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):_________