Abstract of thesis entitled
“An Evidence-based Guideline on Using Oral Probiotic to Prevent Late-
onset Sepsis in Preterm Infants”
Submitted by
Chan Shuk Wa
for the degree of Master of Nursing
at The University of Hong Kong
in July 2016
Late-onset sepsis is one of the most common problems in preterm infants.
According to a study from the National Institute of Child Health and Human
Development (NICHD), nearly 21% of preterm infants developed at least one or more
episodes of blood culture-proven sepsis during hospitalization. It also leads to additional
unpleasant events to the infants, their parents and medical staff, such as lengthen of
hospitalization, repeated invasive diagnostic and therapeutic procedures like long line
insertion, lumbar puncture and blood taking, increased risk of serious medical conditions,
such as meningitis, seizure, increased mortality, medical expenses, parental anxiety and
staff workload.
In recent decades, the use of oral probiotic was suggested to be an effective way
in preventing late-onset sepsis in preterm infants as it can strengthen the host natural
immune responses by developing gastrointestinal colonization.
This dissertation is a translational research that aims to identify the effectiveness
and implementation potential of using probiotic in preterm infants in order to reduce
incidence rate of late-onset sepsis and develop an evidence-based practice (EBP)
guideline for healthcare professionals.
A systematic literature review on the stated topic was done in three electronic
bibliographical databases, including PubMed, Cochrane library and the British Nursing
Index. The quality of the 6 studies found was assessed regarding the randomization,
concealment method, blinding, validity and reliability of outcome measures, dropout
rate and intention to treat, and other biases. And the EBP guideline was generated
based on the synthesis from the retrieved studies. Implementation potential of the
proposed guideline in local neonatal intensive care unit and special care baby unit
setting was discussed by assessing its transferability, feasibility and cost-benefit ratio.
Finally, a detailed implementation plan, including the communication plan, the pilot
study plan and the evaluation plan was proposed.
An Evidence-based Guideline on Using Oral Probiotic to Prevent Late-
onset Sepsis in Preterm Infants
by
Chan Shuk Wa
BNurs (HKU); RN (HK)
A thesis submitted in partial fulfillment of the requirements for
the Degree of Master of Nursing
at The University of Hong Kong
July 2016
i
Declaration
I declare that this dissertation represents my own work, except where due
acknowledgement is made, and that is has not been previously included in a thesis,
dissertation or report to this University or to any other institution for a degree, diploma
or other qualifications.
Signed………………………………………………
Chan Shuk Wa
ii
Acknowledgement
I would like to express my sincere gratitude to my dissertation supervisor, Dr.
Veronica S. F. Lam for her insightful comments and unfailing support throughout the
master dissertation. Her kindness and patience are greatly appreciated.
I am also grateful to my dissertation group members for their active participation
in the group discussion which inspired me a lot of excellent ideas to my dissertation.
Last but not least, I would like to offer sincere thanks to my family and friends for
their support throughout these two years of my master program.
iii
Table of contents Declaration ........................................................................................................................................ i
Acknowledgement ............................................................................................................................ ii
Table of contents ............................................................................................................................. iii
Abbreviations .................................................................................................................................... v
Chapter 1: Introduction ............................................................................................................... - 1 -
Background .................................................................................................................. - 1 -
Affirming the Need ....................................................................................................... - 2 -
Objective and Significance ........................................................................................... - 4 -
Chapter 2: Critical Appraisal ........................................................................................................ - 6 -
Search and Appraisal Strategies ................................................................................... - 6 -
Results .......................................................................................................................... - 8 -
Summary and Synthesis ............................................................................................. - 13 -
Chapter 3: Implementation Potential and Clinical Guideline .................................................... - 22 -
Target population and setting .................................................................................... - 22 -
Transferability ............................................................................................................ - 23 -
Feasibility ................................................................................................................... - 25 -
Cost-Benefit Ratio ...................................................................................................... - 29 -
Evidence-Based Practice Guideline ............................................................................ - 32 -
Chapter 4: Implementation Plan ................................................................................................ - 33 -
Communication Plan .................................................................................................. - 33 -
Pilot Study Plan .......................................................................................................... - 37 -
Evaluation Plan ........................................................................................................... - 40 -
Basis for Implementation ........................................................................................... - 43 -
Conclusion .................................................................................................................................. - 44 -
Bibliography ............................................................................................................................... - 45 -
iv
Appendix .................................................................................................................................... - 48 -
Appendix A: Search strategies .................................................................................. - 48 -
Appendix B: Table of Evidence ................................................................................... - 50 -
Appendix C: Summary or studies results ................................................................... - 56 -
Appendix D: Methodology Checklist 2: Controlled Trials .......................................... - 58 -
Appendix E: SIGN Methodological Quality Coding: levels of Evidence & Grading of
Recommendations ..................................................................................................... - 60 -
Appendix F: Summary of quality of studies ............................................................... - 61 -
Appendix G: Summary of quality of studies (Supplementary information) .............. - 62 -
Appendix H: Estimated cost for the innovation annually .......................................... - 66 -
Appendix I: Evidence-based practice guideline ......................................................... - 68 -
Appendix J: Schedule for implementation and evaluation ........................................ - 75 -
Appendix K: Gantt chart for implementation plan and evaluation ........................... - 76 -
Appendix L: Checklist for the use of oral probiotic to prevent late-onset sepsis in
preterm infants .......................................................................................................... - 77 -
Appendix M: Staff satisfaction survey on the guideline of using oral probiotic in
preterm infants .......................................................................................................... - 78 -
v
Abbreviations
APN Advanced Practice Nurse
BNI British Nursing Index
CFU Colony-forming unit
COS Chief of Service
CRP C-reactive protein
CSF Cerebrospinal fluid
DOM Department Operation Manager
EBP Evidence-based practice
eKG Electronic knowledge gateway
FAO Food and Agriculture Organization of the United Nations
HA Hospital Authority
LGG Lactobacillus rhamnosus GG
NEC Necrotizing enterocolitis
NICHD National Institute of Child Health and Human Development
NICU Neonatal intensive care unit
NS Nurse Specialist
PPE Personal protective equipment
RCT Randomized controlled trial
RDS Respiratory Distress Syndrome
RN Registered Nurse
ROP Retinopathy of prematurity
RR Relative risk
RRI Relative risk increased
RRR Relative risk reduction
SCBU Special care baby unit
SD Standard derivation
sIgA Secretory immunoglobulin A
SIGN Scottish Intercollegiate Guidelines Network
SPSS Statistical Package for the Social Science
VLBW Very low birth weight
WHO World Health Organization
WM Wand manager
- 1 -
Chapter 1: Introduction
Background
According to the World Health Organization (WHO) (2012), there is estimated 15
million preterm infants (born before 37 completed gestation weeks) born every year
with a raising trend. Among these preterm infants, around 10% are extremely preterm
(<28 weeks of gestation) and very preterm (28-32 weeks gestation) (WHO, 2012). These
extremely and very preterm infants usually need to be admitted to neonatal intensive
care unit (NICU) or special care baby unit (SCBU) after birth for further management.
WHO (2012) also pointed out that these extremely and very preterm infants are
facing a higher mortality and morbidity due to various complications, such as feeding
difficulties, severe infections, Respiratory Distress Syndrome (RDS), Necrotizing
Enterocolitis (NEC), Retinopathy of Prematurity (ROP) when compared to term or late
preterm infants. Among all the complications, late-onset sepsis, occurring more than 48
hours after birth, is found to be very common. Nearly 21% of very low birth weight
(VLBW) infants developed at least one or more episodes of blood culture-proven sepsis
during hospitalization according to a study from the National Institute of Child Health
and Human Development (NICHD) (Stroll et al., 2002) which lead to with prolonged
hospitalization, higher mortality and increased medical expenses (Tsai et al., 2014).
Therefore, it is important to address preventative strategies to reduce the incidence of
neonatal sepsis.
- 2 -
When compared with term infants, preterm infants have an immature immune
system. Their delayed natural gastrointestinal colonization may contribute to the easy
development of late-onset sepsis (Mshvidadze et al., 2010; Schwiertz et al., 2003).
Regarding to this situation, the use of oral probiotic was suggested in the past few
decades. Probiotics are live microorganisms that potentially benefit the host (FAO/WHO,
2006) by helping the host to develop gastrointestinal colonization and thus strengthen
the host immune responses (Millar, Wilks & Costeloe., 2003).
Therefore, the use of oral probiotics in preterm infants may be an effective
strategy to reduce the incidence of late-onset sepsis.
Affirming the Need
According to what I observed in my own hospital, around 70% of infants
admitted to NICU were extremely or very preterm infants. 70% of them were suffered
from either definite sepsis (culture-proven by blood or cerebrospinal fluid (CSF)) or
clinical sepsis (high C-reactive protein (CRP) or white blood cell level with clinical signs of
infection) during hospitalization which needs to treat with antibiotic courses for more
than 7 days. For infants who developed definite sepsis, around 20% of the extremely or
very preterm infants admitted, they even need a 6-8 weeks antibiotic course. The long
antibiotic course lengthens the hospitalization period which increases the medical
expenses and causes other problems.
Besides the length of hospitalization, neonatal sepsis also leads to additional
unpleasant events to infants, their parents and medical staff, such as long line insertion
- 3 -
for the long antibiotic course, repeated invasive diagnostic and therapeutic procedures
like lumbar puncture and blood taking, increased mortality, increased parental anxiety
and increased staff workload.
Furthermore, the supplies of NICU services are always in a stressful situation in
Hong Kong. We provide intensive care treatments such as early surfactant replacement
therapy, invasive and non-invasive ventilator support, optimal nutrition management
and urgent surgical interventions to those critically ill infants. The NICU in my hospital
always have full occupancy or even overbooked. As the demand of NICU service is very
large, the extremely preterm and very preterm infants will be transferred to SCBU for
further management once their conditions are stable. The prolonged hospitalization
would increase the demand of NICU and SCBU services. Thus, there is a need to reduce
the incidence rate of late onset sepsis in infant in my unit.
In addition, there is a knowledge gap between the updated researches and the
current practice in Hong Kong. Some nurses in the NICU and SCBU still have questions on
the effectiveness of the probiotic use in preterm infants. Although numerous studies
have shown that neonatal sepsis is strongly associated with increased mortality in
preterm infant and oral probiotic is an effective prophylaxis (Demirel et al., 2013; Jacobs
et al., 2013; Lin et al., 2005; Manzoni et al., 2009; Oncel et al., 2014; Samanta et al.,
2009), probiotic is still not widely use in Hong Kong. Some nurses in my unit still believed
that infection control measures such as good hand hygiene and aseptic technique during
procedures are the only ways to prevent neonatal sepsis. It may because of the lacking
of good quality translation research on this topic in Hong Kong. And there is no updated
- 4 -
systematic review on the effectiveness of probiotic regarding to the types of probiotic
used.
Although some hospitals in Hong Kong already use oral probiotic as a routine
prophylaxis in preterm infants, a well-establish guideline on probiotic use has yet widely
applied in Hong Kong. In review to the high neonatal late-onset sepsis rate in preterm
infants in my hospital, there is a need to develop an evidence based probiotic use
guideline.
Objective and Significance
Significance
The delay of gastrointestinal colonization in preterm infants causes slow
activation of intestinal mucosal immune response which increases chances of infection.
According to Mayer et al. (1999) and Woof et al. (2005), secretory immunoglobulin A
(sIgA) is a main product in the mucosal immune system that defense again dietary and
microbial antigens by performing immune exclusion through adhesion obstacles and
potential danger of antigen invasion into the mucosal tissue and neutralize toxins and
microbial pathogens virulent factors (as cited in Retnaningtyas et al., 2010). And
production of sIgA is trigger by the presence of microflora in intestine during the
neonatal period (Cebra et al., 1999 and Stagg et al., 2003, as cited in Retnaningtyas et al.,
2010). Based on the above theory, administration of probiotic can increase the serum
sIgA level as probiotics are live microorganisms that can accelerate the gastrointestinal
colonization proves in the host (Millar, Wilks & Costeloe., 2003). Thus, administration of
- 5 -
probiotic is suggested to be effective and important on preventing late onset sepsis in
preterm infants.
Objectives of this dissertation
1. To conduct a systematic literature review on the use oral probiotic in preterm
infants in preventing late onset sepsis
2. To integrate and synthesize the findings from most updated literatures found
3. To develop an evidence-based practice guideline on using probiotic in preterm
infants
4. To assess the implementation potential of the proposed guideline in the target
setting
5. To design an evaluation plan for the proposed guideline and make improvement
on current practice
Research question
An evidence-based guideline on using oral probiotic to prevent late onset sepsis
in preterm infants
- 6 -
Chapter 2: Critical Appraisal
To review the effectiveness of probiotic use in preventing late-onset sepsis in
preterm infants and to develop an evidence-based guideline on the use of probiotic, a
comprehensive systemic search was performed.
Search and Appraisal Strategies
A comprehensive literature search was performed in three medical electronic
databases, including the PubMed, Cochrane library and the British Nursing Index (BNI)
from July, 2015 to December, 2015.
Selection Criteria
Relevant articles were selected according to the following criteria.
Inclusion criteria:
Randomized controlled trials (RCTs);
Publication dates from 1st January 2005 to present (approximately 10 years);
Full text;
Studies which targeted in preterm infants with very low birth weight (Gestation
weeks 32 weeks or birth weight 1500 grams);
Received oral probiotic or lactoferrin as prophylaxis;
Outcome measure include incidence rate of late onset sepsis;
Studies performed in hospital settings
- 7 -
Exclusion criteria:
Studies that were not published in English
Preterm infants with major congenital and gastrointestinal anomalies or multiple
complications
Search strategy
Three groups of keywords (MeSH term) were used in the search strategy. The
first group of keywords, concerning about the study targets, included ‘premature infant’,
‘premature infants’, ‘preterm infant’, ‘preterm infants’, ‘neonatal prematurity’ and
‘prematurity’. The second group of keywords, concerning about the interventions,
included ‘lactoferrin’, ‘bovine lactoferrin’, ‘probiotic’ and ‘probiotics’. The third group of
keywords, concerning about the problem identified, including ‘neonatal sepsis’, neonatal
infection’, ‘neonatal infections’, ‘late onset sepsis’, ‘late onset infection’, ‘late onset
infections’ and ‘bacteremia’. The search operator ‘OR’ were applied within the same
group of keywords and the search operator ‘AND’ were applied between different
groups of keywords.
All potentially eligible studies were filtered by limiting to RCTs, publication dates
within 10 years (from 1st January 2005 to present) and full text. Then, the potentially
eligible studies were further limited by reviewing titles and abstracts. After that, content
of the potentially eligible studies were carefully reviewed and those do not match the
inclusion criteria were removed. After relevant studies were identified, a manual search
was performed by reviewing the reference lists of the selected studies. The PRISMA flow
chart of search strategies and the search result are summarized in Appendix A.
- 8 -
Appraisal strategy
To assess the methodological quality of selected studies, the methodology
checklist developed by the Scottish Intercollegiate Guidelines Network (SIGN) was used.
As all selected studies were RCTs, the methodology checklist 2 on controlled trials was
used (Appendix D). Information including method of randomization, concealment
allocation, blinding, and similarity between intervention and control group, validity and
reliability of outcome measures, dropout rate and the use of intention to treat analysis
were taken into account.
Results
Search results
After performing the keywords search stated above, 124 studies in PubMed, 67
studies in Cochrane library and 9 studies in BNI were identified. When limited to RCT,
publication dates from 1st January 2005 to present and full text, 16 studies in PubMed,
37 studies in Cochrane library and 1 study in BNI were identified. All 54 potentially
eligible studies were then screened by titles and abstracts to fulfill the inclusion and
exclusion criteria and 11 of the studies remained. After removal of duplicated studies in
the selected databases, 8 of the studies remained. The remaining 8 studies were then
screened by reading their full content carefully. Two studies were excluded because
their outcome measure do not include incidence rate of late onset sepsis which is our
primary outcome measure in this thesis. Another two studies were excluded because
their target groups do not meet our inclusion criteria. In order to find more potential
studies, the reference lists of the selected 4 studies were reviewed and 2 more relevant
- 9 -
studies were found. Finally, 6 studies were included for review. More details of the
systemic search are shown in the PRISMA flowchart diagram and the search strategies
table in Appendix A.
Data analysis: table of evidences
Data of the six studies found were extracted and summarized in a table of
evidence (Appendix B) to facilitate integrate and synthesize of the studies. The content
of table of evidence included citation of the studies, studies design, level of evidence,
characteristics of the study targets, intervention(s), control, outcome measure(s) and
results. The level of evidence of each study was graded according to the SIGN
Methodological Quality Coding: Levels of Evidence & Grading of Recommendations
(Appendix E).
All six studies yielded from the databases were RCTs which graded level 1
according to SIGN grading system. One of them is high quality RCTs (level 1++) with a
very low risk of bias (Manzoni et al., 2009). The remaining five studies are well-
conducted RCTs (level 1+) with a low risk of bias (Demirel et al., 2013; Jacobs et al., 2013;
Lin et al., 2005; Oncel et al., 2014; Samanta et al., 2009).
Summarize the appraisal results
As mentioned in above, SIGN checklist 2 on controlled trial was used to assess
the methodological quality of the selected studies.
- 10 -
Research question
All the six selected studies addressed an appropriate and clearly focused
question (Demirel et al., 2013; Jacobs et al., 2013; Lin et al., 2005; Manzoni et al., 2009;
Oncel et al., 2014; Samanta et al., 2009). The elements of a well-defined research
question including population, intervention, control and outcome (PICO) are found in
the title of the six selected studies.
Randomization
All six studies used a good randomization method. Three of them were
randomized by a random-number table sequence (Demirel et al., 2013; Lin et al., 2005;
Samanta et al., 2009). The remaining three studies were randomized by computer-
generated randomization lists (Jacobs et al., 2013; Manzoni et al., 2009; Oncel et al.,
2014). For studies that carried out in multiple sites, the randomization was stratified by
centers (Jacobs et al., 2013; Manzoni et al., 2009).
Concealment method
For the concealment method, four out of six studies reported their concealment
method. Three of them used opaque, sequentially numbered, and sealed envelopes to
maintain allocation concealment (Demirel et al., 2013; Lin et al., 2005; Oncel et al., 2014).
One study diluted all doses of the study powder or placebo in prepared milk by an
independent team that not involved in the care of the target group and analysis of the
study to maintain the allocation concealment (Manzoni et al., 2009). The remaining two
- 11 -
studies didn’t mention their concealment method which downgraded their study’s
quality (Jacobs et al., 2013; Samanta et al., 2009).
Blinding
As the target groups of this review are infants, all the treatment allocation were
‘blinded’ to the target population. Besides the target groups, all clinical staff and
investigators were remained unaware of the randomization in three of six studies
(Demirel et al., 2013; Jacobs et al., 2013; Manzoni et al., 2009). In Oncel et al. (2014), the
author mentioned that it was a double blinded study with all clinical staff was blinded.
However, Oncel et al. (2014) didn’t mention about whether investigators were blinded
to the randomization as well. In Samanta et al. (2009), the author also stated that it was
a double blinded study. However, whether investigators or clinical staff was blinded was
not clearly mentioned. Thus, the presences of blinding were not clear in studies of Oncel
et al. (2014) and Samanta et al. (2009). For Lin et al. (2005), only clinical staff was
remained unaware of the randomization, investigators were not blinded. This single
blinded study design may create bias on interventions and thus downgraded the study
quality.
Demographic data between treatment and control groups
All of the six studies compared the maternal clinical and infant’s demographic
and clinical characteristics before the start of trial. Five out of six studies stated that the
background characteristics of intervention and control groups were similar (Demirel et
al., 2013; Lin et al., 2005; Manzoni et al., 2009; Oncel et al., 2014; Samanta et al., 2009)
and three of the studies provided the p-values (Manzoni et al., 2009; Oncel et al., 2014;
- 12 -
Samanta et al., 2009). In the studies of Oncel et al. (2014), the duration of total parental
nutrition in both group were statistically significant with p-value = 0.048. Only Jacobs et
al. (2013) didn’t state that the intervention and control groups were similar but both
groups look reasonably similar in the characteristic table.
Also, the intervention under investigation were appeared to be the only
difference between intervention groups and control group of each study obtained
(Demirel et al., 2013; Jacobs et al., 2013; Lin et al., 2005; Manzoni et al., 2009; Oncel et
al., 2014; Samanta et al., 2009).
Validity and reliability of all relevant outcome measure
All of the six studies clearly stated and used valid and reliable assessment tools to
measure the outcomes. All of the six studies used positive blood culture or CSF culture
result to define sepsis (Demirel et al., 2013; Jacobs et al., 2013; Lin et al., 2005; Manzoni
et al., 2009; Oncel et al., 2014; Samanta et al., 2009). And five out of six studies use Bell’s
classification in the outcome measurement of NEC (Demirel et al., 2013; Jacobs et al.,
2013; Lin et al., 2005; Manzoni et al., 2009; Samanta et al., 2009). Measurements, such
as mortality rate, duration of hospitalization and days reached full enteral feeding were
recorded in the medical records from hospitals accordingly.
Dropout rate and Intention to treat
The dropout rates of all six studies are considered as acceptable which ranged
from 0% to 6.1%. Intention to treat analysis was used in two out of six studies (Jacobs et
al., 2013; Manzoni et al., 2009). Although another two studies didn’t mention whether
- 13 -
intention to treat analysis was used, the numbers of subjects being analyzed match the
numbers of subjects recruited in each group (Lin et al., 2005; Samanta et al., 2009).
However, I still can’t conclude whether intention to treat was used. Oncel et al. (2014)
and Demirel et al. (2013) didn’t use intention to treat analysis as infants who
discontinued intervention were not included into the analysis.
Single/ multiple site(s) study
Two studies were carried out in multiple sites (Jacobs et al., 2013; Manzoni et al.,
2009) and the results are comparable for all sites. The remaining four studies were
carried out in single site (Demirel et al., 2013; Lin et al., 2005; Oncel et al., 2014;
Samanta et al., 2009).
A detailed appraisal result is concluded in Appendix F and supplementary
information was provided in Appendix G.
Summary and Synthesis
Summary of data
Results of the six studies found were extracted and briefly summarized in a table
(Appendix C) to facilitate an organized comparison between studies. And a more
detailed summary of each study was described in Appendix B.
Patient’s characteristics
In all six studies yielded, infants with birth weight 1500grams were recruited
with mean birth weight ranged from 1048g to 1210g (Demirel et al., 2013; Jacobs et al.,
2013; Lin et al., 2005; Manzoni et al., 2009; Oncel et al., 2014; Samanta et al., 2009).
- 14 -
Four studies also recruited infants with gestation weeks 32 weeks (Demirel et al., 2013;
Jacobs et al., 2013; Oncel et al., 2014; Samanta et al., 2009). The mean gestation weeks
of all studies were ranged from 27.8 to 30.14 weeks.
Four of the studies excluded infants who were with major congenital anomalies
(Demirel et al., 2013; Jacobs et al., 2013; Oncel et al., 2014; Samanta et al., 2009).
All six studies were carried out in different countries: Taiwan (Lin et al., 2005);
India (Samanta et al., 2009); Italy (Manzoni et al., 2009); Australia and New Zealand
(Jacobs et al., 2013) and Turkey (Demirel et al., 2013; Oncel et al., 2014).
Study agent and control agent
In the selected studies, various types of probiotic, administration frequency and
daily dosage were applied. Four studies used probiotic including the Lactobacilus species
(Lin et al., 2005; Manzoni et al., 2009; Oncel et al., 2014; Samanta et al., 2009). Three
studies used probiotic including the Bifidobacterium species (Jacobs et al., 2013; Lin et
al., 2005; Samanta et al., 2009). One study included the Streptococcus species (Jacobs et
al., 2013), another one study included bovine lactoferrin as study agent (Manzoni et al.,
2009) and one study included Saccharomyces species (Demirel et al., 2013)
Three studies administrated the probiotic once daily (Demirel et al., 2013;
Manzoni et al., 2009; Oncel et al., 2014) and another two studies would administrate
probiotic twice daily (Lin et al., 2005; Samanta et al., 2009). Jacobs et al. (2013)
suggested administrating the probiotic according to the study targets’ feeding amount. If
- 15 -
the infant can take more than 3 ml milk each feeding, daily dose was suggested. If the
infant can only take less than 3 ml milk each feeding, twice daily dose was suggested.
The total daily doses various from studies. Lin et al. (2005) and Samanta et al.
(2009) suggested dosage according to infant’s body weight with 125mg/kg and 125g/ kg
of study agent respectively and the other four studies suggested a fixed daily dose with
Saccharomyces boulardii 125mg/day (Demirel et al.,2013), bovine lactoferrin 100mg and
Lactobacillus rhamnosus GG(LGG) 6x 109 CFU/day (Jacobs et al., 2013), Bifidobacterium
infantis, Streptococcus thermophiles and Bifidobacterium lactis, 1 x 109 CFU mixed in a
maltodextrin base powder 1.5g (Manzoni et al., 2009) and 5 drops of oil-based
Lactobacillus reuteri x 108 CFU (Oncel et al., 2014).
For the control group, Demirel et al. (2013) used breast milk or formula milk as
placebo, while Lin et al. (2005) and Samanta et al. (2009) used breast milk as placebo.
Manzoni et al. (2009) added 2 ml of 5% glucose solution into breast milk or formula milk
as placebo. Jacobs et al. (2013) added maltodextrin base powder into breast milk or
formula milk as placebo. Finally, Oncel et al. (2014) used oil-based suspension as placebo
in control group.
Duration of treatment
For the duration of treatment, all six studies initiated the treatment when infants
started enteral feeding. Lin et al. (2005) commenced the treatment for infants who had
removed the indwelling umbilical artery or umbilical venous catheter for at least 24
hours. Manzoni et al. (2009) performed the treatment to infants who started feeding
- 16 -
after 3rd day of life. Jacobs et al. (2013) required infants to receive at least 1 ml of milk
every 4 hours to start the treatment.
Three studies ended the treatment when infants discharged from hospital
(Demirel et al., 2013; Lin et al., 2005; Oncel et al., 2014; Samanta et al., 2009). Jacobs et
al. (2013) ended the treatment when infants discharged from hospital or reached the
term corrected age. Manzoni et al. (2009) ended the treatment when infants completed
a 6 weeks (birth weight <100g) or 4 weeks (birth weight 1001-1500g) probiotic courses
respectively.
Methods to deliver treatment
Five studies mixed the study agent or placebo with breast milk or formula milk
(Demirel et al., 2013; Jacobs et al., 2013; Lin et al., 2005; Manzoni et al., 2009; Samanta
et al., 2009). Only Oncel et al. (2014) delivered probiotic in oil-based suspension. All six
studies delivered probiotic orally (Demirel et al., 2013; Jacobs et al., 2013; Lin et al.,
2005; Manzoni et al., 2009; Oncel et al., 2014; Samanta et al., 2009) and three of them
used oro-gastric tube feeding when oral feeding not feasible (Jacobs et al., 2013;
Manzoni et al., 2009; Oncel et al., 2014).
Outcome measure
All six studies included the incidence rate of culture proven sepsis and NEC ≥
stage 2 according to Bell’s classification as their outcome measures (Demirel et al., 2013;
Jacobs et al., 2013; Lin et al., 2005; Manzoni et al., 2009; Oncel et al., 2014; Samanta et
al., 2009). Sepsis were either proven by one positive blood culture (Demirel et al., 2013;
- 17 -
Lin et al., 2005; Samanta et al., 2009) or two separate positive blood culture results
(Jacobs et al., 2013; Manzoni et al., 2009; Oncel et al., 2014) or positive CSF culture
result (Demirel et al., 2013; Jacobs et al., 2013; Manzoni et al., 2009; Samanta et al.,
2009). Two studies included the incidence rate of clinical sepsis (Demirel et al., 2013;
Jacobs et al., 2013). Demirel et al. (2013) defined clinical sepsis by CRP >4.82 mg/L and
Interleukin-6 >4.82mg/L and Jacobs et al. (2013) defined clinical sepsis by CRP >10mg/L
or immature-to-total neutrophil ratio>0.2 and treated with antibiotic ≥ 5days.
Bell’s classification ≥ stage 2 was used to diagnose NEC in all studies found.
Demirel et al. (2013), Lin et al. (2005) and Oncel et al. (2014) required two independent
neonatologists to diagnose NEC. If the two doctors had different opinions, a third
neonatologist would make the decision.
Five out of six studies included mortality rate (Demirel et al., 2013; Jacobs et al.,
2013; Lin et al., 2005; Manzoni et al., 2009; 2014; Samanta et al., 2009). Three of the
studies included days required full enteral feeding and length of hospital stay (Jacobs et
al., 2013; Oncel et al., 2014; Samanta et al., 2009). The above three outcomes were
measured according to hospital’s medical records.
Results and significance
Four out of six studies have shown a statistically significance in reducing
incidence rate of culture proven late onset sepsis in infants (Lin et al., 2005; Manzoni et
al., 2009; Oncel et al., 2014; Samanta et al., 2009). Jacobs et al. (2013) reported that
probiotic can significantly reduce culture proven late onset sepsis in infants with
- 18 -
gestation ≥28 weeks but not in those <28 weeks and Demirel et al. (3013) reported that
probiotic cannot significantly reduce culture proven sepsis.
For clinical sepsis, Demirel et al. (2013) reported that it was statistically
significant in reducing the incidence of clinical sepsis while Jacobs et al. (2013) reported
that it was not statistically significant in reducing the incidence of clinical sepsis.
Five out of six studies have shown a statistically significance in reducing incidence
rate of NEC≥ stage 2 (Jacobs et al., 2013; Lin et al., 2005; Manzoni et al., 2009; Oncel et
al., 2014; Samanta et al., 2009).
Among all six studies, five of them studied the effect of probiotic on mortality.
Three studies have shown a statistically significance in reducing infant mortality (Lin et
al., 2005; Manzoni et al., 2009; Samanta et al., 2009). However, Demirel et al. (2013) and
Jacobs et al. (2013) pointed out that it was not statistically significant in reducing
mortality caused by NEC.
Among all six studies, three of them studies the days required full enter feeding
and length of hospital stay. Samanta et al. (2009) and Oncel et al. (2014) supported the
effect of probiotic, while Jacobs et al. (2013) stated that probiotic has no statistically
significance in these two outcome measures.
As the incidence of late onset sepsis, NEC and morality were presented in
percentage, the effects of these three outcomes are presented in relative risk reduction
- 19 -
(RRR). All the p-values and RRR of relevant outcomes of all studies are summarized in
the table below.
Outcome measure(s) Result
Lin et al. (2005) 1. Sepsis (culture proven) 2. NEC≥ stage 2 3. Mortality
1. p=0.03, RRR=36.2% 2. p=0.04, RRR= 81.3% 3. p=0.009, RRR= 63.6%
Samanta et al. (2009)
1. Sepsis (culture proven) 2. NEC≥ stage 2 3. Mortality 4. Days required full enteral
feeding 5. Length of hospital stay
1. p=0.020, RRR=93% 2. p=0.042, RRR=51.5% 3. p=0.032, RRR=70.1% 4. p=<0.001 5. p=<0.001
Manzoni et al. (2009) 1. Sepsis (culture proven) 2. NEC≥ stage 2 3. Mortality attributable to sepsis
1. p=< 0.001, RRR=73.4% 2. p=0.04, RRR=85.4% 3. p=0.002, RRR=100%
Demirel et al. (2013) 1. Sepsis (clinical) 2. Sepsis (culture proven) 3. NEC≥ stage 2 4. Mortality
1. p=0.03,RRR=41.7% 2. p=0.906,RRR=4.76% 3. p=1.0,RRR=14.3% 4. p=1.0,RRI=0.770%
Jacobs et al. (2013)
1. Sepsis (≥ 28weeks) (culture proven)
2. Sepsis (<28 weeks) (culture proven)
3. Sepsis (clinical) 4. NEC≥ stage 2 5. Mortality ( NEC) 6. Days required full enteral
feeding 7. Length of hospital stay
1. p=0.01, RRR=49.1% 2. p=0.75, RRI=5.5% 3. p=0.52, RRR=9.3% 4. p=0.03, RRR=54.5% 5. p=0.07, RRR=65.0% 6. p=0.09 7. p=0.31
Oncel et al.(2014)
1. Sepsis(culture proven) 2. NEC≥ stage 2 3. Days required full enteral
feeding 4. Length of hospital stay
1. p=0.041, RRR=48% 2. p=0.63, RRR=20% 3. p=0.006 4. p=0.022
p = p-value, RRR= relative risk reduction, RRI= relative risk increased
- 20 -
Synthesis of data
Majority of studies support the effect of probiotics on reducing the incidence
rate of culture proven late onset sepsis. The findings were integrated and synthesized to
develop new guidelines on the probiotic use in preterm infants as follow.
Majority of the studies were carried out in developing counties (Demirel et al.,
2013; Lin et al., 2005; Oncel et al., 2014; Samanta et al., 2009).
All of the studies included infants with birth weight < 1500g (Demirel et al., 2013;
Jacobs et al., 2013; Lin et al., 2005; Manzoni et al., 2009; Oncel et al., 2014; Samanta et
al., 2009). And majority of studies included infants with gestational age < 32 weeks
(Demirel et al., 2013; Jacobs et al., 2013; Oncel et al., 2014; Samanta et al., 2009),
started enteral feeding (Demirel et al., 2013; Lin et al., 2005; Oncel et al., 2014; Samanta
et al., 2009) and excluded those with major congenital anomalies (Demirel et al., 2013;
Jacobs et al., 2013; Oncel et al., 2014; Samanta et al., 2009). Other demographic data
such as sex, mode of delivery and Apgar scores of infants across studies were similar and
have no effect on the results. Therefore, the probiotic therapy could be implemented in
infants with gestation weeks 32 weeks or birth weight < 1500g irrespective of sex,
mode of delivery and Apgar scores.
In term of the type of probiotic used, majority of studies included probiotics
contained either Lactobacillus species (Manzoni et al., 2009; Oncel et al., 2014) or
Bifidobacterium species (Jacobs et al., 2013) or both (Lin et al., 2005; Samanta et al.,
2009). When comparing the effect in term of RRR, Bifidobacteria infantis, Bifidobacteria
- 21 -
bifidum, Bifidobacteria longum and Lactobacillus acidophilus, each 2.5 x 109 CFU
125g/kg twice daily gave the greatest effect on reducing incidence rate of late onset
sepsis, followed by Bovine lactoferrin (100mg) with LGG 6x 109 CFU and Bifidobacterium
infantis, Streptococcus thermophiles and Bifidobacterium lactis, 1 x 109 CFU. Thus,
Bifidobacteria infantis, Bifidobacteria bifidum, Bifidobacteria longum and Lactobacillus
acidophilus, each 2.5 x 109 CFU will be chosen in our guideline and will be given in
125g/kg twice daily as suggested.
Majority of studies suggested mixing probiotic with breast milk or formula milk
(Demirel et al., 2013; Jacobs et al., 2013; Lin et al., 2005; Manzoni et al., 2009; Samanta
et al., 2009) before giving to infants orally (Demirel et al., 2013; Jacobs et al., 2013; Lin et
al., 2005; Manzoni et al., 2009; Oncel et al., 2014; Samanta et al., 2009) or through oro-
gastric tube with milk (Jacobs et al., 2013; Manzoni et al., 2009; Oncel et al., 2014). It is
believed that the probiotic give its effect at the gut. Therefore giving probiotic orally or
through oro-gastric tube is feasible.
Although the above practices are recommended using an evidence-based
approach, we need to assess the implementation potential in Hong Kong hospital setting,
which will be discussed in Chapter 3. Also, an effective evaluation plan should be
prepared for this new innovation, which will be discussed in Chapter 4.
- 22 -
Chapter 3: Implementation Potential and Clinical Guideline
The previous chapters have suggested that using oral probiotic in preterm infants
was effective in preventing late-onset sepsis. In this chapter, the implementation
potential of the proposed innovation in my workplace will be assessed by the
transferability, feasibility and cost-benefit ratio. And an evidence-based practice (EBP)
guideline will be developed at the end.
Target population and setting
The innovation of using oral probiotic in preterm infants will be carried out in the
NICU and SCBU of an acute hospital managed by the Hospital Authority (HA) in Hong
Kong. The NICU residents are mainly preterm infants. Around 70% of them are
extremely preterm or very preterm infants with birth weight ≤ 1500 grams. Most of
them would transfer to SCBU when their conditions are stable for further management.
In the target setting, there are 10 NICU beds and 18 SCBU beds. The NICU
admissions mainly include extremely or very preterm or very low birth weight infants,
infants with severe medical conditions that require urgent medical or surgical
interventions. Referrals are mainly from the labor ward and postnatal ward of my own
hospital or from the NICU of other local hospitals. For the SCBU, the admissions mainly
include late preterm infants (gestation week 32-37) and term infants will comparably
stable medical conditions such as neonatal hypoglycemia, jaundice, clinical sepsis etc. By
general observation, around 6 out of 18 beds in SCBU would admit condition stable
preterm infants transferred from NICU for further management.
- 23 -
According to my observation, there are estimated 200 NICU admissions each
year in my workplace. 70% of the admissions meet the criteria of the target population
mentioned above. That means there are around 140 infants could benefit from the
innovation each year.
Transferability
The similarity of target population, target setting and philosophy of care in my
workplace are compared with those in the studies mentioned in Chapter 2 when
assessing the transferability of the innovation.
Similarity of the target setting
The innovation will be carried out in an acute NICU and SCBU in Hong Kong with
10 NICU beds and 18 SCBU beds. According to the United Nations (2016), Hong Kong is
developing region. Four out of the six studies mentioned in Chapter 2 were carried out
in developing countries also (Demirel et al., 2013; Lin et al., 2005; Oncel et al., 2014;
Samanta et al., 2009) and two of the studies were carried out in developed countries
(Jacobs et al., 2013; Manzoni et al., 2009). The equipment and medical technology of the
proposed target setting are similar to those hospitals in the studies.
Similarity of the target population
According to the data synthesis in Chapter 2, majority of the studies recruited
infants with gestation weeks 32 weeks (Demirel et al., 2013; Jacobs et al., 2013; Oncel
et al., 2014; Samanta et al., 2009). And all studies recruited infants with birth weight
1500grams (Demirel et al., 2013; Jacobs et al., 2013; Lin et al., 2005; Manzoni et al.,
- 24 -
2009; Oncel et al., 2014; Samanta et al., 2009). Also, majority of the studies excluded
infants who were with major congenital anomalies (Demirel et al., 2013; Jacobs et al.,
2013; Oncel et al., 2014; Samanta et al., 2009). Hence, the characteristics of targets in
the proposed setting are similar to those in the studies reviewed.
Philosophy of care
Hospital Authority’s mission is ‘Helping People Stay Healthy’ (Hospital Authority,
2016) and its values emphasize on people-centred care, professional service, committed
staff and teamwork (Hospital Authority, 2016). HA emphasizes on providing the best
quality and people-centered care service to improve patients’ health status through the
teamwork of multidiscipline. In addition, all six studies aimed to provide an evidenced-
based practice to reduce the mortality rate and morbidity rate of preterm infants and
facilitating their development. (Demirel et al., 2013; Jacobs et al., 2013; Lin et al., 2005;
Manzoni et al., 2009; Oncel et al., 2014; Samanta et al., 2009). As my workplace is one of
the HA hospital, we follow HA’s mission and values which match with the philosophy of
care in all six studies found. With similar philosophies of care, the innovation can be
transferred to my workplace for application.
Implementation plan and evaluation time
A 16 weeks pilot study will be launched after 2 weeks’ preparation. Department
approval, materials needed, guidelines and staff training will be prepared in the
preparatory phase. And during the pilot study, feedback from staff will be collected to
assess their compliance and the guidelines will be refined if necessary. Then, a 25 weeks
main study will be carried out as planned. For the evaluation, pilot study evaluation will
- 25 -
be held in week 8-18 after collecting the first slot of data and main study evaluation will
held in week 24-43 to analyze the collected data and prepare the report after the main
study completed. A brief schedule for implementation and evaluation plan is shown in
Appendix J and details will be discussed in Chapter 4.
Feasibility
Besides transferability of the innovation, checking for feasibility is also a key
element of successful implementation.
Freedom to carry out the innovation
Nurses have autonomy and are encouraged to carry out new innovations in my
workplace. In the past years, different new innovations initiated by nurses were carried
out, including the use of sucrose solution in relieving pain during procedures, wearing
personal protective equipment (PPE) to reduce disease transmission when touching
infants and using double phototherapy instead of single phototherapy to treat neonatal
jaundice to shorten infant’s hospitalization duration and readmission rate.
Although probiotics are not medicine, they are supplement that contains live
bacteria and yeasts. Adding supplements into infant’s milk need consensus from the
physicians. In my workplace, different disciplines of the healthcare team work
corporately with each other. Like I mentioned in the above, the innovation of using
double phototherapy to treat neonatal jaundice successfully obtained the consensus
from the physicians. We have regular multi-disciplinary team meetings to ensure good
communication and co-operation in the pediatric department.
- 26 -
Support from administration and organization climate
Nurse autonomy to carry out new innovation is supported in organization level.
HA has been organizing the Hospital Authority Convention every year since 1993. One of
the focuses of the convention is to promote sharing of expert views, ideas and
experience among healthcare professionals. Nurses in my unit are welcome to submit
and present the abstracts of new innovations in the convention to other colleagues. The
best abstract will be awarded every year. Besides that, HA supported us to propose and
carry out evidence based practices. It provides electronic knowledge gateway (eKG) to
facilitate evidence based learning and researches. eKG allows colleagues to access
different medical databases easily which can provide us valuable insight to inspire
innovative ideas and facilitate our research projects.
In administrative level, the Chief of Service (COS), the Department Operation
Manager (DOM) and the Ward Manager (WM) also support nurses to carry out new
evidence based nursing practices. They are currently support us to implement the
guideline to relieve pain of infants by sucrose solution and also the double phototherapy
guideline for jaundice etc. Furthermore, an evidence based research team was
established few years ago in my department. Few projects are now in progress. For
example, the use of iPad in children during blood taking for distraction and using
colostrum as oral flush in neonates. In addition, every staff needs to share at least one
nursing related topic once a year to enhance and update colleagues’ knowledge. With a
supportive environment, the proposed innovation is feasible to carry out.
- 27 -
Consensus among the staff
Nevertheless, most of my colleagues including physicians are willing to provide
better services that can benefit patients. The benefits of the proposed innovation have
good evidence based. It is easy to get nurses and physicians’ cooperation if the
innovation shows benefits to our work and patients. The proposed innovation can lower
the incidence rate of late-onset sepsis in infants and thus decreased staff workload and
increased our satisfaction on work. Comments will also be collected before starting the
pilot study.
Interference with the current staff workload
The proposed innovation is mainly performed by nurses and causes only minimal
interference to the current staff workload when preparing the infants’ milk. One of the
nurses’ responsibilities in NICU and SCBU is to add different kind of supplements into
infants’ milk. For example, medications, polycal powder and human milk fortifier
powder etc. Therefore, we can cluster the proposed innovation with the current staff
routine of adding supplement to minimize unnecessary interference.
Equipment and facilities needed
Most of the equipment and facilities are already available in NICU and SCBU as
follow:
A clean kitchen for staff to prepare infants’ milk;
A up to standard refrigerator to keep human milk and formula milk;
- 28 -
A clean store room to keep the probiotic;
A spare room with chairs and projector for staff training;
Computers, photocopying machine needed to prepare posters, guidelines etc.
Equipment that needed to purchase includes probiotic, stationaries needed to
prepare posters, guidelines. The amount and the cost of equipment will be discussion in
the ‘Cost-Benefit Ratio’ later.
Evaluation tools
For the evaluation, the effectiveness of the probiotic will be measured by the
incidence rate of late onset sepsis among target population as suggested by the six
studies found (Demirel et al., 2013; Jacobs et al., 2013; Lin et al., 2005; Manzoni et al.,
2009; Oncel et al., 2014; Samanta et al., 2009). Blood culture will be performed in
infants with risks or signs of infection and lumbar puncture will be performed in infants
with persistent high fever or severe signs of infections such as relatively high CRP level.
Blood or CSF specimen will be collected under a well-established guideline by trained
physicians and sent to the hospital laboratory for investigations. The neonatal sepsis is
proven by positive blood or CSF culture. In addition, this study will also evaluated by
staffs using questionnaires. A detailed evaluation plan will be discussed in Chapter 4.
Barriers on implementing the innovation
Several potential barriers may appear when we carry out the innovation. Firstly,
the nurses may resistant to change the current practice due to worry about the
- 29 -
uncertainness, increase of workload and knowledge gap on the effectiveness of using
probiotic. Secondly, the proposed innovation needed inter-department coordination,
where we need to discuss with the physicians as well. Finally, we need to explain the
new innovation and obtain the consensus from the parents.
Cost-Benefit Ratio
The third key element of implementation potential is to measure cost-benefit
ratio and it will be assessed by the following aspects.
Potential risks of innovation
No risks were reported from the studies found (Demirel et al., 2013; Jacobs et al.,
2013; Lin et al., 2005; Manzoni et al., 2009; Oncel et al., 2014; Samanta et al., 2009).
However, improper handling of probiotic may contaminate infants’ milk. Therefore, it is
important to train the staff and develop a clear and user-friendly guideline.
Potential benefits of innovation
Oral probiotic can reduce the incidence rate of late onset sepsis (Jacobs et al.,
2013; Lin et al., 2005; Manzoni et al., 2009; Oncel et al., 2014; Samanta et al., 2009).
And it is proven that probiotic can also significantly decrease extremely preterm and
very preterm infant’s rate of developing NEC (Jacobs et al., 2013; Lin et al., 2005;
Manzoni et al., 2009; Samanta et al., 2009), mortality (Lin et al., 2005; Manzoni et al.,
2009; Oncel et al., 2014; Samanta et al., 2009), the time needed to reach full enteral
feeding (Oncel et al., 2014; Samanta et al., 2009) and the length of hospital stay (Oncel
et al., 2014; Samanta et al., 2009). As a result, the medical expenses and staff workload
- 30 -
would decrease. Furthermore, when the length of hospital stay and the incidence of
infection minimized, the infants can suffer less from the medical investigations which
can increase nurses and parents satisfaction. And thus increase hospital and
department’s prestige.
Risks of not trying the proposed innovation
On the other hand, if this innovation could not be promoted in my department,
the incidence rate of late onset sepsis would maintain in a high level which will increase
staff workload and decrease their job satisfaction. When the infants need to stay longer
in the NICU or SCBU, it will decrease the bonding between parents and their child as
their parental time is restricted by short visiting hours in NICU and SCBU.
Costs of the innovation
The cost of the innovation included set-up cost, material cost, non-material cost
and running cost in the subsequent years. A detailed table listing the cost of the
innovation and subsequent running cost is shown in Appendix H.
Set-up cost
The costs of preparing training materials, paper guideline and posters and
materials for evaluations staff are estimated to be $300 each year. This include the cost
of ink and the paper needed, as printer, photocopying machine, computers, projector
and facilities for training are already existing in my workplace. An electronic guideline
will also be uploaded to department’s intranet where the intranet already exists.
- 31 -
Material cost
Firstly, the material cost included the probiotic, which suggested a twice daily
dose during the whole hospitalization period of target infants. A bottle of probiotic
contained 90 capsules which costs HK$245. We estimated that there will be 140 infants
that can benefit from the innovation each year and the average hospitalization period of
each target infant will be 60 days. The estimated costs of probiotic will be HK$245 /90 x
140 infants x 2 (twice daily) x 60 days = HK$ 45734 annually.
Secondly, the material cost includes the staff training time and the time needed
to carry out the new practice, which can be expressed in salary. There are around 40
NICU nurses and 15 SCBU nurses in the target setting and the average salary is around
HK$300 each hour. For the training, each staff would attend a 30 minutes training
session. The estimated cost for training = 0.5 hour x 55 staff x $300 = $8250 annually.
For the time using to carry out the new practice, one staff is responsible for the milk
preparation each time. Around 15 minutes is needed twice daily for adding probiotic
supplement. The estimated cost = 0.25 x 2 x 365 days x $300 =HK$54750 annually.
Non-material cost
For the non-material costs, it includes different stakeholders’ attitude,
compliance and satisfaction level towards the innovation. Frontline staffs including
nurses and physicians may feel stressful and resistance to change when at the beginning
when the new innovation implied. Parents may feel worried to the innovation which
needed extra explanations and causes additional workload to the frontline staff.
- 32 -
Therefore, we should have good communication between different stakeholders prior to
the implementation. A detailed communication plan will be discussed in Chapter 4
Running cost
After the successful implementation of the innovation in the first year, the
running cost of subsequent years include cost of new staff training, cost of probiotic
each year and cost of time that will be used to prepare the milk. It is estimated 10 new
staff will join the NICU and SCBU in each year. The cost for new staff training is
estimated to be HK$1510. The cost of probiotic and time to prepare the milk are similar
each year, which are HK$45734 and HK$54750 respectively.
In conclusion, the cost of innovation in the first year is estimated to be
HK$109034. And the additional annual running cost in each subsequent year is
HK$101994.
Evidence-Based Practice Guideline
Late-onset sepsis is a common problem in extremely preterm and very preterm
infants and oral probiotic is shown to be effective in preventing it. As discussed in the
above, the proposed innovation is transferable and feasible in my workplace. Therefore,
a clinical guideline on using oral probiotic in preterm infants (Appendix I) will be
developed and the recommendations will be graded according to SIGN in Appendix E.
- 33 -
Chapter 4: Implementation Plan
The previous chapter has discussed the implementation potential of the
proposed innovation and developed an evidence based guideline on using oral probiotic
in preterm infants. In this chapter, the implementation plan of the innovation including
the timelines, communication plan, pilot study plan and evaluation plan will be further
discussed. The timelines of a one year implementation plan is shown in Appendix J and
Appendix K. And the basis for implementation will be proposed in the end.
Communication Plan
According to Melnyk & Fineout-Overholt (2005), a good communication plan
should start with identifying stakeholders who are affected by or may affect the
proposed changes or anticipated results of the proposed innovation and prioritize them
according to their influences to the innovation.
Identifying the stakeholders
The stakeholders of a new innovation include the administrators, intended users
of the guideline, other healthcare professionals and the beneficiaries.
The administrators of the pediatric department include the COS, the DOM, the
WM and the Nurse Specialist (NS). They are responsible to approve the proposed
innovation to be implemented in the department and allocate all related resources.
- 34 -
The intended users of the guideline include all nurses in the NICU and SCBU of
the pediatric department. All the nurses are responsible to carry out and evaluate the
effectiveness of the guideline.
Other healthcare professionals include doctors in the pediatric department. We
need their cooperativeness to the guideline as they are responsible to decide when the
preterm infants can start enteral feeding and examine the condition of the infants. Also,
they may worry about the effectiveness of the innovation. Therefore, we need to explain
and communicate with them in detail before implementation.
For the beneficiaries, they are the preterm infants in NICU and SCBU and their
parents. The infants are the one who benefit from the innovation and their parents are
their guardians who have authority to accept or refuse their babies to participate in the
innovation.
Communication process and strategies
Communication with the administrators
In week 1, we will seek the approval from the administrators. To propose a new
innovation, the NS of the pediatric department should be the first one to be reached as
she is the person who responsible to initiate and launch new innovations. A proposal will
be sent to the NS, explaining the affirming needs of the department, objectives of the
innovation, significance of the problem, the evidences and the implementation potential
of the innovation.
- 35 -
After getting feedback and approval from the NS, a revised proposal will be
presented to other administrators in departmental meeting. Besides the administrators
mentioned in the above, other senior medical officers and nursing officers are
coordinately invited to attend the meeting. In the meeting, a detailed proposal,
including the objective, significance, implementation potential, the proposed clinical
guideline and the implementation plan will be presented and discussed.
A committee will then be formed to launch the innovation. The committee
members include the NS, 3 Advanced Practice Nurses (APN) and 3 experienced
Registered Nurses (RN). The NS is responsible to allocate available resources and
negotiate with other administrators, the APNs are responsible for staff training and
mentoring and the RNs are responsible for the preparation and evaluation work. The
details of proposal will be explained to the committee members in the first committee
meeting. Then, regularly meetings will be held to refine details of the proposal and
monitor the process of implementation.
Communication with the intended users
Four identical briefing sessions will be held to all NICU and SCBU nurses in week 2
to gain their support and minimize the resistance of changing. During the briefing
sessions, the background, needs of change, the cost and benefit ratio of the innovation
and the timelines of the implementation will be explained clearly in an evidence based
manner using PowerPoint. To minimize the staff’s resistance to change practice, they are
reassured that only minimal influence to their work and the outcomes of the innovation
would decrease their workload with improving nursing care quality. The new guideline
- 36 -
will be introduced to all nurses during the briefing sessions as well. Each step of the
guideline will be elaborated using words, pictures and demonstration. All staff need to
do a return demonstration and are welcome to ask any questions at the end of the
briefing sessions.
Communication with the other healthcare professionals
To gain support and cooperation from other healthcare professionals, the
innovation will be introduced and explained in the multi-disciplinary team meeting in
week 2. And an email with the proposal, the guideline and the six evidence based
studies attached will be sent to everyone in the department.
Communication with the beneficiaries
The innovation will be introduced to eligible parents when their babies are
admitted and/ or when their babies started enteral feeding. A verbal consent will be
obtained from parents before intervention start. Benefits and risks of the innovation will
be explained to the parents by nurse and they are welcome to ask questions about the
innovation.
Guiding and Sustaining the Changing
To guide the change, we need to provide a clear vision to all the stakeholders.
Besides the briefing sessions, the APNs in the committee will train the nurses-in-charge
in ward to be trainers and role models and make sure their competency on applying the
guideline. Also, the committee members will become the troubleshooters and resource
persons so that other nurses in ward can seek help from them and give opinions anytime.
- 37 -
In addition, a resource manual will be provided and keep updating throughout the
implementation phase. Updated information will be announced in nurse handover each
shift.
To sustain the change, a hardcopy of guideline will be placed in the nursing
station and an electronic copy will be uploaded to department intranet for quick
reference. The committee members will also help to monitor the compliance and the
competency of ward nurses. Revisions to the guideline will be provided if necessary.
Feedback of the staff will be collected in the pilot study.
Pilot Study Plan
A pilot study will be carried out to determine the feasibility of the proposed
innovation in the target setting.
Objective
1. To determine the feasibility and identify any unexpected difficulties of the
proposed innovation.
2. To assess the acceptability of the proposed innovation by nurses, doctors and
parents of preterm infants.
3. To detect any adverse events or safety concerns of the proposed innovation.
4. To test the data collecting methods, including collection of clinical data and
feedback from all stakeholders.
- 38 -
5. To assess the time and budget problems that can occur during the future main
study.
6. To determine whether revisions are needed before implementing the main study.
Target population and recruitment
In the pilot study, eligible subjects will be recruited based on the
recommendations stated in the proposed guideline using convenience sampling method.
Preterm infants who admitted to NICU or SCBU with gestation week 32 weeks and/or
with birth weight 1500 grams with no major congenital anomalies will be recruited
when they start enteral feeding. Parental verbal consents will be obtained before
intervention start.
Timeframe for pilot study
A 16 weeks pilot study (week 3-18) will be launched after the briefing sessions
(Appendix J and Appendix K). The first 6 weeks will be the recruitment period. According
to the admission statistic in the target setting last year, it is estimated 15 eligible
subjects will be recruited in 6 weeks’ time. Interventions will be started after the first
subject being recruited. As the average length of hospitalization of eligible subjects is
around 8 weeks, the last subjects will complete the intervention at around week 16. For
the evaluation, it will be started in week 8-18, refined clinical guideline based on the
pilot test results will be presented in week 18.
- 39 -
Data collection and outcomes measurement
The pilot study will be carried out by the trained NICU and SCBU nurses according
to the proposed guideline. The outcome measures of the pilot study include the
incidence rate of late-onset sepsis; incidence rate of NEC; the length of hospitalization
and the staff satisfaction level. The incidence rate of late-onset sepsis will be measured
by positive blood culture or CSF culture result as suggested by the studies found
(Demirel et al., 2013; Jacobs et al., 2013; Lin et al., 2005; Manzoni et al., 2009; Oncel et
al., 2014; Samanta et al., 2009). Only infants presented with persistent high fever or
severe signs of infection will perform lumbar puncture to collect CSF culture. Incidence
rate of NEC is measured by 2 pediatric physicians using Bell’s stage classification
(Demirel et al., 2013; Lin et al., 2005; Oncel et al., 2014). For the length of hospitalization,
it will be measured by the hospital data in term of days. For the staff satisfaction level, it
will be assessed by the APNs in the committee using assessment form in Appendix M.
During the pilot study, the RNs in the committee are responsible to collect the
demographic data and the clinical outcome measures of recruited subjects and count
the resources used. The APNs in the committee are responsible to monitor staff’s
compliance and assess their competency in performing proposed guideline using
checklist in Appendix L. The staff satisfaction level will be measured in week 8 and 16
respectively.
Evaluation of pilot study
For the evaluation, it will be started in week 8 after first slot of data collected, all
data and feedback collected will be analyzed in week 17 A committee team meeting will
- 40 -
be held at the end of week 17 to discuss data collected and the difficulties encountered
during the pilot study. A refined guideline and details of implementation plan of main
study will be presented and proposed to the administrators in the department meeting
in week 18.
Evaluation Plan
The revised guideline will then be implemented in week 19-36 based on the
experience of pilot study and a detailed evaluation will be carried out afterwards. The
timelines of the main study implementation plan and evaluation plan is shown in
appendix J and K.
Outcomes measures identification
The primary outcome of the innovation is to reduce the incidence rate of late-
onset sepsis in preterm infants admitted to the department. The primary outcome is
measured by blood or CSF cultures (Demirel et al., 2013; Jacobs et al., 2013; Lin et al.,
2005; Manzoni et al., 2009; Oncel et al., 2014; Samanta et al., 2009). Only infants
presented with persistent high fever or severe signs of infection will perform lumbar
puncture to collect CSF culture. And the secondary outcomes are to reduce the
incidence rate of NEC and to shorten length of hospitalization of target subjects.
Incidence of NEC is measured by 2 pediatric physicians using Bell’s stage classification
(Demirel et al., 2013; Lin et al., 2005; Oncel et al., 2014) and length of hospitalization is
measured by hospital data in term of days.
- 41 -
Besides that, high staff compliance rate and increasing staff satisfaction level on
the new guideline are also essential outcomes we would like to achieve. The staff
compliance rate will be measured by the assessment form in appendix L and the staff
satisfaction level will be measured by the staff satisfaction survey in Appendix M
regularly by APNs in the committee during implementation. Qualitative data, such as the
opinions from all stakeholders will also be collected.
Nature of subjects
The nature of subjects will follow the suggestions from the six studies found
(Demirel et al., 2013; Jacobs et al., 2013; Lin et al., 2005; Manzoni et al., 2009; Oncel et
al., 2014; Samanta et al., 2009). Preterm infants that admitted to NICU or SCBU with
gestational age ≤ 32 weeks and/ or birth weight ≤ 1500 grams without major congenital
and gastrointestinal or multiple complications are recruited when they start enteral
feeding.
Sample size
According to the Samanta et al. (2009), the probiotic will be used in the target
setting can achieve a RRR on the incidence rate of late-onset sepsis in preterm infants by
around 93%. Thus, we would like to set our target RRR of late-onset sepsis to be 93%.
The sample size is then calculated by using the Piface sample size calculating
software available on the website http://homepage.cs.uiowa.edu/~rlenth/Power/.
According to the past statistic in the target setting, the incidence rate of definite late-
onset sepsis in extremely and very preterm infants was around 20%. Setting a 80%
- 42 -
statistical power with an absolute RRR on incidence of late-onset sepsis by 93% and a
drop-out rate of 5%, the total number of subjects need to be recruited in the main study
would be 25.
Timing and frequency of taking measurements
For the primary outcome, blood cultures are collected when the eligible subjects
present signs of infection during hospitalization. CSF culture will be collected when
persistent fever or signs of severe infection are indicated by doctors. For the secondary
outcome, incidence rate of NEC would be measured by 2 pediatric physicians using Bell’s
classification whenever signs of NEC presented during hospitalization. For the length of
hospitalization, data will be collected when target subjects discharged. And for the staff
compliance rate and satisfaction level, it will be assessed in the 6th week and then every
4 weeks during implementation in week 24, 28, 32 and 36.
Data analysis
After the implementation, data analysis will be performed in week 37-41 using
the Statistical Package for the Social Science (SPSS), version 22.0 for Window 7. The 3
RNs in the committee will be responsible for the data analysis.
The two tailed ƶ-test for testing one proportion will be used to analyze the
incidence rate of late-onset sepsis and NEC. In addition to the p-values of the outcomes,
results will be presented using relative risk (RR) also. Length of hospitalization will be
analyzed by using one sample t-test. The p-value, mean and standard derivation (SD) will
- 43 -
be given. For all the outcomes measures, a p-value of <0.05 will be considered
statistically significant.
For the staff compliance rate and satisfaction level, they will be presented using
percentage. For the feedbacks collected from all stakeholders, it will be discussed and
analyzed in the committee meetings in week 37-41 by grouping and coding similar ideas
using grounded theory. Finally, the results of the innovations and feedback from the
committee will be presented in a written report in week 41.
Basis for Implementation
The refined guideline will be reported and discussed with the administrators in
week 42 and the finalized EBP guideline will be published in week 43. The proposed
innovation will be considered to be effective and will be fully implemented in the
department if it fulfills the following criteria.
Firstly, no adverse events or safety concerns detected during the implementation.
Secondly, the RRR of incidence rate of late-onset sepsis in extremely or very preterm
infants should reach at least 50% and proven to be statistically significant. Thirdly, the
staff compliance rate is over 90%. Finally, the staff satisfaction level on this innovation
should be over 80% of the total score.
- 44 -
Conclusion
Late onset sepsis is a significant problem in extremely and very preterm infants
that causes additional unpleasant events to the infants, their parents and medical staff.
It lengthen the duration of hospitalization, caused extra invasive diagnostic and
therapeutic procedures like long line insertion, lumbar puncture and blood taking,
increased risk of serious medical conditions, such as meningitis, seizure, increased
mortality, medical expenses, parental anxiety and staff workload.
The use of oral probiotic was proven to be an effective way in preventing late
onset sepsis in preterm infants. Evidences are concluded from six high quality RCTs in
the above chapters. And an EBP guideline for healthcare professionals is developed and
is transferable and feasible to the proposed target setting. A detailed implementation
plan is reviewed to ensure smooth running of the guideline. Thus, it is believed that
more preterm infants can be benefited from the new guideline in the future.
- 45 -
Bibliography
Altman, D.G. (1991). Practical Statistics for Medical Research. Chapman & Hall: New York.
Demirel, G., Erdeve, O., Celik, I. H., & Dilmen, U. (2013). Saccharomyces boulardii for
prevention of necrotizing enterocolitis in preterm infants: a randomized,
controlled study. Acta Paediatrica, 102(12), e560-e565.
FAO/WHO. (2006). Health and nutrition properties of probiotics in food including
powder milk with life lactic acid bacteria. Rome: Food and Agriculture Organization
of the United Nations and World Health Organization. Retrieved December 5 2015,
from ftp://ftp.fao.org/docrep/fao/009/a0512e/a0512e00.pdf
Hospital Authority. (2016). Vision, Mission and Values. Hong Kong: Hospital Authority.
Retrieved February 22 2016, from
https://www.ha.org.hk/visitor/ha_visitor_text_index.asp?Parent_ID=10004&Cont
ent_ID=10009&Ver=TEXT
Jacobs, S. E., Tobin, J. M., Opie, G. F., Donath, S., Tabrizi, S. N., Pirotta, M., Morley, C. J. &
Garland, S. M. (2013). Probiotic Effects on Late-onset Sepsis in Very Preterm
Infants: A Randomized Controlled Trial. Pediatrics: 132(6): 1055-1062.
Lin, H., Su, B., Chen, A., Lin, T., Tsai, C., Yeh, T., & Oh, W. (2005). Oral probiotics reduce
the incidence and severity of necrotizing enterocolitis in very low birth weight
infants. Pediatrics, 115(1), 1-4 4p.
Locke, L. F., Spirduso, W. W., & Sliverman, S. J. (2007). Proposals that work a guide for
planning dissertations and grant proposals (5th ed.). Thousand Oaks: Sage.
Manzoni, P., Rinaldi, M., Cattani, S., Pugni, L., Romeo, M. G., Messner, H., Stolfi, H.,
Decembrino, L., Laforgia, N., Vagnarelli, F., Memo, L., Bordignon, L., Saia, O. S.,
Maule, M., Gallo, E., Mostert, M., Magnani, C., Quereia, M., Bollani, L., Pedicino, R.,
Renzullo, L., Betta, P., Mosca, F., Ferrari, F., Magaldi, R., Stronati, M., Farina, D.
(2009). Bovine Lactoferrin Supplementation for Prevention of Late-Onset Sepsis in
Very Low-Birth-Weight Neonates: A Randomized Trial. Journal of American
Medical Association.302 (13):1421-1428.
Melnyk, B.M., Fineout-Overholt, E. (2005). Evidence-based practice in nursing &
healthcare: A guide to best practice. Philadelphia: Lippincott Williams & Wilkins.
- 46 -
Millar, M., Wilks, M., & Costeloe, K. (2003). Probiotics for preterm infants?. Archives of
Disease in Childhood. Fetal and Neonatal Edition. 88(5):F354-8
Millar, M., Wilks, M., Fleming, P., & Costeloe, K. (2012). Should the use of probiotics in
the preterm be routine? Archives of Disease in Childhood. Fetal and Neonatal
Edition. 97(3): F217-218.
Mshvildadze, M., Neu, J., Shuster, j., Theriaque, D., Li, N., & Mai, V. (2010). Intestinal
microbial ecology in premature infants assessed with non-culture-based
techniques. Journal of Pediatrics. 156(1): 20-25
Munro B.KH. (2005) Statistical methods for health care research (5th ed.). Philadelphia:
Lippincott.
Oncel, M. Y., Sari, F. N., Arayici, S., Guzoglu, N., Erdeve, O., Uras, N., Oguz, S. S., & Dilmen,
U. (2014). Lactobacillus Reuteri for the prevention of necrotizing enterocolitis in
very low birthweight infants: a randomized controlled trial. Archives of Disease in
Childhood. Fetal and Neonatal Edition. 99: F110-F115.
Retnaningtyas, L. P., Etika, R., & Sudarmo, S. M. (2010). EFFECT OF PROBIOTIC
ADMINISTRATION ON THE LEVELS OF FECAL SECRETORY IMMUNOGLOBULIN A IN
PREMATURE INFANTS. Folia Medica Indonesiana, 46(1), 15-23.
Samanta, M., Sarkar, M., Ghosh, P., Ghosh, J., Sinha, M., & Chatterjee, S. (2009).
Prophylactic probiotics for prevention of necrotizing enterocolitis in very low birth
weight newborns. Journal of Tropical Pediatrics. 55(2): 128–131.
Schwiertz, A., Gruhl, B., Löbnitz, M., Michel, P., Radke, M., & Blaut, M. (2003).
Development of the intestinal bacterial composition in hospitalized preterm
infants in comparison with breast-fed, full-term infants. Pediatric Research.
54(3):393-399
Stoll, B. J., Hansen, N., Fanaroff, A. A., Wright, L. L., Carlo, W. A., Ehrenkranz, R. A.,
Lemons, J. A., Donovan, E. F., Stark, A. R., Tyson, J. E., Oh, W., Bauer, C. R., Korones,
S. B., Shankaran, S., Laptool, A. R., Stevenson, D. K., Papile, L. & Poole, W. K.
(2002). Late-onset sepsis in very low birth weight neonates: the experience of the
NICHD Neonatal Research Network. Pediatrics. 110:285-91.
Tsai, M. H., Chu, S. M., Hsu, J. F., Lien, R., Huang, H. R., Chiang, M. C., Fu, R. H., Lee, C. W.,
Huang, Y. C. (2015). Breakthrough bacteremia in the neonatal intensive care unit:
incidence, risk factors, and attributable mortality. American Journal of Infection
Control. 43:20-5.
- 47 -
United Nations. (2016). World Economic Situation and Prospects. New York: United
Nations. Retrieved January 30 2016, from
http://www.un.org/en/development/desa/policy/wesp/wesp_current/2016wesp_
full_en.pdf
World Health Organization. (2012). Born too soon: the global action report on preterm
birth. Geneva: World Health Organization. Retrieved December 5 2015, from
http://apps.who.int/iris/bitstream/10665/44864/1/9789241503433_eng.pdf?ua=1
- 48 -
Appendix
Appendix A: Search strategies Sc
reen
ing
Incl
ud
ed
Elig
ibili
ty
Records after duplicates removed (n = 10)
Full-text articles assessed for eligibility
(n = 10)
Full-text articles excluded, with reasons
(n = 4)
Studies included in qualitative synthesis
(n = 6)
Records identified through database searching
(n = 54 )
Iden
tifi
cati
on
Additional records identified through other sources
(n =2 )
Records screened (n = 56)
Records excluded (n = 43)
- 49 -
Appendix A: Search strategies (cont’d)
Keyword PubMed Cochrane library
British Nursing Index
S1: premature infant 74927 5481 596
S2: premature infants 77454 5481 595
S3: preterm infant 82936 5110 906
S4: preterm infants 82461 5110 906
S5: neonatal prematurity 77644 825 120
S6: prematurity 84466 2369 186
S7: S1 OR S2 OR S3 OR S4 OR S5 OR S6 96125 8789 1446
S8: lactoferrin 7216 296 3
S9: bovine lactoferrin 1665 61 1
S10: probiotic 15556 1820 138
S11: probiotics 13383 1930 138
S12: S8 OR S9 OR S10 OR S11 22713 2856 141
S13: neonatal sepsis 8513 636 86
S14: neonatal infection 21314 1009 425
S15: neonatal infections 26072 1009 425
S16: late onset sepsis 1278 119 18
S17: late onset infection 3430 217 27
S18: late onset infections 4042 217 28
S19: bacteremia 36616 1406 79
S20: S13 OR S14 OR S15 OR S16 OR S17 OR S18 OR S19
66500 5132 532
S7 AND S12 AND S20 124 67 9
Articles remained after limited to RCTs and within 10 years (2005-2015), full text
16 37 1
Articles remained after screening title and abstract
5 6 0
Articles remained after removal of duplicated with other databases
8
Articles remained after screening full paper
4
Manual search from the references of potential eligible studies
2
Total 6
- 50 -
Appendix B: Table of Evidence Lin, H., Su, B., Chen, A., Lin, T., Tsai, C., Yeh, T., & Oh, W. (2005). Oral probiotics reduce the incidence and severity of necrotizing enterocolitis in very low birth weight infants. Pediatrics, 115(1), 1-4 4p.
Study type Sample characteristics Intervention Control Outcome measure(s) Result: (p=p-value)
Lin et al.(2005)/ RCT (1+)
neonates
Birth weight <1500g
Started to fed enterally
mean birth weight (g):
(1104/ 1071 : probiotic
/control)
mean gestational age
(wks): (28.5/ 28.2 :
probiotic /control)
Exclusion criteria:
Died or had NEC before
7th day of life
Probiotic: (n=180)
Placebo: (n=187) Primary measure(s)
1. number of infants with sepsis
(culture proven):
a. probiotic: 22 (12.2%)
b. control: 36 (19.3%)
2. number of infants with NEC≥
stage 2:
a. probiotic: 2 (1.1%)
b. control: 10 (5.5%)
3. mortality:
a. probiotic: 7 (3.9%)
b. control: 20 (10.7%)
1. number of infants with
sepsis (culture proven):
a. probiotic Vs control:
p=0.03, RRR=36.2%
2. number of infants with
NEC≥ stage 2:
a. probiotic Vs control:
p=0.04, RRR=81.3%
3. mortality:
a. probiotic Vs control:
p=0.009, RRR=63.6%
1. Probiotic group: combination
Lactobacillus acidpphilus and
Bifidobacterium infantis 125mg/kg
per dose mixed with breast milk
2. Control group: breast milk or
donor milk only
3. Probiotics mixed with breast milk
or donor milk before feeding
4. Feeding was started when the
infant had stable vital signs, active
bowel sound without abdominal
distension, no bile or blood from
the nasogastric tube, and no
indwelling umbilical artery or
umbilical venous catheter for at
least 24 hours.
Duration:
twice per day until discharge
- 51 -
Samanta, M., Sarkar, M., Ghosh, P., Ghosh, J., Sinha, M., & Chatterjee, S. et al.(2009). Prophylactic probiotics for prevention of necrotizing enterocolitis in very low birth weight newborns. Journal of Tropical Pediatrics. 55(2): 128–131.
Study type Sample characteristics Intervention Control Outcome measure(s) Result: (p=p-value)
Samanta et al.(2008)/ RCT (1+)
Inclusion criteria:
Born < 32 completed
weeks gestation
Birth weight <1500g
Started feed enterally
mean birth weight (g):
(1172/ 1210 : probiotic
/control)
mean gestational age
(wks): (30.12/ 30.14 :
probiotic /control)
Exclusion criteria:
Died before 2nd day of
life
had major congenital or
gastrointestinal
anomalies
infants expired due to
other neonatal illnesses
Probiotic: (n=91)
Placebo: (n=95) Primary measure(s)
1. days reached full enteral
feeding (SD):
a. probiotic: 13.76 2.28
b. control: 19.2 2.02
2. length of hospital stay (days)
(SD):
a. probiotic: 17.17 3.23
b. control: 24.07 4
3. mortality:
a. probiotic: 4 (4.4%)
b. control: 14 (14.7%)
4. number of infants with NEC≥
stage 2:
a. probiotic: 5 (1.1%)
b. control: 15 (15.8%)
5. number of infants with sepsis
(culture proven):
a. probiotic: 13 (14.3%)
b. control: 28 (29.5%)
1. days reached full enteral
feeding:
a. probiotic Vs control:
effect size= -2.39;
p=<0.001
2. days of hospital stay:
a. probiotic Vs control:
effect size= -2.14;
p=<0.001
3. mortality:
a. probiotic Vs control:
p=0.032, RRR=70.1%
4. number of infants with
NEC≥ stage 2:
a. probiotic Vs control:
p=0.042, RRR=93.0%
5. number of infants with
sepsis (culture proven):
a. probiotic Vs control:
p=0.020, RRR=51.5%
1. Probiotic group: mixture of
Bifidobacteria infantis,
Bifidobacteria bifidum,
Bifidobacteria longum and
Lactobacillus acidophilus, each2.5 x
109 CFU in breast milk, 125g/kg
2. Control group: breast milk only
3. Feeding was started when the
infant had stable vital signs, active
bowel sound without abdominal
distension, no bile or blood from
the nasogastric tube.
Duration:
4. Twice daily until discharge
CFU: colony-forming units
- 52 -
Manzoni, P., Rinaldi, M., Cattani, S., Pugni, L., Romeo, M. G., & Messner, H., et al. (2009). Bovine Lactoferrin Supplementation for Prevention of Late-Onset Sepsis in Very Low-Birth-Weight Neonates: A Randomized Trial. Journal of American Medical Association.302(13):1421-1428.
Study type Sample characteristics Intervention Control Outcome measure(s) Result: (p=p-value)
Manzoni et al.(2009) / prospective RCT (1++)
Inclusion criteria:
1500g
younger than 3 days
mean birth weight (g):
(1142/1138/1109 :
BLF/BLF+LGG/control)
mean gestational age
(wks): (29.6/29.8/29.5:
BLF/BLF+LGG/control)
Exclusion criteria:
Parental consent lacking/
refused
Ongoing antifungal
prophylaxis
Early onset sepsis (before
day 3 of life)
Liver failure (APT, ALP -
glutamyl transferase and
DB 3-fold higher than
reference range)
BLF alone
(n=153)*
BLF+LGG (n=151)
Placebo: (n=168) Primary measure(s)
1. Incidence of late-onset sepsis
(culture proven):
a. BLF: 9 (5.9%)*
b. BLF+LGG: 7 (4.6%)
c. Control: 29 (17.3%)
2. Mortality attributable to sepsis
(culture proven):
a. BLF: 0 (0%)*
b. BLF+LGG: 1 (0.7%)
c. Control: 8 (4.8%)
Secondary measure(s)
3. NEC≥ stage 2
a. BLF: 3 (1.9%)*
b. BLF+LGG: 0 (0%)
c. Control: 10 (6.0%)
1. Incidence of late-onset
sepsis (culture proven):
a. BLF Vs control:
p=0.002, RRR=65.9%*
b. BLF+LGG Vs control:
p<0.001,RRR=73.4%
2. Mortality attributable to
sepsis (culture proven):
a. BLF Vs control:
p=0.008, RRR=100%*
b. BLF+LGG Vs control:
p=0.04, RRR=85.4%
3. NEC≥ stage 2
a. BLF Vs control: p=
0.09 (not significant),
RRR=68.3% *
b. BLF+LGG Vs control:
p=0.002,RRR=100%
1. BLF group: BLF (100mg/d)* or
2. BLF + LCC group: BLF (100mg/d) +
LGG (6x 109 CFU/d)
3. Control group: 2ml of 5% glucose
solution
4. Begin in 3rd day of life
5. 1 daily dose
6. Diluted in milk
7. Neonates not feeding in the first 48
hours received the drug(s) or
placebo by oro-gastric tube
Duration:
Start form 3rd of life
6 weeks (birth weight <1000g)
4weeks (birth weight 1001-1500g)
VLBW: very low birth weight; APT: aspartate aminotransferase; ALP: alanine aminotransferase; DB: direct bilirubin serum; BLF: bovine lactoferrin; LGG: Lactobacillus rhamnosus GG; late-onset sepsis: occurs more than 3rd of life and before discharge; * The BLF only group is not our focus in this dissertation
- 53 -
Demirel, G., Erdeve, O., Celik, I. H., & Dilmen, U. (2013). Saccharomyces boulardii for prevention of necrotizing enterocolitis in preterm infants: randomized, controlled study. Acta Paediatrica, 102(12), e560-e565.
Study type Sample characteristics Intervention Control Outcome measure(s) Result: (p=p-value)
Demirel et al.(2013)/ RCT (1+)
neonates
Birth weight 1500g
Born 32 completed
weeks gestation
Started to fed enterally
mean birth weight (g):
(1164/ 1131 : probiotic
/control)
mean gestational age
(wks): (29.4/ 29.2 :
probiotic /control)
Exclusion criteria:
With major congenital
malformations
Died within 7th day
after intervention started
(exclude from analysis)
Probiotic: (n=135)
Placebo: (n=136) Primary measure(s)
1. number of infants with NEC≥
stage 2:
a. probiotic: 6 (4.4%)
b. control: 7 (5.1%)
2. mortality:
a. probiotic: 5 (3.7%)
b. control: 5 (3.7%)
Secondary measure(s)
3. number of infants with sepsis
(clinical):
a. probiotic: 47 (34.8%)
b. control: 65 (47.8%)
4. number of infants with sepsis
(culture proven):
a. probiotic: 20 (14.8%)
b. control: 21 (15.4%)
Primary measure(s)
1. number of infants with
NEC≥ stage 2:
a. probiotic Vs control:
p=1.0 (not
significant),,
RRR=14.3%
2. mortality:
a. probiotic Vs control:
p=1.0 (not
significant),,
RRI=0.770%
Secondary measure(s)
3. number of infants with
sepsis (clinical):
a. probiotic Vs control:
p=0.03, RRR=41.7%
4. number of infants with
sepsis (culture proven):
a. probiotic Vs control:
p=0.906 (not
significant),,
RRR=4.76%
1. Probiotic group: Saccharomyces
boulardii 125mg/kg per dose
mixed with breast milk or formula
milk
2. Control group: breast milk or
formula milk
3. Probiotics mixed with breast milk
or donor milk before feeding
4. Feeding was started when the
infant had stable vital signs, active
bowel sound without abdominal
distension, no bile or blood from
the nasogastric tube.
Duration:
once per day until discharge
- 54 -
Jacobs, S. E., Tobin, J. M., Opie, G. F., Donath, S., Tabrizi, S. N., & Pirotta, M., et al. (2013). Probiotic Effects on Late-onset Sepsis in Very Preterm Infants: A Randomized Controlled Trial. Pediatrics: 132(6): 1055-1062.
Study type Sample characteristics Intervention Control Outcome measure(s) Result: (p=p-value)
Jacobs et al. (2013)/ RCT (1+)
Inclusion criteria:
Born < 32 completed
weeks gestation
Birth weight <1500g
Younger than 3 day
of life
mean birth weight
(g): (1063/ 1048 :
probiotic /control)
mean gestational age
(wks): (27.9/ 27.8 :
probiotic /control)
Exclusion criteria:
had major congenital
or chromosomal
anomalies
death was considered
likely within 72 hours
of birth
mother was taking
non-dietary probiotic
supplements
Probiotic:(n=548) Placebo: (n=551) Primary measure(s)
1. number of infants (≥ 28weeks) with
at least 1 episode of late onset
sepsis (culture proven):
a. probiotic: 18 (5.5%)
b. control: 34 (10.8%)
2. number of infants (<28 weeks) with
at least 1 episode of late onset
sepsis (culture proven):
a. probiotic: 54 (24.7%)
b. control: 55 (23.4%)
Secondary measure(s)
3. number of infants with NEC≥ stage 2:
a. probiotic: 11 (2.0%)
b. control: 24 (4.4%)
4. mortality caused by NEC:
a. probiotic: 4 (0.7%)
b. control: 11 (2.0%)
5. Days required full enteral feeding:
a. probiotic: 71
b. control: 74
6. length of hospital stay (days):
a. probiotic: 12
b. control: 12
7. number of infants with clinical sepsis:
a. probiotic: 75 (13.7%)
b. control: 83 (15.1%)
Primary measure(s)
1. Number of infants (≥ 28weeks) with at
least 1 episode of late onset sepsis
(culture proven):
a. probiotic Vs control: p=0.01, RRR=
49.1%
2. Number of infants (<28 weeks) with at
least 1 episode of late onset sepsis
(culture proven):
a. probiotic Vs control: p=0.75(not
significant), RRI= 5.5%
Secondary measure(s)
3. number of infants with NEC≥ stage 2:
a. probiotic Vs control: p=0.03, RRR=
54.5%
4. mortality caused by NEC:
a. probiotic Vs control: p=0.07(not
significant), RRR= 65.0%
5. Days required full enteral feeding:
a. probiotic Vs control: p=0.09 (not
significant)
6. length of hospital stay:
a. probiotic Vs control: p=0.31 (not
significant)
7. number of infants with clinical sepsis
a. probiotic Vs control: p=0.52 (not
significant), RRR=9.3%
1. Probiotic group: combination
Bifidobacterium infantis (BB-02
300 x 106), Streptococcus
thermophiles (TH-4 350 x 106)
and Bifidobacterium lactis (BB-
12 350 x 106) with 1 x 109 CFU
per 1.5g, in a maltodextrin base
powder
2. Control : maltodextrin powder
3. Administered only when an
infant was receiving at least 1
ml of milk every 4 hours
4. Withheld during periods when
infants were nil orally
5. Daily dose of 1.5g of study
powder, reconstituted with 3 ml
breast milk or formula. OR twice
daily of 0.75 study powder,
reconstituted with 1.5ml breast
milk or formula
6. Administered by gastric tube or
mouth
Duration:
Until discharge or term corrected
age
Late onset sepsis: culture proven (blood/ CSF) after 2 days of life treated with antibiotics ≥ 5days
- 55 -
Oncel, M. Y., Sari, F. N., Arayici, S., Guzoglu, N., Erdeve, O., & Uras, N., et al.(2014). Lactobacillus Reuteri for the prevention of necrotizing enterocolitis in very low birthweight infants: a randomized controlled trial. Arch Dis Child Fetal Neonatal Ed. 99: F110-F115.
Study type Sample characteristics Intervention Control Outcome measure(s) Result: (p=p-value)
Oncel et al.(2014)/ RCT (1+)
Inclusion criteria:
Born 32 completed
weeks gestation
Birth weight 1500g
Survived to fed enterally
mean birth weight (g):
(1071/ 1048 : probiotic
/control)
mean gestational age
(wks): (28.2/ 27.9 :
probiotic /control)
Exclusion criteria:
had major congenital
malformations
lacks of parental consent
infants died within the
1st week of life (would
not have had
opportunity to benefit
from the intervention)
Probiotic: (n=213)
(n=200, analyzed)
Placebo: (n=211)
(n=200, analyzed)
Primary measure(s)
1. number of infants with
NEC≥ stage 2:
a. probiotic: 8 (4.0%)
b. control: 10 (5.0%)
Secondary measure(s)
2. number of infants with
culture proven sepsis:
a. probiotic: 13 (6.5%)
b. control: 25 (12.5%)
3. days required full enteral
feeding (SD):
a. probiotic: 9.1 3.2
b. control: 10.1 4.3
4. length of hospital stay
(days):
a. probiotic: 38
b. control: 46
1. number of infants with
NEC≥ stage 2:
a. probiotic Vs control:
p=0.63(not
significant),
RRR=20.0%
Secondary measure(s)
2. number of infants with
culture proven sepsis:
a. probiotic Vs control:
p=0.041, RRR=48.0%
3. days required full enteral
feeding
a. probiotic Vs control:
(effect size = -0.3125;
p= 0.006)
4. length of hospital stay
a. probiotic Vs control:
p= 0.022
1. Probiotic group: oil-based suspension
containing 1 x 108 CFU/day (5 drops)
of lyophilized Lactobacillus reuteri
2. Control group: oil-based suspension
3. Feeding was started when the infant
had stable vital signs, active bowel
sound without abdominal distension,
no bile or blood from the nasogastric
tube.
4. Started with the first feed of the
infants
5. Suction of oral secretions
6. 5 drops of study powder in oil-based
will be placed in the posterior
oropharynx of the infants
7. For infants without per oral feeds, 5
drops of study substance will be
administered through a gastric tube
followed by a flash of 0.5 ml of sterile
water
8. Follow by breast milk or formula milk
Duration:
Once daily until discharge
- 56 -
Appendix C: Summary or studies results Lin (2005) Samanta (2009) Manzoni (2009) Demirel (2013) Jacobs (2013) Oncel (2014)
Inclusion criteria
1. <1500g 2. Started enteral
feeding
1. <1500g 2. <32 weeks 3. Started enteral
feeding
1. 1500g 2. Younger than 3 day of
life
1. 32 weeks
2. 1500g 3. Started enteral
feeding
1. <32 weeks 2. <1500g 3. Younger than 3 day of
life
1. 32 weeks
2. 1500g 3. Started enteral
feeding
Exclusion criteria
1. Died or had NEC before 7th day of life
1. Died before 2nd day of life
2. With major congenital and GI anomalies
1. Ongoing antifungal prophylaxis
2. Had sepsis before 3rd of life
3. Liver failure
1. With major congenital malformations
2. Died within 7th day after intervention started (exclude from analysis)
1. With major congenital or chromosomal anomalies
2. Died before 3rd of life 3. Mother taking
nondietary probiotic supplements
1. With major congenital malformations
2. Died before 7th day of life (exclude from analysis)
Places Taiwan (Developing)
India (Developing)
11 sites in Italy (Developed)
Turkey (Developing)
8 sites in Australia 2 in New Zealand (Developed)
Turkey (Developing)
Study period
54 months 6 months 10 months 9 months 37 months 12 months
Study agent Lactobacillus acidophilus and Bifidobacterium infantis
Bifidobacteria infantis, Bifidobacteria bifidum, Bifidobacteria longum and Lactobacillus acidophilus, each 2.5 x 10
9 CFU
Bovine lactoferrin (BLF) + Lactobacillus rhamnosus GG (LGG)
Saccharomyces boulardii Bifidobacterium infantis, Streptococcus thermophiles and Bifidobacterium lactis, 1 x 10
9 CFU, in a
maltodextrin base powder
Lactobacillus reuteri , 1 x 10
8 CFU (oil-based)
Control agent
Breast milk Breast milk 2ml of 5% glucose solution
Breast milk of formula milk
Maltodextrin base powder
Oil-based suspension
Dose Frequency
125mg/ kg twice daily 125g/ kg twice daily Daily Daily 1.5 g study powder daily or 0.75 g twice daily
Daily
Total daily dose
250mg/kg 250g/kg BLF (100mg/day) + LGG (6x 10
9 CFU/day)
250mg ( 5x 109 CFU/day) Study powder 6x 10
9
CFU/day Lactobacillus reuteri , 1 x 10
8 CFU/day
Duration (start)
Baby start enteral feeding without UVA/C at least 24 hours
Baby start enteral feeding
3rd
day of life Baby start enteral feeding
Baby start enteral feeding receiving at least 1 ml of milk every 4 hours
Baby start enteral feeding
Stop when Sign of feeding intolerance
NA NA Two or more signs of feeding intolerance
Nil per oral Two or more signs of feeding intolerance
- 57 -
Appendix C: Summary or studies results (cont’d) Lin (2005) Samanta (2009) Manzoni (2009) Demirel (2013) Jacobs (2013) Oncel (2014)
Duration (end)
Until discharge Until discharge 6 weeks (birth weight <1000g); 4 weeks (birth weight 1001-1500g)
Until discharge Until discharge or term corrected age
Until death or discharge
Study agent mixed with
Breast milk or donor milk Breast milk Breast milk or formula Breast milk of formula Breast milk or formula Oil-based suspension
Method of feeding
Oral Oral Oral or oro-gastric tube Oral Oral or oro-gastric tube Oral or oro-gastric tube
Outcome measure
1. Sepsis 2. NEC≥ stage 2 3. Mortality
1. Sepsis 2. NEC≥ stage 2 3. Mortality 4. Days required full
enteral feeding 5. Length of hospital
stay
1. Sepsis 2. NEC≥ stage 2 3. Mortality attributable
to sepsis
1. Sepsis (clinical) 2. Sepsis (culture
proven) 3. NEC≥ stage 2 4. Mortality
1. Sepsis (≥ 28weeks) 2. Sepsis (<28 weeks) 3. Clinical sepsis 4. NEC≥ stage 2 5. Mortality ( NEC) 6. Days required full
enteral feeding 7. Length of hospital stay
1. Sepsis 2. NEC≥ stage 2 3. Days required full
enteral feeding 4. Length of hospital stay
Measuring tool
1. Culture proven (blood) after randomization
2. Bell’s classification: 2 independent neonatologists
3. Hospital data
1. Culture proven (blood/ CSF) beyond 5
th of life
2. Bell’s classification 3. Hospital data 4. Hospital data 5. Hospital data
1. 2 Culture proven (blood/ CSF) beyond 3rd of life
2. Bell’s classification 3. Hospital data
1. C-reactive protein (CRP) >4.82 mg/L and Interleukin-6 >4.82mg/L
2. 1 culture proven (blood/ CSF/ urine)
3. Bell’s classification 4. Hospital data
1. 2 Culture proven (blood/ CSF) beyond 2nd of life and antibiotic ≥ 5days
2. Same as above 3. CRP >10mg/L and /or
immature-to-total neutrophil ratio>0.2 and antibiotic ≥ 5days
4. Bell’s classification 5. Hospital data 6. Hospital data 7. Hospital data
1. 2 Culture proven (blood)
2. Bell’s classification: 2 independent neonatologists
3. Hospital data 4. Hospital data
Result (p-value, relative risk reduction)
1. p=0.03,RRR=36.2% 2. p=0.04,RRR= 81.3% 3. p=0.009,RRR= 63.6%
1. p=0.020,RRR=93% 2. p=0.042,RRR=51.5% 3. p=0.032,RRR=70.1% 4. p=<0.001 5. p=<0.001
1. p=< 0.001, RRR=73.4%
2. p=0.04, RRR=85.4% 3. p=0.002, RRR=100%
1. p=0.03,RRR=41.7% 2. p=0.906,RRR=4.76% 3. p=1.0,RRR=14.3% 4. p=1.0,RRI=0.770%
1. p=0.01,RRR=49.1% 2. p=0.75,RRI=5.5% 3. p=0.52, RRR=9.3% 4. p=0.03,RRR=54.5% 5. p=0.07,RRR=65.0% 6. p=0.09 7. p=0.31
1. p=0.041, RRR=48% 2. p=0.63, RRR=20% 3. p=0.006 4. p=0.022
- 58 -
Appendix D: Methodology Checklist 2: Controlled Trials
S I G N
Methodology Checklist 2: Controlled Trials
Study identification (Include author, title, year of publication, journal title, pages)
Guideline topic: Key Question No: Reviewer:
Before completing this checklist, consider:
1. Is the paper a randomised controlled trial or a controlled clinical trial? If in doubt, check the study design algorithm available from SIGN and make sure you have the correct checklist. If it is a controlled clinical trial questions 1.2, 1.3, and 1.4 are not relevant, and the study cannot be rated higher than 1+
2. Is the paper relevant to key question? Analyse using PICO (Patient or Population Intervention Comparison Outcome). IF NO REJECT (give reason below). IF YES complete the checklist.
Reason for rejection: 1. Paper not relevant to key question 2. Other reason (please specify):
SECTION 1: INTERNAL VALIDITY
In a well conducted RCT study… Does this study do it?
1.1 The study addresses an appropriate and clearly focused
question.
Yes
Can’t say
No
1.2 The assignment of subjects to treatment groups is randomised.
Yes
Can’t say
No
1.3 An adequate concealment method is used.
Yes
Can’t say
No
1.4 The design keeps subjects and investigators ‘blind’ about
treatment allocation.
Yes
Can’t say
No
1.5 The treatment and control groups are similar at the start
of the trial.
Yes
Can’t say □
No
1.6 The only difference between groups is the treatment
under investigation.
Yes
Can’t say
No
- 59 -
1.7 All relevant outcomes are measured in a standard, valid
and reliable way.
Yes
Can’t say
No
1.8 What percentage of the individuals or clusters recruited
into each treatment arm of the study dropped out before
the study was completed?
1.9 All the subjects are analysed in the groups to which they were randomly allocated (often referred to as intention to treat analysis).
Yes
Can’t say
No
Does not
apply
1.10 Where the study is carried out at more than one site, results are comparable for all sites.
Yes
Can’t say
No
Does not
apply
SECTION 2: OVERALL ASSESSMENT OF THE STUDY
2.1 How well was the study done to minimise bias? Code as follows:
High quality (++)
Acceptable (+)
Low quality (-)
Unacceptable – reject 0
2.2 Taking into account clinical considerations, your evaluation of the methodology used, and the statistical power of the study, are you certain that the overall effect is due to the study intervention?
2.3 Are the results of this study directly applicable to
the patient group targeted by this guideline?
2.4 Notes. Summarise the authors’ conclusions. Add any comments on your own
assessment of the study, and the extent to which it answers your question and mention
any areas of uncertainty raised above.
- 60 -
Appendix E: SIGN Methodological Quality Coding: levels of Evidence & Grading of Recommendations
LEVELS OF EVIDENCE
1++ High quality meta-analyses, systematic reviews of RCTs, or RCTs with a very low risk of bias
1+ Well-conducted meta-analyses, systematic reviews, or RCTs with a low risk of bias
1- Meta-analyses, systematic reviews, or RCTs with a high risk of bias
2++ High quality systematic reviews of case control or cohort or studies High quality case control or cohort studies with a very low risk of confounding or bias and a high probability that the relationship is causal
2+ Well-conducted case control or cohort studies with a low risk of confounding or bias and a moderate probability that the relationship is causal
2- Case control or cohort studies with a high risk of confounding or bias and a significant risk that the relationship is not causal
3 Non-analytic studies, e.g. case reports, case series
4 Expert opinion
GRADES OF RECOMMENDATIONS
At least one meta-analysis, systematic review, or RCT rated as 1++, and directly applicable to the target population; or A body of evidence consisting principally of studies rated as 1+, directly applicable to the target population, and demonstrating overall consistency of results
A body of evidence including studies rated as 2++, directly applicable to the target population, and demonstrating overall consistency of results; or Extrapolated evidence from studies rated as 1++ or 1+
A body of evidence including studies rated as 2+, directly applicable to the target population and demonstrating overall consistency of results; or Extrapolated evidence from studies rated as 2++
Evidence level 3 or 4; or Extrapolated evidence from studies rated as 2+
Good practice points
Recommended best practice based on the clinical experience of the guideline development group
- 61 -
Appendix F: Summary of quality of studies
Lin (2005) Samanta
(2009) Manzoni (2009)
Demirel (2013)
Jacobs (2013)
Oncel (2014)
1.1 The study addresses an appropriate and clearly focused question.
Yes Yes Yes Yes Yes Yes
1.2 The assignment of subjects to treatment groups is randomised.
Yes Yes Yes Yes Yes Yes
1.3 An adequate concealment method is used.
Yes Not
mentioned Yes Yes
Not mentioned
Yes
1.4 The design keeps subjects and investigators ‘blind’ about treatment allocation.
No Not
mentioned Yes Yes Yes
Not mentioned
1.5 The treatment and control groups are similar at the start of the trial.
Yes Yes Yes Yes Not
mentioned Yes
1.6 The only difference between groups is the treatment under investigation.
Yes Yes Yes Yes Yes Yes
1.7 All relevant outcomes are measured in a standard, valid and reliable way.
Yes Yes Yes Yes Yes Yes
1.8 What percentage of the individuals or clusters recruited into each treatment arm of the study dropped out before the study was completed?
0% 0% 2.6% 2.86% 2.72% 6.1%
1.9 All the subjects are analyzed in the groups to which they were randomly allocated (often referred to as intention to treat analysis).
NA NA Yes No Yes No
1.10 Where the study is carried out at more than one site, results are comparable for all sites.
NA NA Yes NA Yes NA
2.1 How well was the study done to minimise bias?
1+ 1+ 1++ 1+ 1+ 1+
High quality (++); Acceptable (+); Low quality (-); Unacceptable – reject 0
- 62 -
Appendix G: Summary of quality of studies (Supplementary information) Section 1: Internal validity 1.2 The assignment of subjects to treatment groups is randomised.
Lin (2005) By a random-number table sequence
Samanta (2009) By a random-number table sequence
Manzoni (2009) By computer-generated randomization lists; Lactoferrin : lactoferrin and probiotic: control is 1:1:1 ratio using computer-generated randomization lists; Randomization was stratified by centers
Demirel (2013) By computer-generated sequential numbers
Jacobs (2013) By computer-generated randomization lists; Randomization was stratified by center with 1:1 ratio, Infants from multiple births were randomized individually
Oncel (2014) By computer-generated randomization lists
1.3 An adequate concealment method is used.
Lin (2005) The allocations were contained in opaque, sequentially numbered, sealed envelopes
Samanta (2009) The concealment method was not mentioned
Manzoni (2009) All doses including placebo were diluted in prepared milk to maintain blinding
Demirel (2013) The allocations were contained in opaque, sequentially numbered, sealed envelopes
Jacobs (2013) The concealment method was not mentioned
Oncel (2014) The allocations were contained in opaque, sequentially numbered sealed envelopes
1.4 The design keeps subjects and investigators ‘blind’ about treatment allocation.
Lin (2005) Single blinded; investigators were not blinded, only clinical staff remained unware of the randomization
Samanta (2009) Double blinded; Author didn’t mentioned which parties were blinded
Manzoni (2009) Double blinded; Clinical and research staff remained unware of the randomization
Demirel (2013) Double blinded; Clinical and research staff remained unware of the randomization
Jacobs (2013) Double blinded ; Apart from pharmacist, all staff and parents were blinded to the randomized allocation
Oncel (2014) Double blinded; Clinical staff were blinded but the author didn’t mention about whether investigators were blinded as well
1.5 The treatment and control groups are similar at the start of the trial.
Lin (2005) Author stated that the maternal clinical and infant’s clinical and demographic characteristics did not differ
- 63 -
between the 2 groups
Samanta (2009) Author stated that the maternal clinical and infant’s demographic and clinical characteristics were similar with p-values >0.05
Manzoni (2009) The demographic and nutritional characteristics were compared and p-values shown to be non-significant
Demirel (2013) Author stated that the maternal clinical and infant’s clinical and demographic characteristics did not differ between the 2 groups
Jacobs (2013) Author didn’t state that the demographic data of patients were similar and p values were not shown
Oncel (2014) Author stated that the maternal clinical and infant’s demographic and clinical characteristics were similar with p-values >0.05 except the duration of TPN with p=0.048
1.7 All relevant outcomes are measured in a standard, valid and reliable way.
Lin (2005) Sepsis: positive blood culture NEC ≥ Bell’s stage 2: 2 independent doctor Mortality: hospital data
Samanta (2009) Sepsis: positive blood or CSF culture taken beyond 5 days of age NEC: ≥ Bell’s stage 2 Mortality: hospital data Days required full enteral feeding: hospital data Length of hospital stay: hospital data
Manzoni (2009) Sepsis: occurred ≥ 72 hours of life. Detection of clinical signs by physician, presence of positive blood or CSF culture results NEC: ≥ Bell’s stage 2 Mortality attributable to sepsis: hospital data
Demirel (2013) Sepsis (clinical): C-reactive protein >4.82 mg/L and Interleukin-6 >4.82mg/L Sepsis (definite): 1 culture proven (blood/ CSF/ urine) NEC: ≥ Bell’s stage 2 Mortality: hospital data
Jacobs (2013) Definite sepsis: 2 separate cultures (blood/ CSF) of same species and treated with antibiotic ≥ 5 days Sepsis (clinical): CRP >10mg/L and /or immature-to-total neutrophil ratio>0.2 and treated with antibiotic ≥ 5days NEC: ≥ Bell’s stage 2
- 64 -
Mortality caused by NEC” hospital data Days required full enteral feeding: hospital data Length of hospital stay: hospital data
Oncel (2014) Sepsis: culture proven NEC: ≥ stage 2
1.8 What percentage of the individuals or clusters recruited into each treatment arm of the study dropped out before the study was completed?
Lin (2005) Dropout rate: 0% (intervention group); 0% (control group)
Samanta (2009) Dropout rate: 0% (intervention group); 0% (control group)
Manzoni (2009) Dropout rate: 1.31% (lactoferrin group); 2.6% (lactoferrin and probiotic group); 1.79% (control group)
Demirel (2013) Dropout rate: 2.17% (intervention group); 2.86% (control group)
Jacobs (2013) Dropout rate: 2.37% (intervention group); 2.72% (control group)
Oncel (2014) Dropout rate: 6.1% (intervention group); 5.5% (control group)
1.9 All the subjects are analyzed in the groups to which they were randomly allocated (often referred to as intention to treat analysis).
Lin (2005) Author didn’t mention that intention to treat analysis was used
Samanta (2009) Author didn’t mention that intention to treat analysis was used
Manzoni (2009) Author used intention to treat analysis
Demirel (2013) Participants who discontinued intervention were not included into the analysis
Jacobs (2013) Author used intention to treat analysis
Oncel (2014) Participants who discontinued intervention were not included into the analysis
1.10 Where the study is carried out at more than one site, results are comparable for all sites.
Lin (2005) Carried at in the NICU of China Medical University Hospital in Taiwan
Samanta (2009) Carried out in the NCU of Medical College and Hospital, Kolkata, India
Manzoni (2009) Carried out at 11 tertiary Italian neonatal intensive care units
Demirel (2013) Carried in the NICU of Samsun Maternity and Child Health Hospital, Samsun, Turkey
Jacobs (2013) Carried out in 2 perinatal hospitals in Australia and 2 perinatal hospitals in New Zealand
Oncel (2014) Carried in the NICU of Zekai tahir Burak Maternity Teaching Hospital, Ankara, Turkey
- 65 -
SECTION 2: OVERALL ASSESSMENT OF THE STUDY 2.4 Comments:
Lin (2005) Oral probiotics (Lactobacillus acidophilus and Bifidobacterium infantis) significantly reduces the incidence and severity of NEC, incidence of sepsis and mortality in very low birth weight infants. Potential bias The intervention was not masked to investigators and breast milk team
Samanta (2009) Oral probiotics (Bifidobacteria infantis, Bifidobacteria bifidum, Bifidobacteria longum and Lactobacillus acidophilus, each 2.5 x 109 CFU) significantly reduces the incidence of NEC, sepsis, mortality, days required full enteral feeding, length of hospital stay in very low birth weight infants, BUT NOT the severity of NEC.
Manzoni (2009) Oral lactoferrin (BLF) in combination with probiotic (LGG) decreased the incidence of late-onset sepsis NEC, and mortality attributable to sepsis in VLBW and ELBW infants.
Demirel (2013) It is statically significant in reducing the rate of clinical sepsis. All other outcomes, including definite sepsis, rate of NEC stage II or above can’t react statically significant.
Jacobs (2013) It is statically significant in reducing late onset sepsis in infants≥ 28weeks but no those <28 weeks. It was also significant in reducing NEC≥ 2nd Bell’s stage. But the decrease in mortality, days required full enteral feeding and length of hospital stay were not supported.
Oncel (2014) Probiotic (Lactobacillus reuteri , 1 x 108 CFU/day) is not statistically significant in reducing culture proven late onset sepsis but is statistically significant in reducing NEC, days required full enteral feeding and length of hospital stay
However, in this study, infants who died before 7th day of life were excluded from analysis which downgraded its quality of study. The reason explained by the author is infants would not have had opportunity to benefit from the intervention if they died within 7 days of life.
- 66 -
Appendix H: Estimated cost for the innovation annually
Estimated cost for the first year:
Items Price per unit (HKD) Quantity Amount (HKD)
Set-up cost:
Rooms for training 0 1 0
Chairs for training 0 15 0
Computer 0 1 0
Projector 0 1 0
Stationary 5 10 50
Hard copy guideline 5 6 30
Electronic copy guideline 0 1 0
Poster for promotion 20 5 100
Checklist for staff competency
1 60 60
Evaluation forms for staff satisfaction
1 60 60
Sub-total: 300
Material cost:
Probiotic 245/90 (one capsule)
140x60x2 45734
Salary of staff for training (30 minutes)
$300 0.5x55 8250
Salary of staff for milk preparation (15 minutes each day)
$300 0.25x 2x 365
54750
Sub-total: 108734
Total: 109034
- 67 -
Appendix H: Estimated cost for the innovation annually (cont’d)
Estimated additional annual running cost in the subsequent year
Items Price per unit (HKD) Quantity Amount (HKD)
Set-up cost:
Rooms for training 0 1 0
Chairs for training 0 15 0
Computer 0 1 0
projector 0 1 0
Stationary (bought in the first year)
0 10 0
Hard copy guideline (printed in the first year)
0 6 0
Electronic copy guideline 0 1 0
Poster for promotion (printed in the first year)
0 5 0
Checklist for new staff competency
1 10 10
Material cost:
Salary of new staff for training (30 minutes)
$300 0.5x10 1500
Probiotic 245/90 (one capsule)
140x60x2 45734
Salary of staff for milk preparation (15 minutes each day)
$300 0.25x 2x 365
54750
Total: 101994
- 68 -
Appendix I: Evidence-based practice guideline
Title
The title of the clinical guideline is named as “Evidence-based Guideline on Using
Oral probiotic to prevent Late-Onset Sepsis in Preterm Infants”
Background of clinical issue
Late-onset sepsis is a common problem in extremely preterm and very preterm
infants with delayed natural gastrointestinal colonization (Mshvidadze et al., 2010;
Schwiertz et al., 2003). The use of oral probiotic is shown to be effective in preventing
neonatal late-onset sepsis in different randomized controlled trial studies (Demirel et al.,
2013; Jacobs et al., 2013; Lin et al., 2005; Manzoni et al., 2009; Oncel et al., 2014;
Samanta et al., 2009). Besides that, the use of probiotic can also reduce incidence rate of
NEC≥ stage 2 (Jacobs et al., 2013; Lin et al., 2005; Manzoni et al., 2009; Oncel et al., 2014;
Samanta et al., 2009), infant mortality (Lin et al., 2005; Manzoni et al., 2009; Samanta et
al., 2009) and shorten length of hospitalization (Oncel et al., 2014; Samanta et al., 2009).
Aim
To develop an evidence-based guideline on using oral probiotic for preventing
late-onset sepsis in preterm infants
Objective
The objectives of this guideline are to:
- 69 -
To summarize all available high level of evidences for the use of oral probiotic on
preterm infants
To formulate and evidence-based guidelines for NICU and SCBU nurses on using
oral probiotic on preterm infants
To streamline and standardize the clinical practice on using oral probiotic on
preterm infants
Target Users
This guideline is intended to provide instructions for NICU and SCBU nurses of all
level on the usage of oral probiotic in preterm infants.
Target group
Preterm infants who are admitted to NICU or SCBU with gestational age ≤ 32
weeks and/ or birth weight ≤ 1500 grams, excluding those with major congenital
anomalies.
Recommendations
The recommendations are graded according to Scottish Intercollegiate
Guidelines Network (SIGN).
Recommendation 1
1.1 The eligible population should include preterm infants with gestational week 32
weeks and/or with birth weight 1500 grams.
(Grade of recommendation: A)
- 70 -
Evidences:
Infants with gestational age < 32 weeks [Samanta et al., 2009 (1+), Demirel et al.,
2013 (1+), Jacobs et al., 2013 (1+), Oncel et al., 2014 (1+)] and/ or birth weight < 1500g
[Lin et al., 2005 (1+), Samanta et al., 2009 (1+), Manzoni et al., 2009 (1++), Demirel et al.,
2013 (1+), Jacobs et al., 2013(1+), Oncel et al., 2014 (1+)] were eligible population to the
intervention.
1.2 Intervention should start when eligible population started enteral feeding.
(Grade of recommendation: A)
Evidences:
Four studies mentioned that oral probiotic should be used when infant started to
feed enterally [[Lin et al., 2005 (1+), Samanta et al., 2009 (1+), Demirel et al., 2013 (1+),
Oncel et al., 2014 (1+)].
1.3 The patient should be excluded from the innovation if they have major congenital
anomalies.
(Grade of recommendation: A)
Evidences:
Four studies exclude infants with major congenital anomalies. [Samanta et al.,
2009 (1+), Demirel et al., 2013 (1+), Jacobs et al., 2013(1+), Oncel et al., 2014 (1+)].
- 71 -
Recommendation 2
2.1 Probiotic contains Lactobacillus and/or Bifidobacterium infantis species should be
used.
(Grade of recommendation: A)
Evidences:
Five studies use probiotic that contains Lactobacillus species [Lin et al., 2005 (1+),
Samanta et al., 2009 (1+), Manzoni et al., 2009 (1++), Oncel et al., 2014 (1+)] and/or
Bifidobacterium infantis species [Lin et al., 2005 (1+), Samanta et al., 2009 (1+), Jacobs et
al., 2013(1+)].
2.2 Probiotic mixture of Bifidobacteria infantis, Bifidobacteria bifidum, Bifidobacteria
longum and Lactobacillus acidophilus, each 2.5 billion CFU should be used.
(Grade of recommendation: B)
Evidences:
After comparing the effect on the incidence rate of late onset sepsis in all six
studies found, probiotic with mixture of Bifidobacteria infantis, Bifidobacteria bifidum,
Bifidobacteria longum and Lactobacillus acidophilus, each 2.5 billion CFU provided a 93%
of RRR significantly (p=0.020) [Samanta et al., 2009 (1+)].
2.3 The probiotic mixture should administrate according to infant’s body weight.
(Grade of recommendation: B)
- 72 -
Evidences:
After comparing the RRR in the incidence rate of late onset sepsis in all six
studies found, probiotic used in Samanta et al. (2009) provided the highest RRR.
Samanta et al. (2009) suggested that the probiotic mixture should be administrated
according to infant’s body weight *Samanta et al., 2009 (1+)].
2.4 The probiotic mixture should be administrated at a dose of 125g/kg twice daily.
(Grade of recommendation: B)
Evidences:
After comparing the RRR in the incidence rate of late onset sepsis in all six
studies found, probiotic used in Samanta et al. (2009) provided the highest RRR.
Samanta et al. (2009) suggested that the probiotic mixture should be administrated
twice daily with the dosage being 125g/kg till discharge [Samanta et al., 2009 (1+)].
Recommendation 3
3.1 Intervention should start when infants started enteral feeding with stable vital
signs, active bowel sound without abdominal distension and no bile or blood aspirated
from gastric tube.
(Grade of recommendation: A)
Evidences:
- 73 -
In four of the studies found, intervention should only start when infants started
enteral feeding with stable vital signs, active bowel sound without abdominal distension
and no bile or blood from gastric tube [Lin et al., 2005 (1+), Samanta et al., 2009 (1+),
Demirel et al., 2013 (1+), Oncel et al., 2014 (1+)]
3.2 Intervention should stop when signs of feeding intolerance were observed.
(Grade of recommendation: B)
Evidences:
Interventions were stopped in three of the studies found when signs of feeding
intolerance were observed in infants [Lin et al., 2005 (1+), Demirel et al., 2013 (1+),
Oncel et al., 2014 (1+)]
3.3 The intervention should end when infant discharge.
(Grade of recommendation: A)
Evidences:
In five of the studies found, intervention were ended when infants discharged
from hospital [Lin et al., 2005 (1+), Samanta et al., 2009 (1+), Demirel et al., 2013 (1+),
Jacobs et al., 2013(1+), Oncel et al., 2014 (1+)].
3.4 The probiotic should mix with breast milk or formula milk.
(Grade of recommendation: A)
- 74 -
Evidences:
The probiotic was mixed with breast milk [Lin et al., 2005 (1+), Samanta et al.,
2009 (1+), Manzoni et al., 2009 (1++), Demirel et al., 2013 (1+), Jacobs et al., 2013(1+)]
or formula milk [Manzoni et al., 2009 (1++), Demirel et al., 2013 (1+)] before
administrate to infants.
3.5 The probiotic can be administrated orally or through oro-gastric tube.
(Grade of recommendation: A)
Evidences:
In the six studies found, five of them mentioned that the probiotic were
administrated orally [Lin et al., 2005 (1+), Samanta et al., 2009 (1+), Manzoni et al., 2009
(1++), Demirel et al., 2013 (1+), Jacobs et al., 2013(1+)] or through orogastric tube
[Manzoni et al., 2009 (1++), Jacobs et al., 2013(1+), Oncel et al., 2014 (1+)]. Oncel et al.
(2014) stated that suctioning oral secretions should be done before administrating
probiotic and the probiotic was placed in the posterior oropharynx of the infants. For
infants without per oral feeds, probiotic were administered through a gastric tube
followed by a flash of 0.5 ml sterile water [Oncel et al., 2014 (1+)].
- 75 -
Appendix J: Schedule for implementation and evaluation Timeslot Phase Action
Week 1 Approval seeking
Hold meetings to explain the innovation Form the committee
Week 2 Get support and introduce the innovation to the staff
Hold briefing sessions to the intended users Introduce the innovation in the multi-disciplinary
team meeting Email details of innovations to all staffs in
department Provide hardcopy and electronic copy of
guideline to staff
Week 3-8 Pilot study and evaluation
Recruit eligible patients
Week 3-16 Perform intervention Identify unexpected difficulties Collect feedback from staff
Week 17 Analyze data collected in the pilot study
Week 18 Evaluate the pilot study and refine the guideline
Week 19-28 Main study and evaluation
Recruit eligible patients
Week 19-36 Perform intervention Collect relevant clinical data
Week 24,28,32,36
Assess the staff satisfaction level
Week 37-41 Data analysis Prepare evaluation report
Week 42-43 Present and discuss the results found to the administrators
Refine and finalize the guideline
- 76 -
Appendix K: Gantt chart for implementation plan and evaluation Week
Item 1
2 3
4 5
6 7
8 9 10
11 12
13 14
15 16
17 18
19 20
21 22
23 24
25 26
27 28
29 30
31 32
33 34
35 36
37 38
39 40
41 42
43 44
Approval seeking
Staff briefing
Pilot study recruitment
Pilot study intervention
Pilot study evaluation
Main study recruitment
Main study intervention
Main study evaluation
- 77 -
Appendix L: Checklist for the use of oral probiotic to prevent late-onset sepsis in preterm infants Date: ________________________
Please tick the appropriate box.
Content Pass Fail Remarks
Recruitment of eligible population
Gestation week 32 weeks and/or with birth weight 1500 grams
Free from major congenital anomalies
Started enteral feeding
Parental verbal content obtained
Preparation of oral probiotic
Correct dose of probiotic mixture powder (125g/kg) prepared
Probiotic mixture powder is well mixed with infant’s milk
Preparation procedures are clean and smooth
Application of oral probiotic
Correct time of administration
Correct frequency of administration (twice daily)
The vital signs of infants are stable
Correct route of administration (orally or oro-gastric tube)
Other comments:
- 78 -
Appendix M: Staff satisfaction survey on the guideline of using oral probiotic in preterm infants Date: ____________________________
Rank: ____________________________
Please circle the most appropriate answer
Item Strongly Disagree
Slightly Disagree
Neutral Slightly Agree
Strongly Agree
1. The briefing session is useful and clear. 1 2 3 4 5
2. The guideline can improve the quality of nursing care in the department.
1 2 3 4 5
3. The guideline is clear and easy to follow. 1 2 3 4 5
4. The workload of new guideline is acceptable. 1 2 3 4 5
5. I am competent in carrying out the new guideline 1 2 3 4 5
6. The resource manual is easy to locate. 1 2 3 4 5
7. The resource manual is useful. 1 2 3 4 5
8. There is adequate support from the committee members. 1 2 3 4 5
9. The new guideline is effective in reducing late-onset sepsis in preterm infants.
1 2 3 4 5
10. In general, I am satisfied with the implementation of new guideline.
1 2 3 4 5
11. Any other comments or suggestion?
_________________________________________________________________________________________________________________________________
Total score (for official use only):_________