU stay. As previously described, a
variety of statistical methods were utilised: Spearman’s rho for continuous dependent and
independent variables; Mann-Whitney and Kruskal-Wallis for mixed categorical and
continuous variables; and Fishers exact test for categorical dependent and independent
variables. Only one independent variable had a statistically significant effect on the OAS.
The critical illness measurement PELOD was significantly associated with oropharyngeal
colonisation with x2=6.166 with df =2, and a p-value of 0.046.
4.4.4 Research Question Four - What is the relationship between dysfunctional oral
health in critically ill children and PICU-related healthcare-associated infections (HAI)?
Eight (17.4%) participants developed a PICU-related HAI during their critical illness.
Compared with the 38 participants free of HAI, these eight participants had an increased
median length of stay in the PICU (p=0.002), a higher median OAS on day two of
admission to PICU (p=0.245) indicating moderate dysfunction, and a higher critical
illness score (PELOD: p=0.072). The demographic and clinical characteristics of the
70
population are summarised in Table 4.6. The types of infection, isolated strains and
results of oropharyngeal flora sampling are given in Table 4.7.
Table 4.6.
Demographics and clinical characteristics of population developing HAI during PICU admission (n=8)
Admission source n (f)
RCH ward 2 (25.0%)
RCH department of emergency medicine 0
Outside RCH retrieval 2 (25.0%)
Operating Theatre 4 (50.0%)
Primary Diagnosis
Trauma 2 (25.0%)
Respiratory failure 1 (12.5%)
Post-operative 1 (12.5%)
Neurology 1 (12.5%)
Haematology / Oncology 2 (25.0%)
Sepsis 1 (12.5%)
Other (liver etc) 0
Participants receiving mechanical ventilation 7 (87.5%)
Non-survivors of PICU 1 (12.5%)
Dentate status
Dentate 2 (25.0%)
Non- Dentate 6 (75.0%)
Min Max Median
Age (months) 2.0 168.0 25.0
Length of PICU stay (hours) 107.0a 972.0 259.5
OAS (on day 2) 5.0 8.0 5.5
PIM2 0.57 6.16 2.91
PELOD (day 2) 0 21 10.5b ap=0.002 b p=0.072
71
Table 4.7.
Oropharyngeal sampling results and PICU-related HAI
Day No.
Oropharyngeal colonisation Type of HAI Pathogen in HAI Day of HAI
0
Commensal flora & Haemophilus influenzae
Pneumonia
(ETT aspirate) Haemophilus
influenzae
2
2 Commensal flora & Acinetobacter baumannii
4 Commensal flora & Acinetobacter baumannii
0 Commensal flora &
Escherichia coli
Pneumonia (ETT aspirate)
Pseudomonas aeruginosa
2
2 Commensal flora & Candida sp.
4 Candida albicans
6 Pseudomonas aeruginosa & Candida albicans
8 Commensal flora
0
Commensal flora &
Haemophilus influenzae
Pneumonia (ETT aspirate)
Candida sp.
2
2-6 Commensal flora & Candida sp.
8
Commensal flora, Haemophilus influenzae & Candida sp.
Pneumonia (ETT aspirate)
Staphylococcus aureus & Haemophilus influenzae
8
10 Not done
12 Commensal flora & Candida albicans
14-18 Commensal flora
20 Klebsiella pneumoniae, Enterobacter cloacae & Candida albicans
22-24 Commensal flora
26 Klebsiella pneumoniae, Enterobacter cloacae & commensal flora
28 Not done
30-32 Commensal flora
72
Day No.
Oropharyngeal colonisation Type of HAI Pathogen in HAI Day of HAI
34-36 Nil flora
38 Commensal flora
0 Nil Blood-stream infection (Blood culture)
Enterococcus faecalis 4
2-4 Commensal flora & Staphylococcus aureus
0-6 Commensal flora Bacteraemia
(Blood Culture) Escherichia coli 6
0
Staphylococcus aureus & commensal flora
Pneumonia (ETT aspirate)
Candida sp. 10
2 Not done
4
Staphylococcus aureus, Candida sp. & commensal flora
6-18 Candida sp.
20 Nil
0-4 Not Done Pneumonia
(ETT aspirate)
Candida sp. 14
6-8 Nil
10 Commensal flora & Candida sp.
12-14 Nil
16 Candida albicans
18-22 Nil
0-4 Candida albicans Blood-stream infection (Blood culture)
Stenotrophomonas Maltophilia
18
6-10 Nil
12-14 Pseudomonas aeruginosa
16
18 Stenotrophomonas maltophilia Nil
20 Stenotrophomonas maltophilia
22-24 Serratia marcescens & commensal flora
73
During their PICU stay, two participants acquired a blood-stream infection or bacteraemia
(Enterococcus faecal is (n=1) and Escherichia coli (n=1)) without concurrent oropharyngeal
colonisation. One participant developed pneumonia (Pseudomonas aeruginosa) on day two
of admission to PICU, and the same organism was isolated in their oropharyngeal flora four
days later.
In the five remaining participants (four with pneumonia, one with bacteraemia), the
pathogens found in their blood or endo-tracheal tube (ETT) aspirates were also isolated from
oropharyngeal sampling - simultaneously in one participant, and before any other sampled
site of colonisation in the remaining four participants. One participant, who was admitted
with a pneumonia caused by Haemophilus influenzae, developed two PICU-related HAI’s
during their critical illness, one of which was associated with oropharyngeal colonisation
prior to its development.
4.5 SUMMARY
Forty-six participants were recruited to the CCOH study between the 2nd of March 2008 and
the 10th of December 2008. Greater than half (n=32; 62.6%) of critically ill children in the
CCOH study had oral dysfunction during their critical illness, as measured by the OAS.
Forty-one percent of critically ill children in the study had pathogenic oropharyngeal
colonisation during their critical illness. The oral hygiene care that the participants received
during the course of their critical illness varied widely.
74
There was neither upward nor downward trend in the incidence of pathogenic colonisation or
oral dysfunction, nor a change in median OAS over increasing length of stay in PICU. After
examining multiple patient characteristics and PICU therapies for influence on oral health,
only the critical illness measurement PELOD was significantly associated with oropharyngeal
colonisation.
Eight (17.4%) participants developed a PICU-related HAI during their critical illness. In five
participants (four with pneumonia, one with bacteraemia), the pathogens found in their blood
or ETT aspirates were also isolated from oropharyngeal sampling - simultaneously in one
participant, and before any other sampled site of colonisation in the remaining four
participants.
75
CHAPTER 5 - DISCUSSION
5.1 INTRODUCTION
It is inevitable that some children in the general population will become critically
unwell and require intensive treatment in a Paediatric Intensive Care Unit (PICU).
Whilst critically ill, some children may have characteristics and receive PICU therapies
which put them at an increased risk for poor oral health and pathogenic oropharyngeal
colonisation. The oral cavity is fragile, and because of this, may easily become
dysfunctional and harbour pathogenic microorganisms. Pathogenic microorganisms,
such as Staphylococcus aureus, can cause severe systemic illnesses including
pneumonia and blood-stream infection. Worryingly, PICU nurses are ill-informed
regarding best practice for oral hygiene and its significance for their patients.
The primary goal of the Critically ill Children’s Oral Health (CCOH) study was to
describe the oral health of critically ill children. In addition, the study sought to examine
the influence of time in PICU, patient characteristics and PICU therapies on oral health,
and explore the relationship between poor oral health and PICU-acquired healthcare-
associated infections (HAI). Current knowledge regarding this area is limited and
heavily reliant upon extrapolation of adult data. Given the importance placed on
maintaining oral health in the adult critical care population by international institutions
such as the Centers for Disease Control and Prevention (CDC), the current limitations
of the paediatric research need to be addressed.
76
5.2 KEY FINDINGS
5.2.1 What is the status of oral health in critically ill children during admission to a
PICU?
The CCOH study aimed to describe the status of oral health in critically ill children
through use of the Oral Assessment Scale (OAS) and microbial colonisation in the
oropharynx. Greater than half (n=32; 62.6%) of critically ill children in the CCOH study
had oral dysfunction during their critical illness, as measured by the OAS. Oral
dysfunction manifested in various ways - development of ulcers, dental plaque, cavities,
cracked lips, decreased salivary flow, or generalised inflammation and infection. These
oral dysfunctions represent a breakdown in the local and systemic health of these
critically ill children. Decreased salivary flow, or xerostomia, causes a change in the
immunological defences within the oral cavity facilitating adhesion of pathogenic
organisms. Cracked lips, generalised infection and ulcers, display a breakdown in the
primary defence mechanism of the mouth, allowing infiltration of these pathogenic
organisms into the wider circulatory system. A build up of dental plaque has been
shown in adult populations to be a reservoir for respiratory pathogens, such as
Staphylococcus aureus and Pseudomonas aeruginosa (Fourrier et al., 1998;
Scannapieco et al., 1992). It could be concluded from this that oral dysfunction via each
of these mechanisms potentially contributes to the systemic health of children during
critical illness. Oral dysfunction, as an indicator of poor oral health, has been shown to
be equally prevalent in adult critical care studies (Munro, Grap, Elswick et al., 2006;
77
Scannapieco et al., 1992). Evidence-based oral hygiene practices need to be instituted to
improve the oral health of critically ill children.
Oral dysfunction, as quantified by the OAS, may be symptomatic of underlying
microflora changes. Forty-one percent of critically ill children in the study had
pathogenic oropharyngeal colonisation during their critical illness, which reinforces
previous work by Thorburn (2009). The most common pathogenic oropharyngeal
colonisation was with Candida sp., with similar prevalence seen in Singhi et al’s (2008)
descriptive study examining generalised Candidaemia in the PICU. Candida sp. are
frequently present in the oropharynx of healthy children (Hannula et al., 1999) as part of
commensal flora, however within the CCOH study the Candida sp. were considered
pathogenic when prevalent in high numbers (≥106 colony forming units (cfu) per litre)
indicating opportunistic colonisation. Candida sp. frequently are opportunistic
pathogens of blood-stream and respiratory infections in critically ill patients and those
with compromised immune status (Singhi et al., 2008). Over the last decade there has
been an increase in the incidence of opportunistic Candida sp. bloodstream and
respiratory infections in hospitals (Singhi et al., 2008), particularly in the critically ill.
Primary colonisation of Candida sp. within the oropharynx facilitates the entry of a
potential systemic pathogen to the respiratory and circulatory system.
In addition to Candida sp., the other potential respiratory and systemic pathogens
colonising the oropharynx of critically ill children during this study included
Staphylococcus aureus, Haemophilus influenzae, Enterococcus sp., and Pseudomonas
aeruginosa in accord with previous adult and paediatric studies (Fourrier et al., 1998;
Garrouste-Orgeas et al., 1997; Rubenstein et al., 1992; Scannapieco et al., 1992;
78
Thorburn et al., 2009). Respiratory and systemic pathogens are not usually prominent
members of the oral commensal flora of healthy adults (Scannapieco et al., 1992) or
children (Kononen, 2005). This suggests that the oropharynx of critically ill children
can act as a reservoir for these pathogens which have the potential to cause systemic
infections including pneumonia and blood-stream infections, which are consequently
associated with substantial morbidity and mortality (Chan et al., 2007; D. Jones &
Munro, 2008; Safdar et al., 2005).
Half of the participants (n=23; 50%) within the CCOH study had absent commensal
oropharyngeal flora during their critical illness. The absence of commensal microflora
has previously been described as poor or dysfunctional oral health (Marsh & Percival,
2006), as the unique functions of the microflora are absent. Within the oral cavity this
involves the commencement of the process of digestion and a barrier against
colonisation or overgrowth of potentially pathogenic micro-organisms. Oral hygiene
protocols for critically ill children need to stratify for PICU therapies and patient
characteristics associated with the absence of commensal flora, as well as provide
strategies to support the regeneration of commensal oropharyngeal flora and prevent
opportunistic overgrowth of pathogenic microorganisms.
Within this study, information surrounding oral care practices was collected by audit
and while this may not be completely reflective of actual practice, the nurses caring for
these children demonstrated wide variation in oral hygiene practices. A large proportion
of the critically ill children described in the study had oral dysfunction and the nurses’
practice surrounding choice of instruments, solutions and frequency of oral care was not
informed by patient characteristics or current research. Only 8% of dentate children had
79
their teeth brushed within 48 hours of admission to PICU, 12.5% of children received
no oral hygiene on day two of admission to PICU and cotton swabs were used in 35%
of patients. Even though a high proportion of the population had Candida sp. isolated
on their oropharyngeal swabs, only 13% received oral anti-fungal therapy. It would be
difficult to suggest that this nursing practice was informed by evidence, and this may be
reflected in the frequency of oral dysfunction in the critically ill children. Like all
patient management practices, oral health can be best managed through the institution of
evidence-based oral hygiene protocols specific to critically ill children.
5.2.2 How does the oral health of critically ill children change during their admission
to PICU?
Our findings of no change or trend, either positively or negatively, surrounding oral
health over PICU admission within the CCOH study, challenges other studies within
critical care. The previous descriptive study completed by Franklin et al (2000) found a
statistically significant increase in mean plaque scores during PICU stay
(admission=22.5; discharge=25.8; p=0.001) and gingival inflammation (admission=4.1;
discharge=5.5; p=0.006). Fourrier et al (1998), in their descriptive study on adult critical
care, also found a statistically significant increase in dental plaque on patients remaining
ICU for five days or greater (day 0: 1.1 ± 0.7; day 5: 1.6 ± 0.7; p<0.05). Their study also
found the frequency of colonisation by aerobic pathogens increased over length of ICU
stay (day 0: 27%; day 5: 40% and day 0: 31%; day 5: 38%; day 10: 46%) but the
number of participants in each group was too low to reach statistical significance.
Worsening oral health over time spent in PICU has been used to demonstrate the
80
ineffectiveness of current oral hygiene practices (Franklin et al., 2000), specifically the
efficacy of foam swabs; this study however, disputes this finding.
The reason for the difference between the CCOH study results to previous adult
(Fourrier et al., 1998) and paediatric (Franklin et al., 2000) studies in critical care may
be due to several factors. Considering the difference in critical care outcomes and the
diversity in baseline oral health components between the adult and paediatric
populations, a variation in study findings could be expected. The study by Franklin et
al., (2000) is the only other study found examining the change in composition of the
oral health of critically ill children over time in PICU. The inclusion criteria in the
CCOH study were broader than in Franklin et al’s., (2000), and included all patients
regardless of dentate status, which is arguably a more accurate reflection of critically ill
children. The oral assessments completed in the CCOH study were carried out by bed-
side clinicians, in comparison to dental care assistants in the study by Franklin et al.,
(2000). However the OAS has been shown to have sound reliability and validity in
multiple studies (Andersson et al., 1999; Barnason et al., 1998; Eilers et al., 1988;
Ferozali et al., 2007; Holmes & Mountain, 1993; Jiggins & Talbot, 1999) using bed-side
clinicians. Given the study by Franklin et al., (2000) is a singular study, and the
inclusion criteria and assessment methods utilised in the two studies were substantially
different, variation in results should be expected.
81
5.2.3 How is the oral health of critically ill children affected by patient characteristics
or PICU therapies?
Literature suggests that PICU therapies (e.g. intubation (Jiggins & Talbot, 1999),
antibiotics (Jiggins & Talbot, 1999; Sixou et al., 1996)), and patient characteristics
including age (Kononen, 2000), severity of critical illness (Rubenstein et al., 1992;
Thorburn et al., 2009), admission source (Toltzis, Hoyen, Spinner-Block, Salvator, &
Rice, 1999), neutropaenia (Sixou et al., 1996) and admission diagnosis (Sixou et al.,
1996; Thorburn et al., 2009)), would have a relationship with the oral health of critically
ill children. Unexpectedly, the majority of patient characteristics and PICU therapies
examined in this study had little or no relationship with the status of oral health of
participants. An increase in the severity of critical illness, as measured by the Pediatric
Logistic Organ Dysfunction Score (PELOD), demonstrated a statistically significant
positive association (p=0.046) with pathogenic or absent oropharyngeal flora, in
comparison to commensal flora. It has been established in adult studies that severe
illness alters the oropharyngeal flora (Fourrier et al., 1998; Scannapieco et al., 1992;
Thorburn et al., 2009). The CCOH study findings correlate with the study by
Rubenstein et al. (Rubenstein et al., 1992), which found that PICU patients who were
colonised orally with pathogenic microflora such as Candida sp. had higher levels of
critical illness (Pediatric Risk of Mortality: 12.9 ± 2.8 vs. 4.9 ± 1.0; p=0.01).
PELOD (Leteurtre et al., 1999) measures the severity of critical illness by quantifying
the severity of multiple organ dysfunction. The link between organ dysfunction and
pathogenic or absent oropharyngeal colonisation is logical when recognising the support
and treatment organ dysfunction often requires within the PICU. Therapies including
82
antibiotics, intubation and fluid restriction, all of which have previously been suggested
as possible causes of microbial flora changes within the oral cavity (Kite & Pearson,
1995; Sixou et al., 1996), are instituted to support a child with worsening organ
dysfunction. Jiggins and Talbot (1999) previously developed a ‘risk assessment for oral
disease’ for use in the PICU, and while not validated, it is comprised of therapies used
to support children during organ dysfunction, including ventilation, sedation, dialysis
and diuretics. The development and validation of a risk assessment for poor oral health
in critically ill children would be a valuable addition to an evidence-based oral hygiene
protocol, allowing for stratification of ‘at risk’ children.
Within the CCOH study, critically ill children who were orally intubated did not have a
significantly higher incidence of dysfunctional oral health or pathogenic oropharyngeal
colonisation, in comparison to self-ventilating or nasally intubated patients. This
contrasts previous studies which describe nurses as being ‘uneasy’ or ‘unwilling’ to
provide oral hygiene to patients who were orally intubated as they viewed the risk of
accidental extubation as being too great (Binkley et al., 2004; Fitch et al., 1999; Rello et
al., 2007b). Either this attitude was not active within the RCH study group, or the oral
hygiene practices did not have an effect on the outcome of oral health. An analysis of
the effect of oral hygiene practices on oral health was not within the scope of this study.
Owing to both the variety of oral hygiene practices provided to each participant
throughout their admission to PICU and the size of a master’s thesis, this additional
record was not undertaken. Further research needs to be conducted to explore PICU
nurses’ current knowledge and practice surrounding oral health and oral hygiene.
83
As the only patient characteristic to have a relationship with worsening oral health was
severity of critical illness, oral hygiene protocols need to be focussed on increasing
support for children with severe levels of critical illness. Other patient and clinical
characteristics, such as age, dentate status and admission diagnosis, need to be
incorporated into oral hygiene protocols for practical purposes. Protocols need to be
developed to encompass a variety of solutions, instruments and frequency with
stratification for severity of critical illness. To ensure feasibility, the oral hygiene
protocols need to take into consideration the attitudes and knowledge of PICU nurses
surrounding oral health and oral hygiene.
5.2.4 What is the relationship between dysfunctional oral health in critically ill children
and PICU-related healthcare-associated infections (HAI)?
Of the population described by this study, 17.4% (eight) developed a PICU-related HAI
during their critical illness. HAIs are a common, serious problem in critically ill
children and are associated with substantial morbidity and mortality along with
increased attributable costs. Within the CCOH study, in comparison to the participants
who did not develop a PICU-related HAI, this group had higher critical illness scores
(PELOD: p=0.072) and longer length of admission (p=0.002). Notably within the scope
of this study, the participants who developed a PICU-related HAI had a median OAS of
5.5 (5.0-8.0) on day two of admission to PICU, indicating moderate dysfunction, in
comparison to the remaining PICU population having a median OAS of 5.0 (5.0 – 8.0).
While this relationship did not reach statistical significance (p=0.245) and sample size
does not allow further statistical analysis, this trend towards poor oral health in critically
ill children who developed a PICU-related HAI may indicate a potential relationship
84
between the development of HAI and poor oral health. Adult critical care studies have
confirmed that poor oral health (Munro, Grap, Elswick et al., 2006), and pathogenic
oropharyngeal colonisation (Abele-Horn et al., 1997; Garrouste-Orgeas et al., 1997;
Pugin, Auckenthaler, Lew et al., 1991), increases the risk of HAI, such as pneumonia.
Further research regarding this area in paediatrics is required before causal relationships
can be suggested.
Not all participants in the CCOH study who had pathogenic oropharyngeal colonisation
during their admission to PICU developed a PICU-related HAI. However, of the eight
participants who did develop a PICU-related HAI, six (75%) participants (five
pneumonias, one blood-stream infection) had the causative pathogens isolated from
oropharyngeal sampling previously or simultaneously. While these results are limited
by small sample size, they echo previous findings in adult critical care by Fourrier et al
(1998) and Munro et al. (2006). The CCOH study results suggest that the oropharynx
of critically ill children could be a reservoir of potential systemic bacterial and fungal
pathogens. PICU-related HAI as a result of translocation of pathogens from the
oropharynx to the respiratory and cardiovascular systems are physiologically plausible.
Oral hygiene treatment strategies should be directed towards reducing the prevalence of
pathogenic oropharyngeal colonisation and improving oral health.
The most common oropharyngeal pathogen to be associated with a consecutive PICU-
related HAI within the study was Candida sp., with three participants being colonised in
their oropharynx prior to developing clinically evident pneumonia, caused by Candida
sp. Critically ill patients and those with compromised immune status are prime targets
for opportunistic Candida infection (Singhi et al., 2008). The known predisposing
85
conditions for increased colonisation by Candida sp. include prematurity, antibiotic
therapy, central venous catheter, use of antacids, immunosuppression, immuno-
deficiency and malnutrition (Singhi et al., 2008), all of which are frequently present or
occur during critical illness in childhood. The other oropharyngeal pathogens associated
with a consecutive PICU-related HAI within the CCOH study were Haemophilus
influenzae, which later was the causative agent for pneumonia, and Stenotrophomonas
maltophilia, which later was the causative agent for a blood-stream infection. Whether
the oropharynx was the first site of colonisation for the pathogens, or whether the
oropharyngeal colonisation was coincidental or causative, has not been established by
this study due to limitations with sample size and design, however it does suggest that
further research in this area needs to be completed.
While traditional definitions of ‘pathogenic’ microorganisms have been used in the this
study, in the immunocompromised patient a significant proportion of bacteraemias have
been associated with opportunistic commensal oral bacteria including Streptococci sp.
(Lucas, Beighton, Roberts, & Challacombe, 1997). Oral Streptococci, particularly the
Streptococci oralis group of the viridans Streptococci, are frequently isolated from
blood cultures of febrile neutropaenic patients and the incidence of septicaemia caused
by commensal bacteria in this clinical sub-group is increasing (Lucas et al., 1997).
However, the majority of critically ill children have a symbiotic relationship with their
commensal oropharyngeal flora, therefore patients who are at risk of developing
opportunistic infections from commensal flora need to be identified and oral hygiene
strategies tailored to facilitate this.
86
5.3 STRENGTHS AND LIMITATIONS OF THE STUDY
Limitations of the CCOH study include its observational design and low sample size
(n=46). Consequently, the study lacks the ability to generate powered correlations
(Bhopal, 2008; Friis & Sellers, 2009) and is limited in its generalisability. However, the
study was intended to be exploratory. Considering the paucity of current literature
surrounding the oral health of critically ill children, the CCOH study has broadened the
body of knowledge available on the subject. Although the study design did not allow the
comparison of results to healthy controls, the significance of the results is not
diminished. Of the 46 participants, greater than half (n=32) had moderate oral
dysfunction during their critical illness, which has the potential to significantly affect
systemic health, no matter what the comparative prevalence in healthy populations.
One measurement of the status of oral health within the study was accomplished
through the OAS. A potential weakness of the study may be the effectiveness of the
OAS to adequately measure oral health, however substantial evidence exists to attest to
its historical reliability and validity (Andersson et al., 1999; Barnason et al., 1998;
Eilers et al., 1988; Ferozali et al., 2007; Holmes & Mountain, 1993; Jiggins & Talbot,
1999). Due to the OAS primarily objective design and given that other authors have
previously completed inter-rater reliability testing (Andersson et al., 1999; Barnason et
al., 1998; Eilers et al., 1988; Ferozali et al., 2007; Holmes & Mountain, 1993; Jiggins &
Talbot, 1999) and recent authors (Gibson & Nelson, 2000; Ross & Crumpler, 2007)
have not repeated testing, inter-rater reliability testing was not undertaken prior to this
study. While the OAS was completed by multiple assessors, education surrounding its
use is extensive and multi-faceted. Previous researchers (Munro, Grap, Jablonski et al.,
87
2006) have advocated the measurement of oral health via dental plaque and salivary
volume, however these measurements are elements of the OAS, so were not repeated.
Information on the sub-groups of the pathogens isolated on oropharyngeal swabs was
limited as semi-quantitative analysis was undertaken, which did not sub-classify all
bacterial and fungal types. The bacterial and fungal types chosen for analysis were
informed by previous research in adults and paediatrics and advised by a panel of
experts consisting of senior microbiologists and clinicians. Bed-side clinicians obtained
the oropharyngeal swabs rather than a single investigator, however nurses were trained
individually by the principal investigator and information was placed at the bedside
regarding the saliva sampling protocol.
While limitations are present in the study, it is the first of its kind to fully describe the
oral health of critically ill children and provides background information for the
development of further studies. In comparison to previous studies, oral health was
systematically described using a validated assessment scale in combination with
microbiological analysis. The CCOH study was successful in examining the effect of
length of time in PICU, PICU therapies and patient characteristics on oral health.
A further strength of the CCOH study, in comparison to previous research in the area, is
the wide inclusion criteria. The study participants are an accurate representation of
critically ill children, including all patients regardless of their age, dentate status or
ventilatory support. However, the study did not include children with a primary
diagnosis of a cardiac condition as the RCH PICU does not have a cardiac service. The
participants were heterogeneous which, while challenging when undertaking
88
correlations, is necessary to adequately and completely describe a complicated
population. Future development of oral hygiene protocols will also need to encompass
this heterogeneity.
5.4 STUDY FINDINGS AND THEIR IMPLICTIONS FOR CLINICAL PRACTICE
The study results covered in this chapter, their clinical significance and supporting
literature, lead to the following conclusions:
Oral health is frequently dysfunctional during critical illness in childhood at the
RCH PICU. A large percentage of RCH critically ill children show signs of poor
oral health including inadequate salivary flow, dental plaque and ulcers.
Potential systemic pathogens are frequently isolated from the oropharynx of
critically ill children including Candida sp., Staphylococcus aureus,
Haemophilus influenzae, Enterococcus sp., and Pseudomonas aeruginosa.
Critically ill children in the PICU at the RCH frequently have absent
oropharyngeal flora, which is indicative of poor oral health.
PICU nurses working at the RCH conduct a wide variation in oral hygiene
practices some of which are not supported by current research.
In contrast to previous research (Fourrier et al., 1998; Franklin et al., 2000), the
oral health of critically ill children admitted to the RCH PICU did not worsen
over length of stay in PICU.
The severity of critical illness has a significant positive relationship with
pathogenic and absent colonisation of the oropharynx. Oral hygiene practices
should be more aggressive for children with high levels of critical illness.
89
A large percentage of PICU-related healthcare-associated infections involve
preceding or simultaneous colonisation of the oropharynx by the causative
pathogen.
While further study is required to determine the full merit of these conclusions, given
their potential impact upon clinical practice, further investigation appears warranted.
5.5 SUMMARY
Given the prevalence of dysfunctional oral health during childhood critical illness in the
RCH PICU population, and combined with the potential systemic consequences,
evidence based paediatric oral hygiene practices should be developed. These orl hygiene
practices should be researched to ensure effectiveness in improving the oral health of
critically ill children and be developed to take into considerations risk factors for
developing poor oral health.
90
CHAPTER 6 - CONCLUSION
6.1 Introduction
Within adult critical care research and practice, clinicians have acknowledged the
profound effect oral health has on systemic health (Abele-Horn et al., 1997; A. Berry &
Davidson, 2006; Binkley et al., 2004; Brennan et al., 2004; Chan et al., 2007; Cutler,
2005; Dennesen et al., 2003; Ewig et al., 1999; Fitch et al., 1999; Fourrier et al., 1998;
Garrouste-Orgeas et al., 1997; D. Jones & Munro, 2008; McNeill, 2000; Munro & Grap,
2004; Munro, Grap, Elswick et al., 2006). Despite this, little is known surrounding the
oral health of critically ill children and its relationship with systemic health. The
purpose of the Critically ill Children’s Oral Health (CCOH) study was to describe the
state of oral health during childhood critical illness, examine the relationship between
poor oral health and a variety of clinical characteristics, and explore the relationship
between poor oral health and systemic infections. A prospective observational cohort
study was undertaken at a single tertiary-referral Paediatric Intensive Care Unit (PICU).
The status of oral health was measured using the Oral Assessment Scale (OAS) (Eilers
et al., 1988) and culturing of the oropharyngeal flora. Information was also collected
concerning the use of supportive PICU therapies, clinical characteristics of the children
and the occurrence of PICU related healthcare-associated infections (HAI).
91
6.2 RESEARCH QUESTION CONCLUSIONS
Oral health has the potential to influence systemic health during critical illness in
childhood. This study has demonstrated that oral health is frequently dysfunctional and
the oropharynx frequently harbours potential systemic pathogens during childhood
critical illness. Considering this, it is worrying that PICU nurses working at the Royal
Children’s Hospital (RCH) had variable oral hygiene practices, some of which are not
supported by current research. The only clinical characteristic which had a relationship
with dysfunctional oral health described in this study, was severity of critical illness.
This had a significant positive relationship with pathogenic or absent colonisation of the
oropharynx. In contrast to previous research (Fourrier et al., 1998; Franklin et al., 2000),
the oral health of critically ill children admitted to the RCH PICU did not worsen over
length of stay in PICU. In addition to the physiological plausibility of translocation of
oropharyngeal flora to the blood stream and respiratory tract, this study found a large
percentage of PICU-related HAI involved preceding or simultaneous colonisation of the
oropharynx by the causative pathogen. While further study is required to determine the
full merit of the conclusions generated in the CCOH study, given their potential impact
upon clinical practice, further investigation would appear warranted.
6.3 RECOMMENDATIONS AND FURTHER RESEARCH
The prevalence of poor oral health during childhood critical illness in the RCH PICU
population, combined with potential systemic consequences supports the development
of evidence-based paediatric oral hygiene practices. The development of these
evidence-based practices should include well-controlled clinical trials incorporating all
92
aspects of oral care interventions including solution, instruments and frequency. These
interventions could then be summarised in a protocol to help create useful and
appropriate clinical practices. Preferably, the protocol should involve regular oral
assessment utilising a validated oral assessment scale, be tiered by severity of critical
illness and be governed by the practical elements of dentate status, conscious level,
intubation status and developmental age.
There is currently minimal research exploring the relationship between dysfunctional
oral health and deteriorating systemic health within paediatric critical illness.
Specifically the link between pathogenic oropharyngeal colonisation and ventilator
associated pneumonia (VAP) should be further examined. Considering the mortality,
morbidity and economic consequences of VAP, this needs to be addressed.
There is little information available surrounding current PICU oral hygiene nursing
practice. While elements of oral hygiene practices were collected as part of the CCOH
study, this is indicative of a small sample of nurses from a single Australian PICU.
Additionally, no information is available concerning the attitudes, beliefs, training and
knowledge of health care workers within PICU surrounding oral health. Current theory
surrounding oral hygiene provision is based on extrapolation from adult literature
(Binkley et al., 2004; Fitch et al., 1999; Rello et al., 2007a). A multi-centre survey
should be undertaken to ascertain the current state of oral health practices and the
attitudes, beliefs, training and knowledge of health care workers within PICU.
In summary, given the findings of the CCOH study, several recommendations can be
made:
93
Evidence based oral hygiene practices and protocols need to be instituted to
improve the oral health of critically ill children.
Oral hygiene practices and protocols should be directed towards reducing the
prevalence of pathogenic oropharyngeal colonisation and improving oral health.
Oral hygiene practices and protocols need encompass the heterogeneity of the
PICU population.
Oral hygiene practices and protocols need to be focussed on increasing support
with severe levels of critical illness.
The development and validation of a risk assessment for poor oral health in
critically ill children would be a valuable addition to an evidence-based oral
hygiene protocol, allowing for stratification of ‘at risk’ children.
Oral hygiene protocols need to take into consideration the attitudes and
knowledge of PICU nurses surrounding oral health and oral hygiene.
Further research needs to be conducted to explore PICU nurses’ current
knowledge and practice surrounding oral health and oral hygiene.
Further research needs to be conducted surrounding the link between poor oral
health as pathogenic oropharyngeal colonisation and the development of HAI.
Patients who are at risk of developing opportunistic infections from commensal
flora need to be identified and oral hygiene strategies tailored to facilitate this.
94
6.4 CONCLUSION
The CCOH study has established that oral health is often clinically and
microbiologically dysfunctional during critical illness in childhood. Future research
should focus on the development of evidence-based oral hygiene practices for
integration into PICU clinical practice. While the CCOH study has contributed to the
body of knowledge available surrounding the oral health of the critically ill child,
further research needs to be completed to ascertain the relationship between oral health
and HAI in this population.
95
REFERENCES
A'Court, C., Garrard, C., Crook, D., Bowler, I., Conlong, C., Peto, T., et al. (1993). Microbiological lung surveillance in mechanically ventilated patients, using non-directed bronchial lavage and quantitative culture. Quarterly journal of medicine, 86, 635-648. Abele-Horn, M., Dauber, A., Bauernfeind, A., Russwurm, W., Seyfarth-Metzger, I., Gleich, P., et al. (1997). Decrease in nosocomial pneumonia in ventilated patients by selective oropharyngeal decontamination (SOD). Intensive Care Medicine, 23, 187-195. Alexander, J., Tregea, S., & Slater, A. (2008). Australian and New Zealand Paediatric Intensive Care Registry 2006. Brisbane: Australian and New Zealand Intensive Care Society. Andersson, P., Persson, L., Hallberg, I., & Renvert, S. (1999). Testing an oral assessment guide during chemotherapy treatment in a Swedish care setting: a pilot study. Journal of Clinical Nursing, 8, 150-158. ANZICS: Adult patient database. (2008). ANZICS Patient Database Annual Report 2007. Sydney: ANZICS. Aschengrau, A., & Seage, G. (2008). Essentials of epidemiology in public health (2nd ed.). Sudbury, Massachusetts: Jones and Bartlett Publishers. Aydogdu, A., & Gursel, G. (2008). Predictive factors for septic shock in patients with ventilator-associated pneumonia. Southern Medical Journal, 101(12), 1222-1226. Baghurst, P., Norton, L., & Slater, A. (2008). The application of the risk-adjusted control charts using the Paediatric Index of Mortality 2 for monitoring paediatric intensive care performance in Australia and New Zealand. Intensive Care Medicine, 34(7), 1282-1288. Barnason, S., Graham, J., Wild, M., Jenson, L., Rasmussen, D., Shulz, P., et al. (1998). Comparison of two endotracheal tube securement techniques on unplanned extubation, oral mucosa, and facial skin integrity. Heart and Lung, 27(6), 409-417. Beck, S. (1979). Impact of a systematic oral care protocol on stomatitis after chemotherapy. Cancer Nursing, 185-199. Berry, A., & Davidson, P. (2006). Beyond comfort: Oral hygiene as a critical nursing activity in the intensive care unit. Intensive and Critical Care Nursing, 22, 318-328. Berry, A., Davidson, P., Masters, J., & Rolls, K. (2007). Systematic literature review of oral hygiene practices for intensive care patients receiving mechanical ventilation. American Journal of Critical Care, 16(6), 552-562. Bhopal, R. (2008). Concepts of epidemiology: integrating the ideas, theories, principles and methods of epidemiology (2nd ed.). Oxford, New York: Oxford University Press. Binkley, C., Furr, L., Carrico, R., & McCurren, C. (2004). Survey of oral care practices in US intensive care units. American Journal of Infection Control, 32(3), 161-169. Bonten, M., Gaillard, F., Van Tiel, G., Smeets, G., Van Der Geest, S., & Stobberingh, E. (1994). the stomach is not a source for colonization of the upper respiratory tract and pneumonia in ICU patients. Chest, 105, 878-884. Bracken, M., Fleming, L., Hall, P., Van Stiphout, N., Bossley, C., Biggart, E., et al. (2009). The importance of nurse led home visits in the assessment of children with problematic asthma. Archives of the diseases of childhood, Epub ahead of print Jun 21. Brennan, M., Bahrani-Mougeot, F., Fox, P., Kennedy, T., Hopkins, S., Boucher, R., et al. (2004). The role of oral microbial colonization in ventilator-associated pneumonia.
96
Oral Surgery Oral Medicine Oral Pathology Oral Radiology And Endodontics, 98(6), 665-671. Chan, E., Ruest, A., O'Meade, M., & Cook, D. (2007). Oral decontamination for prevention of pneumonia in mechanically ventilated adults: systematic review and meta-analysis. British Medical Journal, 334, 889-900. Costello, T., & Coyne, I. (2008). Nurses knowledge of mouth care practices. British Journal of Nursing, 17(4), 264-268. Cutler, C. (2005). Improving oral care in patients receiving mechanical ventilation. American Journal of Critical Care, 14(5), 389-395. Daly, C., Mitchell, D., Grossberg, D., Highfield, J., & Stewart, D. (1997). Bacteraemia caused by periodontal probing. Australian Dental Journal, 42(2), 77-80. Dennesen, P., van der Ven, A., Vlasved, M., Lokker, L., Ramsay, G., Kessels, A., et al. (2003). Inadequate salivary flow and poor oral mucosal status in intubated intensive care unit patients. Critical Care Medicine, 31(3), 781-786. Donnelly, J., Bellm, L., Epstein, J., Sonis, S., & Symonds, R. (2003). Antibicrobial therapy to prevent or treat oral mucositis. The Lancet: Infectious Diseases, 3, 405-412. Edwards, E., Powell, C., Mason, B., & Oliver, A. (2009). Prospective cohort study to test the predictability of the Cardiff and Vale paediatric early warning system. Archives of the diseases of childhood, 94, 602-606. Eilers, J., Berger, A., & Peterson, M. (1988). Development, testing and application of the oral assessment guide. Oncology Nursing Forum, 15(3), 325-330. Eulmesekian, P., Perez, A., Minces, P., & Ferrero, H. (2007). Validation of pediatric index of mortality (PIM2) in a single pediatric intensive care unit of Argentina. Pediatric Critical Care Medicine, 8(1), 54-57. Ewig, S., Torres, A., El-Ebriary, M., Fabregas, N., Hernandez, C., Gonzalez, J., et al. (1999). Bacterial colonization patterns in mechanically ventilated patients with traumatic and medical head injury. American Journal of Respiratory and Critical Care Medicine, 159, 188-198. Feider, L. (2007). Survey of Oral Care Practices for the Orally Intubated Adult Critically Ill Patient. University of Washington, Washington. Ferozali, F., Johnson, G., & Cavangnaro, A. (2007). Health benefits and reductions in bacteria from enhanced oral care. Specialised Care for Dentistry, 27(5), 168-176. Fitch, J., Munro, C., Glass, C., & Pellegrini, J. (1999). Oral care in the adult intensive care unit. American Journal of Critical Care, 8(5), 314-318. Fourrier, F., Duvivier, B., Boutigny, H., Roussel-Delvallez, M., & Chopin, C. (1998). Colonization of dental plaque: a source of nosocomial infections in intensive care unit patients. Critical Care Medicine, 26(2), 301-308. Franklin, D., Senior, N., James, I., & Roberts, G. (2000). Oral health status of children in a Paediatric Intensive Care Unit. Intensive Care Medicine, 26, 319-324. Friis, R., & Sellers, T. (2009). Epidemiology for Public Health Practice (4th ed.). Sudbury, Massachusetts: Jones and Bartlett Publishers. Garner, J., Jarvis, W., Emori, T., Horan, T., & Hughes, J. (1996). CDC definitions of Nosocomial Infections. St. Louis: Mosby. Garrouste-Orgeas, M., Chevret, S., Arlet, G., Marie, O., Rouveau, M., Popoff, N., et al. (1997). Oropharyngeal or gastric colonisation and nosocomial pneumonia in adult Intensive Care Unit Patients. American Journal of Respiratory and Critical Care Medicine, 156, 1647-1655. Gibson, F., & Nelson, W. (2000). Mouth care for children with cancer. Paediatric Nursing, 12(1), 18-22.
97
Grap, M., Munro, C., Ashtiani, B., & Bryant, S. (2003). Oral care interventions in critical care: frequency and documentation. American Journal of Critical Care, 12(2), 113-118. Grap, M., Munro, C., Elswick, R., Sessler, C., & Ward, K. (2004). Duration of action of a single, early application of chlorhexidine on oral microbial flora in mechanically ventilated patients: a pilot study. Heart and Lung, 33(2), 83-91. Hannula, S., Jousimies-Somer, H., Takala, A., Syryanen, R., Kononen, E., & Asikainen, S. (1999). Age-related acquisition of oral and nasopharyngeal yeast species and stability of colonization in young children. Oral Microbiology and Immunology, 14, 176-182. Holmes, S., & Mountain, E. (1993). Assessment of oral status: evaluation of three oral assessment guides. Journal of Clinical Nursing, 2, 35-40. Horan, T., Andrus, M., & Dudeck, M. (2008). CDC/NHSN surveillance definition of health care-associated infection and criteria for specific types of infections in the acute care setting. American Journal of Infection Control (AJIC), 36(5), 309-328. Hugonnet, S., Sax, H., Eggimann, P., Chevrolet, J., & Pittet, D. (2004). Nosocomial bloodstream infection and clinical sepsis. Emerging Infectious Diseases, 10(1), 76-81. Inwald, D., Tasker, P., Peters, M., & Nadel, S. (2009). Emergency management of children with severe sepsis in the United Kingdom - the results of the Paediatric Intensive Care Society sepsis audit. Archives of the Diseases of Childhood, Jan 8, publish ahead of print. Jenkinson, H., & Lamont, R. (2005). Oral microbial communities in sickness and in health. Trends in Microbiology, 13(12), 589-595. Jiggins, M., & Talbot, J. (1999). Mouth care in PICU. Paediatric Nursing, 11(10), 23-26. Joint Faculty of Intensive Care Medicine. (2003). Minimum standards for intensive care units: Joint Faculty of Intensive Care Medicine. Jones, D., & Munro, C. (2008). Oral care and the risk of bloodstream infections in mechanically ventilated adults: a review. Intensive and Critical Care Nursing, 24, 152-161. Jones, H. (2005). Oral care in intensive care units. Specialised Care in Dentistry, 25(1), 6-11. Kite, K., & Pearson, L. (1995). A rational for mouth care: the integration of theory with practice. Intensive and Critical Care Nursing, 11, 71-76 Kononen, E. (2000). Development of oral bacterial flora in young children. Annals of Medicine, 32(2), 107-112. Kononen, E. (2005). Anaerobes in the upper respiratory tract in infancy. Anaerobe, 11, 131-136. Kononen, E., Asikainen, S., Karjalainen, J., & Jousimies-Somer, H. (1994). The oral gram-negative anaerobic microflora in young children: longitudinal changes from edentulous to dentate mouth. Oral Microbiology and Immunology, 9, 136-141. Kononen, E., Kanervo, A., Takala, A., Asikainen, S., & Jousimies-Somer, H. (1999). Establishment of oral anaerobes during the first year of life. Journal of Dental Research, 78(10), 1634-1639. Lacroix, J., & Cotting, J. (2005). Severity of illness and organ dysfunction scoring in children. Pediatric Critical Care Medicine, 6(3), S126-S134. Law, V., Seow, W., & Townsend, G. (2007). Factors influencing oral colonization of mutans streptococci in young children. Australian Dental Journal, 52(2), 93-100.
98
Leteurtre, S., Martinot, A., Duhamel, A., Gauven, F., Grandbastien, B., Nam, V., et al. (1999). Development of a pediatric multiple organ dysfunction score: use of two strategies. Medical Decision Making, 19(4), 399-410. Leteurtre, S., Martinot, A., Duhamel, A., Proulx, F., Grandbastien, B., Cotting, J., et al. (2003). Validation of the paediatric logistic organ dysfunction (PELOD) score: prospective, observational, multicentre study. The Lancet, 362(9379), 192-197. Lucas, V., Beighton, D., Roberts, G., & Challacombe, S. (1997). Changes in the oral streptococcal flora of children undergoing allogeneic bone marrow transplantation. Journal of Infection, 35, 135-141. Marsh, P., & Percival, R. (2006). The oral microflora - friend or foe? Can we decide? International Dental Journal, 56(4 (Supplement)). McClure, J., Cooke, R., Lal, P., Pickles, D., Majjid, S., Grant, C., et al. (2009). Outcome of late-onset hospital-acquired pneumonia related to causative organism. Journal of Hospital Infections, Jan 20, Epub ahead of print. McNeill, H. (2000). Biting back at poor oral hygiene. Intensive and Critical Care Nursing, 16, 367-372. Munro, C., & Grap, M. (2004). Oral health and care in the intensive care unit: state of the science. American Journal of Critical Care, 13(1), 25-33. Munro, C., Grap, M., Elswick, R., McKinney, J., Sessler, C., & Hummel, R. (2006). Oral health status and development of ventilator-associated pneumonia: a descriptive study. American Journal of Critical Care, 15(5), 453-460. Munro, C., Grap, M., Jablonski, R., & Boyle, A. (2006). Oral health measurement in nursing research: state of the science. Biological Research for Nursing, 8(1), 35-42. Napenas, J., Brennan, M., Bahrani-Mougeot, F., Fox, P., & Lockhart, P. (2007). Relationship between mucositis and changes in oral microflora during cancer chemotherapy. Oral Surgery Oral Medicine Oral Pathology Oral Radiology And Endodontics 103, 48-59. O'Reilly, M. (2003). Oral care of the critically ill: a review of literature and guidelines for practice. Australian Critical Care, 16(3), 101-110. Passos, J., & Brand, L. (1966). Effects of agents used for oral hygiene. Nursing Research, 15(3), 196-202. Paulsson, G., Wardh, I., Andersson, P., & Ohrn, K. (2008). Comparison of oral assessments between nursing staff and patients on medical wards. European journal of cancer care, 17, 49-55. Pugin, J., Auckenthaler, R., Lew, D., & Suter, P. (1991). Oropharyngeal decontamination decreases incidence of ventilator-associated pneumonia. JAMA, 265(20), 2704-2710. Pugin, J., Auckenthaler, R., Miller, N., Janssens, J., Lew, P., & Suter, P. (1991). Diagnosis of ventilator-associate pneumonia by bacteriologic analysis of bronchoscopic and non-bronchoscopic 'blind' bronchoalveolar lavage fluid. American Review of Respiratory Disorders, 143, 1121-1129. Rello, J., Koulenti, D., Blot, S., Sierra, R., Diaz, E., De Waele, J., et al. (2007a). Oral care practices in intensive care units: a survey of 59 European ICUs. Intensive Care Medicine, 33(6), 1066-1070. Rello, J., Koulenti, D., Blot, S., Sierra, R., Diaz, E., De Waele, J., et al. (2007b). Oral care practices in intensive care units: a survey of 59 European ICUs. Intensive Care Medicine, 33, 1066-1070. Roberts, M. (1998). Antibiotic resistance in oral / respiratory bacteria. Critical reviews in oral biology and medicine, 9(4), 522-540.
99
Ross, A., & Crumpler, J. (2007). The impact of an evidence-based practice education program on the role of oral care in the prevention of ventilator-associated pneumonia. Intensive and Critical Care Nursing, 23, 132-136. Rubenstein, J., Kabat, K., Shulman, S., & Yogev, R. (1992). Bacterial and fungal colonization of endotracheal tubes in children: a prospective study. Critical Care Medicine, 20(11), 1544-1549. Safdar, N. M. D. M. S., Dezfulian, C. M. D., Collard, H. R. M. D., & Saint, S. M. D. M. P. H. (2005). Clinical and economic consequences of ventilator-associated pneumonia: A systematic review. Critical Care Medicine, 33(10), 2184-2193. Sakaquicki, M., Shime, N., S., F., & Hashimoto, S. (2008). Current problems in the diagnosis and treatment of hospital-acquired methicillin-resistant Staphylococcus aureus pneumonia. Journal of Anaesthesia, 22(2), 125-130. Santanae, M., Leite, H., de Carvalho, W., & Lopes, E. (2009). Hypophosphatemia in critically ill children: Prevalence and associated risk factors. Pediatric Critical Care Medicine, Jan 30, Epub ahead of print. Scannapieco, F., Stewart, E., & Mylotte, J. (1992). Colonization of dental plaque by respiratory pathogens in medical intensive care patients. Critical Care Medicine, 20(6), 740-745. Schleder, B. (2003). Taking charge of ventilator-associated pneumonia. Nursing Management, 34(8), 27-33. Shann, F., Pearson, G., Slater, A., & Wilkinson, K. (1997). Paediatric index of mortality (PIM): a mortality prediction model for children in intensive care. Intensive Care Medicine, 23, 201-207. Singh, N., Gayowski, T., Wagener, M., & Marino, I. (1999). Pulmonary infiltrates in liver transplant recipients in the intensive care unit. Transplantation, 67(8), 1138-1144. Singhi, A., Raman Rao, D., & Chakrabarti, A. (2008). Candida colonization and candidemia in a pediatric intensive care unit. Pediatric Critical Care Medicine, 9(1), 91-95. Sixou, J., Medeiros-Batista, O., & Bonnaure-Mallet, M. (1996). Modifications of the microflora of the oral cavity arising during immunosuppressive chemotherapy. Oral Oncology: European journal of cancer, 32B(5), 306-310. Sixou, J., Medeiros-Batista, O., Ganddemer, V., & Bonnaure-Mallet, M. (1998). The effect of chemotherapy on the supragingival plaque of pediatric cancer patients. Oral Oncology, 34, 476-483. Slater, A., Shann, F., & Pearson, G. (2002). PIM2: a revised version of the Paediatric Index of Mortality. Neonatal and Pediatric Intensive Care, 29, 278-285. Somerville, R. (1999). Oral care in the intensive care setting: a case study. Nursing in Critical Care, 4(1), 7-13. Stiefel, K., Damron, S., Sowers, N., & Velez, L. (2000). Improving oral hygiene for the seriously ill patient: implementing research-based practice. Medsurg Nursing, 9(1), 40-43. Suljagic, V., Cobeljic, M., Jankovic, S., Mirovic, V., Markovic-Denic, L., Romic, P., et al. (2005). Nosocomial bloodstream infections in ICU and non-ICU patients. American Journal of Infection Control (AJIC), 33(6), 333-340. Sullivan, P., Juszczak, E., Bachlet, A., Lambert, B., Vernon-Roberts, A., Grant, H., et al. (2005). Gastrostomy tube feeding in children with cerebral palsy: a prospective, longitudinal study. Developmental medical and childhood neurology, 47(2), 77-85.
100
Tablan, O., Anderson, L., Besser, R., Bridges, C., & Hajjeh, R. (2003). Guidelines for preventing health-care associated pneumonia. Altanta: Centers for Disease Control and Prevention. Thorburn, K., Jardine, M., Taylor, B., Reilly, N., Sarginson, R., & van Saene, H. (2009). Antibiotic resistant bacteria and infection in children with cerebral palsy requiring mechanical ventilation. Pediatric Critical Care Medicine, 10(2), 222-226. Thukral, A., Kohli, U., Lodha, R., Kabra, S., & Kabra, N. (2007). Validation of the PELOD score for multiple organ dysfunction in children. Indian Pediatrics, 44(9), 683-686. Thukral, A., Lodha, R., Irshad, M., & Arora, N. (2006). Performance of Pediatric Risk of Mortality (PRISM), Pediatric Index of Mortality (PIM) and PIM2 in a pediatric intensive care unit in a developing country. Pediatric Critical Care Medicine, 7(4), 356-361. Toltzis, P., Hoyen, C., Spinner-Block, S., Salvator, A., & Rice, L. (1999). Factors that predict preexisting colonization with antibiotic-resistant gram negative bacilli in patients admitted to a Pediatric Intensive Care Unit. Pediatrics, 103(4), 719-723. Tombes, M., & Galluci, B. (1993). The effects of hydrogen peroxide rinces on the normal oral mucosa. Nursing Research, 42(6), 332-337. Turton, P. (2008). Ventilator-associated pneumonia in paediatric intensive care: a literature review. Nursing in Critical Care, 13(5), 241-248. Veldman, A., Trautschold, T., Weiss, K., Fischer, D., & Bauer, K. (2006). Characteristics and outcome of unplanned extubation in ventilated preterm and term newborns on a neonatal intensive care unit. Paediatric Anaesthesia, 16(9), 968-973. White, R. (2000). Nurse assessment of oral health: a review of practice and education. British Journal of Nursing, 9(5), 260-266. Wolfer, A., Silvani, P., Musicco, M., & Salvo, I. (2007). Pediatric Index of Mortality 2 score in Italy: a multicenter, prospective, observational study. Intensive Care Medicine, 33(8), 1407-1413. Yung, M., Wilkins, B., Norton, L., & Slater, A. (2008). Glucose control, organ failure and mortality in pediatric intensive care. Pediatric Critical Care Medicine, 9(2), 147-152. Zolldann, D., Thiex, R., Waitschies, B., Lutticken, R., & Lemmen, S. (2005). Periodic surveillance of nosocomial infections in a neurosurgery Intensive Care Unit. Infection, 33(3), 115-121.
101
APPENDICES
Appendix A
Oral Assessment Scale (OAS)
Lips – feel, observe 1 = Smooth, pink, moist
2 = Dry or cracked
3 = Ulceration or bleeding
Tongue – feel, observe 1 = Smooth, pink, moist
2 = Coated / shiny appearance, increased/decreased redness
3 = Thick and large, inflamed, blistered or ulcered
Saliva - observe 1 = Thin, watery and plentiful
2 = Thick or decreased
3 = Ropy or absent
Gingiva/oral mucosa – observe 1 = Smooth, pink, moist
2 = Generally pale, with small amount of reddened areas or ulcers, dry
3 = Bleeding, inflamed, multiple ulcers, very dry and oedematous
Teeth - observe 0 = Non-dentate
1 = Clean, no debris
2 = Plaque/debris in localised area
3 = Plaque/debris generalised, cavities visable
Tool modified from Beck (1979).
Appendix B
Pediatric Logistic Organ Dysfunction score (PELOD)
0 1 10 20 Neurological * Glasgow Coma Score 12-15 7-11 4-6 3 and Pupillary Reactions Both reactive N/A Both fixed N/A
Cardiovascular † Heart rate (beats/min) < 12 years ≥12 years
≤195 ≤150
N/A N/A
>195 >150
N/A N/A
Systolic Blood pressure (mmHg) <1 month 1 month-1year 1-12 years >12years
<140 <55 <100 <140
N/A N/A N/A N/A
35-65 35-75 45-85 55-95
<35 <35 <45 <55
Renal Creatinine (µmol/L) <7 days 7 days-1 year 1-12 years >12 years
<140 <55 <100 <140
N/A N/A N/A N/A
≥ ≥ ≥ ≥
N/A N/A N/A N/A
Respiratory PaO2 (kPa)/FiO2 ratio >9.3 and N/A ≥ or N/A PaCO2 (kPa) ≤11.7 and N/A >11.7 N/A Mechanical ventilation § No Ventilation Ventilation N/A N/A Haematological White blood cell count (x10 9/L) ≥4.5
and 1.5-4.4 or
<1.5 N/A
Platelets ≥35 <35 N/A N/A Hepatic Aspartate transaminase (IU/L) <950
and ≥950 or
N/A N/A
Prothrombin time ¶ (or INR) >60 (<1.40)
≤60 (≥1.40)
N/A N/A
Note: PaO2=arterial oxygen pressure. FiO2=fraction of inspired oxygen. PaCO2=arterial carbon dioxide pressure. INR=International normalised ratio. *Glasgow Coma Score: use lowest value. If patient is sedated, record estimated GCS before sedation. Assess patient only if known or suspected acute central nervous system disease. Pupillary reactions: non-reactive pupils must be >3mm. Do not assess after iatrogenic pupillary dilatation. †Heart rate and systolic blood pressure: do not assess during crying iatrogenic agitation. § PaO2: use arterial measurement only. ¶ Percentage of activity. PaO2/FiO2 ratio, which cannot be assessed inpatients with intracardiac shunts, is considered as normal in children with cyanotic heart disease. PaCo2 may be measured from arterial, capillary or venous samples. Mechanical ventilation: the use of mask ventilation is not counted as ventilation. Score = 0; no oral dysfunction Score = 1-103; increasing organ dysfunction
Tool modified from Lacroix & Cotting, 2005; Leteurtre et al., 1999; Leteurtre et al., 2003; Santanae et al., 2009; Thukral et al., 2007; Yung et al., 2008.
103
Appendix C
Paediatric Index of Mortality 2 (PIM2)
PIM2 Logistic Regression Model (n=20787)
Variables Coefficient 95% CI
Absolute (SBP -120), mmHg 0.01395 0.01054 to 0.01735
Pupils fixed to light (Y/N) 3.0791 2.7712 to 3.3869
100xFiO2 / PaO2, mmHg-1 0.2888 0.2015 to 0.3760
Absolute (base excess) mmol/l 0.1040 0.0919 to 0.1161
Mechanical ventilation (Y/N) 1.3352 1.1188 to 1.5516
Elective admission (Y/N) -0.9282 -1.1795 to -0.6768
Recovery post procedure (Y/N) -1.0244 -1.3235 to -0.7254
Bypass (Y/N) 0.7507 0.3971 to 1.1043
High risk diagnosis (Y/N) 1.6829 1.5185 to 1.8473
Low risk diagnosis (Y/N) -1.5770 -2.0244 to -1.1296
Constant -4.8841 -5.1132 to -4.6549
Note: SBP = Systolic Blood Pressure; mmHg = Millimetres of mercury; FiO2 = Fraction of inspired oxygen; PaO2 = Partial pressure of oxygen; Mmol/l = Millimols per litre Result 0.01 = 1% risk of mortality Result 99.9 = 99% risk of mortality
104
Appendix D
Clinical Pulmonary Infection Score (CPIS)
Parameter Value Score
Temperature (Celsius) 36.5 – 38.4 0
28.5 – 39.0 1
<36.0 - >39.0 2
White Blood Cells (mm-3) 4.0 – 11.0 0
11.0 – 17.0 1
>17.0 2
Secretions Nil 0
+ 1
++ 2
FiO2 / PaO2 (ratio in kPa) >33 0
<33 2
(unless ARDS, then 0)
Chest radiograph infiltrates clear 0
patchy 1
localised 2
Culture of tracheal aspirate No or < 1+ pathogenic bacteria cultured
0
>1+ pathogenic bacteria cultured
1
Note: FiO2 = Fraction of inspired oxygen; PaO2 = Partial pressure of oxygen; kPa = Kilopascals; ARDS = Acute Respiratory Distress Syndrome
Appendix E
Data Extraction Tool Participant _______________ UR No______________ Database No______________ Primary Dx_______________ Survivor YES/NO Dentate / Non-Dentate Age_______ Admission Source ________________ Adm Date/Time ____________ D/C Date/Time_____________
OAS PIM 2
PELOD CPIS Antibiotics
Y/N
Intubated
Y/N
PO Anti-fungal
Y/N
Neutro-penia
Y/N
Oral Hygiene (Freq, solution, implement)
Day 0
Day 2
Day 4
Day 6
Day 8
Day 10
Day12
106
Appendix F
Oropharyngeal Colonisation & Healthcare-Associated Infections
Participant _______________
Oropharyngeal
PICU Related Healthcare Associated Infection
DAY 0:
DAY 2:
DAY 4:
DAY 6:
DAY 8:
DAY 10:
DAY 12:
Appendix G
Ethical approval from Royal Children’s Hospital, Brisbane HREC
ROYAL CHILDREN'S HOSPITAL & HEALTH SERVICE DISTRJCT ETHICS COMMIITEE
Professor John Peam (Chair) 3365 5323 Mo Amanda Smith (Co-ordinator) 36369167
Ms Amanda UIlman Paediatric ICU Royal Children's Hospital & Health Service District Herston QLD 4029
Dear Ms lJIlman,
Queensland Government
Queensland Health
Lc:vel3. RCH Foundation Building Royal Children's Hospital
Herstoo QLD 4029 Australia Telephone (07) 3636 9167 Facsimile (07) 3365 54SS
23" Oclober 2007
Oropharyngeal flora in criticaUy ill children: changing compositioD during time spent in the
PICU.
Many thanks for your letter of the 28th September with responses to queries raised by the Committee in relation to the above project.
This was tabled at our meeting on the 22nd October and the Committee are happy to give their approval for this important work.
Please do Dot hesitate to contact me should you have any queries.
With kindest regards,
Professor John Peam Chair Royal Children's Hospital and Health Service District Ethics Committee
Cc: Ethics Commitlee files (professor John Peam) Members of the Ethics Committee
108
Appendix H
Ethical approval from the Queensland University of Technology HREC
University Human Research Ethics Committee
HUMAN ETHICS APPROVAL CERTIFICATE
NHMRC Registered Committee Number EC00171
Dear Mrs Amanda Ullman
A UHREC should clearly communicate its decisions about a research proposal to the researcher and the final decision to approve or reject a proposal should be communicated to the researcher in wrrting. This Approval Certificate serves as your written nolioo that the proposal has met the requirements of Ihe Nalional Slatemenl on Research involving Human Participation and has been approved on that basis. You are therefore authorized to commence activities as oumned in your proposal application, subject to any specific and standard conditions detailed in this document
Within this Approvat Certificate are:
• Project Details • Participant Details • Conditions of Approval (Specific and Standard)
Researchers should report to the UHREC, via the Research Ethics Officer. events that might affect continued ethical acceptability of the project. including, but not limited to:
(a) serious or unexpected adverse effects on participants; and (b) proposed significant changes in the conduct, the participant profile or the risks of the proposed research,
Further information regarding your ongoing obligations regarding human based research can be found yia the Research Ethics website hllp:lAw.w.research.quLedu.au/ethics/ or by contacting the Research Ethics Coordinator on 07 3138 2340 or [email protected]
If any details wlthm thiS Approval
thiS cerllficate J I lficale are incorrect please advise Research Ethics within 10 days of receipt of
Research Eth,cs Orrrcer =±�ti: 1&;;;-___________ _
(on behalf of the Chairperson U C)
I Projlu::t Detaits
Category of Approval : Approved by other HREC
Approved Until: 28/0412011
Approval Number: 0800000252
Date
Project Title: Oral health of critically itl children: changing composition during time spent in the PICU
Project Chief Invest igator: Mr5 Amanda Ullman
Other Project Slaff/Students: Or Peter lewis , Prof Glenn Gardner , Ms Debbie long, Ms Desley Hom , Or M3rk Coulth3rd
Experiment Summary: Investigate whether the composition of the oropharyngeal nora of critically ill children undergoes similar changes to that seen in the adult ICU and paediatric Oflcology.
I Participant Det ails
PartiCipants: Approximately 90 aged between 1 day and 15 years
locationis of the Work: PaediatriC Intensive Care Unit, Royal Children's Hospitat, Herston
RM Report No. E801 VersiO!1 2
109
University Human Research Ethi�s commi:!e HUMAN ETHICS APPROVAL CERTIFICATE
NHMRC Registered Committee Number EC00171
I Conditic>ns of Approval
Specific Conditions of Approval:
No special conditions placed on approval by the UHREC. Standard conditions apply.
Standard Conditions of Approval:
The University's standard conditions of approval require the research team to:
1 Conduct the project in accordance with University policy. NHMRC I AVCC guidelines and regulations, and the provisions of any relevant Slate I Territory or Commonwealth regulations or legislation;
2. Respond 10 the requests and instructions of the University Human Research Ethics Committee (UHREC);
3. Advise the Research Ethics Officer immedialely if any complaints are made, or expressions of concern are raised. in relation to the project;
4. Suspend or modify the project if the risks to participants are found to be disproportionate to the benefits. and immediately advise the Research Ethics Officer of this action;
5. Stop any involvement of any participant if continuation of the research may be harmful to that person. and immediately advise the Research Ethics Officer of this action;
6. Advise the Research Ethics Officer of any unforeseen development or events that might affect the continued ethical acceptability of the project;
7. Report on the progress of the approved project at least annualty, or at intervals determined by the Committee;
8. (Where the research is publicly or privately funded) publish the results of the project is such a way to permit scrutiny and contribute to public knowledge; and
9. Ensure that the results of the research are made available to the participants.
Modifying your Ethical C learance:
The University has an expedited mechanism for the approval of minor modifications to an ethical clearance (this includes changes to the research team, subject pool, lesting instruments, etc). In practice this mechanism enables researchers to conduct a number of projects under the same ethical clearance.
Any proposed modification to the project or variation to the ethical clearance must be reported immediately to the Committee (via the Research Ethics Officer), and cannot be implemenled until the Chief Investigator has been notified of the Committee's approval for the change f variation.
Requests for changes f variations should be made in writing 10 the Research Ethics Officer. Minor changes (changes to the subject pool, the use of an additional instrument, etc) will be assessed on a case by case basis and interim approval may be granted subject to ratification at the subsequent meeting of the Committee.
It generally takes 7 -14 days to process and notify the Chief Investigator of the outcome of a request for a minor change I variation.
Major changes to your project must also be made in writing and will be considered by the UHREC. Depending upon the nature of your request. you may be asked to submit a new application form for your project.
Audits:
All active ethical clearances are subject to random audit by the UHREC, whiCh will include the review of the signed consent forms for participants, whether any modifications f variations to the project have been approved, and the data storage arrangements.
End 01 Oocumenl
RM Report No. E801 Version 2
110
Appendix I
Parent / guardian information and consent forms
ROYAL CHILDREN’S HOSPITAL AND HEALTH SERVICE DISTRICT PARENT/GUARDIAN INFORMATION SHEET
Project Title: Oropharyngeal flora in critically ill children: changing composition during time spent in the PICU. Investigators: Ms Amanda Ullman (PICU Registered Nurse) Ms Debbie Long (PICU Nurse Researcher)
Ms Desley Horn (PICU Nurse Unit Manager) Contact Person:Ms Amanda Ullman PICU Registered Nurse Level 2, Surgical Building Royal Children's Hospital Ph: 07 3636 7957 Introduction You are invited to participate in a study that aims to assess whether the composition of the flora in the oral cavity of critically ill children changes during the course of their admission to the Paediatric Intensive Care Unit. We are sorry that your child is sick in Intensive Care and realise that this is a very stressful time for you. Thank you for taking the time to consider allowing your child to be part of this study. Background Oral health has a profound effect on general health. It is estimated that upwards of 500 species of bacteria normally inhabit the mouth. These bacteria have an important role in providing protection from illness-causing bacteria and organisms such as Staphylococcus aureus. However we think that during illness the oral cavity of critically ill children can change to become a source of infection, which could easily be transmitted to both the blood stream and the respiratory tract. Thorough studies have investigated this subject in adult intensive care and paediatric oncology patients, but no-one has investigated this in critically ill children. We know that children have different ‘normal’ flora (or bacteria) in their mouth before they get sick. The children within the Paediatric Intensive Care often have significant risk factors for developing illness-causing bacteria in the mouth, such as needing a tube to help them breathe or antibiotics.
111
As we mentioned previously, bacteria and organisms in the mouth can quickly transmit to both the blood stream and the respiratory tract. In a critically ill child, this can lead to pneumonia, a bloodstream-infection called septicaemia, or a urinary tract infection. These infections are often referred to as nosocomial infections. Nosocomial infections can be life threatening, increase PICU and hospital length of stays and have considerable cost implications related to medications and therapies. We think that the identification and prevention of illness-causing bacteria through a simple task of oral cares can have a significant impact on the lives of critically ill children. The results of this study aim to provide the foundations of further research where evidence based oral care regimes in critically ill children will be evaluated. Description of Research We would like to find out whether the bacteria in the mouths of critically ill children changes during the course of their admission to the Paediatric Intensive Care Unit (PICU). To do this we need to place a cotton-tipped swab in your child’s mouth for approximately thirty seconds, to collect a sample of their saliva. This will be sent to the laboratory to identify the bacteria/organisms that are present. We will take oral swabs on the day that your child is admitted to the PICU, and then every second day, while they are admitted to PICU. The results from these swabs will inform us whether the oral flora has changed from being good, ‘normal’ bacteria, to infection-causing ‘bad’ organisms. Your child would receive our standard treatment for oral hygiene. In addition to standard treatment, every day the nurses would assess the oral cavity for signs of infection or inflammation using an Oral Assessment Scale. If there is evidence of a large amount of inflammation or infection, a referral would be made to the Oral Health Service, within the Royal Children’s Hospital. If your child develops a nosocomial infection during the course of their admission to the PICU, we would then compare the organisms/bacteria found in the mouth with the bacteria/organisms found in the infection. Are there any risks involved? No, there are no risks involved. The oral swabs are minimally invasive and are not painful. Having the swab taken is similar to having a small soft toothbrush gently placed in your mouth. Are there any benefits? Yes, your child’s oral health status will be closely monitored and if their oral health deteriorates, they will be referred to the Oral Health service within the Royal Children’s Hospital for follow up. Withdrawing from the Study Your decision whether or not for your child to participate in this research will not prejudice their future relations with the Royal Children's Hospital and Health Service District. If you decide for your child to participate, you are free to withdraw your consent and to discontinue participation at any time. The decision to withdraw from the study will not affect their routine medical treatment or their relationship with the people treating them.
112
Confidentiality This project is being conducted in accordance with National Health and Medical Research Council ethical guidelines. All information will be kept in the strictest confidence in a locked filing cabinet and only the researchers will have access to the information. No information about the project will be published in any form that would allow any individual to be recognised. Auditors, ethics committees and regulatory authorities may access the anonymous data to ensure that the research is being properly conducted and meeting ethical rules and regulations. However, identifying data is not used. Contact If you wish to discuss any aspect of this study please phone Amanda Ullman on (07) 3636 7957. If you have concerns about the conduct of this study please contact the Chair of the Royal Children's Hospital Ethics Committee on (07) 3636 5323 or the Executive Director of Medical Services on (07) 3636 7590. Thank you for your consideration of this project.
113
Appendix J
ROYAL CHILDREN'S HOSPITAL AND HEALTH SERVICE DISTRICT PARENT/GUARDIAN CONSENT FORM
PARTICIPANT I have read the above information. I have asked all of my questions and I have gotten answers. I agree to enrol my child in this study. ______________________________________________________________ Signature of Parent or Guardian Date INVESTIGATOR I have fully explained to the parent/guardian ……………………………… the nature and purpose of the research and the procedures to be employed as described above and such risks as are involved in their performance, and I have provided the parent/guardian with a copy of the Parent/Guardian Information Sheet. ______________________________________________________________ Signature of Investigator Date ______________________________________________________________ Print Name Position INDEPENDENT WITNESS I have witnessed the receipt of an Information Sheet by the parent/guardian and the exchanging of information between the investigator and the parent/guardian about the study. ______________________________________________________________ Signature of Witness Date ______________________________________________________________ Print Name Position
114
Appendix K
Youth assent information and consent forms
ROYAL CHILDREN'S HOSPITAL AND HEALTH SERVICE DISTRICT CHILD/ADOLESCENT INFORMATION SHEET Project Title: Oropharyngeal flora in critically ill children: changing composition during time spent in the PICU. Investigators: Ms Amanda Ullman (PICU Nurse) Ms Debbie Long (PICU Nurse) Ms Desley Horn (PICU Nurse) Introduction This letter tells you about a project we are doing at the hospital. Your Doctor has told us that you need to stay in the Children’s Intensive Care Unit for the next few days. In your mouth you have hundreds of little good bacteria that help keep your mouth and teeth healthy. We know that when adults and kids with cancer get sick and need to stay in hospital, the bacteria in their mouth goes from being good to bad, which can make them get even sicker. We would like to ask you for your permission to participate in our project to help figure out whether this happens when you stay in the Children’s Intensive Care Unit. What will happen? In the Children’s Intensive Care Unit we have a team of special doctors and nurses looking after you and your family will be able to visit you here. If you decide to participate in our project a nurse will look in your mouth to see if your gums, teeth and tongue are healthy, like when you go to the dentist. Every second day they will put a cotton-tipped swab in the corner of your mouth for about 30 seconds. The swab is like placing a small soft toothbrush in your mouth. We will then send that swab to a scientist’s laboratory that will look at it under a microscope and see if the bacteria in your mouth are good or bad. Risks & Discomfort While you’re in the Children’s Intensive Care Unit the nurses will clean your teeth like normal. The nurse will put the cotton-tipped swab in your mouth which feels bit like a toothbrush. In the children’s intensive care unit we will try to minimise the amount of
115
pain or discomfort that you might have. This study will not involve any more pain or discomfort than what you have already discussed with your parents or doctor. Benefits We want to make sure that children staying in intensive care, like you, receive the best possible care. The information that we get from this project will hopefully tell us how we can take care of you better. Withdrawing from the Study Your decision whether or not to participate in this project will not affect the care or treatment that you receive at this hospital. If you decide to participate, you can say no and withdraw your decision and discontinue participation at any time. If you like, you can talk more about this project to your Mum or Dad, or any of the Doctors or Nurses. If you are happy to participate in our project, please tick yes or no at the bottom of the page by your name. Your name will not be in our project, so people who read it will not know who you are. Thank you for helping us understand more about children who are sick in the intensive care unit. Name: ____________________________________ Yes No