929 Feeding guideline

Evidence-Based Practice Guideline

for the

Management of Feeding

in Monash Newborn

13 November 2008

Developed with support from the William Buckland Foundation

“Ensuring Infants in Monash Newborn Receive the Best Possible Care”

© Southern Health Clinical Practice Guideline for the Management of Feeding in Monash Newborn

Disclaimer:

The information in this report is a summary of that available and is primarily designed to give

readers a starting point to consider currently available research evidence. Whilst appreciable

care has been taken in the preparation of the materials included in this publication, the authors

and Southern Health do not warrant the accuracy of this document and deny any

representation, implied or expressed, concerning the efficacy, appropriateness or suitability of

any treatment or product. In view of the possibility of human error or advances of medical

knowledge the authors and Southern Health cannot and do not warrant that the information

contained in these pages is in every aspect accurate or complete. Accordingly, they are not

and will not be held responsible or liable for any errors of omissions that may be found in this

publication. You are therefore encouraged to consult other sources in order to confirm the

information contained in this publication and, in the event that medical treatment is required,

to take professional expert advice from a legally qualified and appropriately experienced

medical practitioner.


Contents 3

Contents Key principles ............................................................................................................... 5 1. Introduction .............................................................................................................. 7

1.1 Background ....................................................................................................... 7 1.2 Development ..................................................................................................... 7 1.3 Scope ............................................................................................................... 8 1.4 Exclusions ......................................................................................................... 8 1.5 Users ............................................................................................................... 8 1.6 Statement of intent ............................................................................................ 8

2. Establishing readiness for milk feeds .......................................................................... 11 2.1 Readiness for enteral feeds ............................................................................... 11 2.2 Readiness for introduction of sucking feeds ......................................................... 18

3. Trophic feeds (minimal enteral nutrition) .................................................................... 27 4. Tube feed methods .................................................................................................. 30

4.1 Transpyloric or gastric tube feeds ...................................................................... 30 4.2 Nasogastric or orogastric tube feeds ................................................................... 31 4.3 Intermittent or continuous feeds ........................................................................ 32 4.4 Indwelling or intermittently placed tubes ............................................................ 33 4.5 Rate of introduction and advancement of milk feeds ............................................. 34 4.6 Frequency of feeds ........................................................................................... 36 4.7 Maximum tube feed volume per day ................................................................... 36 4.8 Size of feeding tubes ........................................................................................ 38 4.9 Positioning for tube feeds .................................................................................. 38

5. Sucking feed methods .............................................................................................. 41 5.1 Breast or bottle feeding .................................................................................... 41 5.2 Non-nutritive sucking ....................................................................................... 44 5.3 Demand or scheduled feeds .............................................................................. 47 5.4 Breastfeeding methods ..................................................................................... 50 5.5 Methods of bottle feeding .................................................................................. 61

6 Feed intolerance and necrotising enterocolitis ............................................................... 67 6.1 Necrotising enterocolitis .................................................................................... 67 6.2 Gastric residuals .............................................................................................. 68 6.3 Abdominal girth ............................................................................................... 70 6.4 Vomiting ......................................................................................................... 71 6.5 Signs of feed intolerance................................................................................... 72

7. Breastmilk or formula ............................................................................................... 73 7.1 Benefits of breastmilk ....................................................................................... 73 7.2 Availability of breastmilk ................................................................................... 75

8. Role of parents in establishing feeding ........................................................................ 77 8.1 Importance of parent involvement ..................................................................... 77 8.2 Communication, information and education ......................................................... 78 8.3 Encouragement, positive feedback and support ................................................... 79 8.4 Summary ........................................................................................................ 79

9. Dissemination and implementation ............................................................................ 80 9.1 Dissemination.................................................................................................. 80 9.2 Implementation ............................................................................................... 81

10. Outcomes and audit ............................................................................................... 82 10.1 Process measures .......................................................................................... 82 10.2 Outcome measures ........................................................................................ 82

11. Guideline development process ................................................................................ 83 11.1 Guideline Development Group ......................................................................... 83 11.2 Methodology .................................................................................................. 85 11.3 Searches ....................................................................................................... 87 11.4 Critical appraisal tables ................................................................................... 88

12. References ............................................................................................................ 89 Annex 1 Acronyms and Abbreviations ............................................................................ 93 Annex 2 Definitions ...................................................................................................... 94


4


5

Key principles This guideline provides a guide to establishment of feeding in Monash Newborn. While some

aspects of the guideline are quite technical and very detailed, there are three key principles

that underpin all of the recommendations.

Establishment of feeding infants in Monash Newborn is based on

supporting breastfeeding and provision of breastmilk

clinicians and parents working closely together

actively planning progression of feeding and communicating plans clearly.

Supporting breastfeeding and provision of breastmilk

Breastfeeding and breastmilk have many benefits for both mother and baby, so all care in

Monash Newborn is designed to support provision of breastmilk and establishment of

breastfeeding.

In Monash Newborn we encourage and support mothers to provide breastmilk and skin-to-skin

contact. We look for opportunities for breast contact and non-nutritive sucking at the breast

and are flexible in timing feeds to support breastfeeding when mothers are available. We also

avoid actions that might make it less likely that infants establish breastfeeding, so where

possible we avoid introducing formula or bottle feeds.

Clinicians and parents working closely together

In Monash Newborn we know that every infant and every family is unique. To ensure we meet

the needs of infants and their families, we work closely with parents. As part of this process,

throughout the introduction and establishment of feeding, we discuss and provide information

on the potential options with parents. We record and respect the decisions that parents make

about the care of their infant and support them to reach the goals they set.

Recognising the central importance of the relationship between parents and their infant, we

encourage parents to be involved in delivering care whenever possible. This might include

holding the infant during tube feeds or providing skin-to-skin contact.

Actively planning progression of feeding and communicating plans clearly

In Monash Newborn, clinicians and parents regularly discuss, document and review plans for

introducing, increasing and managing feeds.

In Monash Newborn we develop feeding plans as a multidisciplinary team, including the

parents. These feeding plans are clearly documented in the infant‘s medical record and

reviewed at least once every 24 hours.

We work together to look for opportunities to support infants to progress towards establishing

feeding and we support and encourage families at each step in the process.

In Monash Newborn we aim to give infants the best possible care and their families the best

possible support. This guideline is one part of our work to achieve that aim.


6


Introduction 7

1. Introduction

1.1 Background

Infants in neonatal intensive care units (NICUs) and special care nurseries are often initially

unable to have normal sucking feeds. These infants may require intravenous nutrition until

they are ready to digest milk, and tube feeds until they are able to gain all their nutritional

requirements through breast or bottle feeds.

Establishing mature feeding behaviour and independence from tube feeds is a pre-requisite for

infants to be discharged home. Achieving this goal can require a lengthy process consisting of

many weeks between commencement of milk feeds and establishment of full sucking feeds.

International studies have found substantial differences in length of hospital stay of infants in

NICU. The majority of these differences are thought to be due to variations in care.

Recognition of maturity of feeding and control of breathing are likely to be key factors

contributing to the differences in discharge timing.

More effective, standardised strategies for the introduction and establishment of oral feeding

have the potential to significantly reduce length of stay for infants in NICU and special care

nurseries. Earlier discharge will result in benefits to babies and their families, improved access

to neonatal cots, reduced pressure on neonatal services and significant cost savings to health

services.

1.2 Development

This guideline was developed by the Monash Newborn Feeding Guideline Development Group,

Monash Medical Centre, Clayton, Southern Health in 2007. Monash Newborn includes both the

NICU and special care nursery. The guideline was developed in response to variability in

practice in management of feeding which was identified by senior nursing and medical staff in

Monash Newborn.

A Monash Newborn Feeding Guideline Steering Group (the Steering Group) was established to

facilitate and direct the development, implementation and evaluation of evidence-based

feeding guidelines in Monash Newborn at Monash Medical Centre, Southern Health. The scope

of the guideline was determined by the Steering Group. Details of the composition of the

Steering Group are provided in Chapter 10.

The process undertaken to develop this guideline was evidence-based. Initially a search was

undertaken to identify existing evidence-based guidelines which could be adapted for local use.

The published literature, key clinical practice guideline websites and websites of a variety of

professional organisations were searched. No high quality evidence-based guidelines for the

management of feeding (as assessed by the AGREE Criteria1) were identified.

A multidisciplinary Guideline Development Group (GDG) was convened that included senior

and junior medical and nursing staff and relevant allied health staff from Monash Newborn, as

well as parent representatives and specialists in evidence-based practice and guideline

development. Chapter 10 provides further details of the composition of the GDG.

The GDG created a list of clinical questions that should be addressed in a guideline for the

management of feeding in infants in Monash Newborn, and systematic searches were

undertaken to identify evidence to answer these questions. Where evidence was found the

GDG made recommendations based on this evidence, integrated with clinical expertise and

parent advice and perspectives. Where evidence was not found to answer a clinical question

the GDG made a consensus recommendation based on clinical expertise and parent

preferences.


Introduction 8

The levels of evidence found, and grades of recommendation made are identified in the text of

the guideline, as described in Table 1. A full description of the development process can be

found in Chapter 10.

It is intended that this guideline be updated every 5 years, to reflect changes in the available

evidence and any relevant local changes. This guideline will therefore be due for review in

2012.

The most recent edition of this guideline, including any recent updates, is available on the

Southern Health website.

1.3 Scope

The Steering Group agreed that the guideline would focus on the process of establishing

feeding in infants in both the NICU and the special care nursery and would exclude questions

about the content of feeds (types of formula, nutritional content, fat type, etc).

1.4 Exclusions

This guideline has not been developed to apply to:

Infants of Diabetic Mothers (including mothers with gestational diabetes). Infants of

Diabetic Mothers (IDM) in Monash Newborn should be managed according to the

Southern Health Clinical Pathway for Infants of Diabetic Mothers.

Infants with congenital anomalies that have a direct impact on their feeding

management.

Infants requiring surgery which has a direct impact on their feeding management.

Infants with a neurological condition which has a direct impact on their feeding

management.

Infants with severe gastro-oesophageal reflux.

Some of the content of this guideline may be appropriate for infants in these groups, however

clinicians should take extra caution.

Clinicians should also take extra caution with infants with significant relevant co-morbidities or

chronic illnesses and consult with appropriate specialists.

1.5 Users

This guideline is intended for use by clinicians involved in the care and management of feeding

in infants in Monash Newborn. It has been primarily developed for use in Monash Newborn at

Monash Medical Centre Clayton in south east Melbourne, Australia. Adaptation for local use may

be necessary in other contexts.

1.6 Statement of intent

This guideline has been developed to ensure infants in Monash Newborn receive the best

possible feeding management, based on the best available evidence integrated with clinical

expertise and parent/carer advice and perspectives, and consistent with World Health

Organisation recommendations on supporting breastfeeding in the care of ill and very preterm

babies. The guideline aims to promote consistency of care and to provide clinicians in Monash

Newborn with recommendations for the management of the process (but not the content) of

feeding in infants. The guideline also aims to form the foundation of information on

establishment of feeding for parents and carers of infants in Monash Newborn.

Guidelines are a tool used to improve patient care and do not replace the central role of clinical

expertise and judgement in determining appropriate patient care. Clinicians should apply this


Introduction 9

guideline in the context of the individual patient and their family. Variations from this guideline

should be documented in the patient‘s medical record at the time the relevant decision is

made.

The recommendations in this guideline were based on the best evidence available at the time

of searching, and should be read with an awareness of any more recent evidence.

A number of supporting documents have been developed to facilitate the implementation of

the recommendations of this guideline into practice. These include a feeding plan, summaries

for hospital staff and evidence-based information for parents and carers of infants in Monash

Newborn. These documents are available from Monash Newborn.


Introduction 10

Table 1: Key to evidence statements and grades of recommendation2

Levels of evidence

1++ High quality meta-analyses, systematic reviews of randomised controlled

trials (RCTs), or RCTs with a very low risk of bias

1+ Well conducted meta-analyses, systematic reviews of RCTs, or RCTs with a

low risk of bias

1- Meta-analyses, systematic reviews of RCTs, or RCTs with a high risk of

bias

2++ High quality systematic reviews of case-control or cohort studies.

High quality case-control or cohort studies with a very low risk of

confounding 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, eg case reports, case series

4 Expert opinion

Grades of recommendation

Note: The grade of recommendation relates to the strength of the evidence on

which the recommendation is based. It does not reflect the clinical importance of

the recommendation.

A At least one meta-analysis, systematic review of RCTs, 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

B 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 1++ or 1+

C 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 2++

D Evidence level 3 or 4; or Extrapolated evidence from studies rated 2+; or

The consensus opinion of the guideline development group

Good practice points Recommended best practice based on the clinical experience of the

guideline development group


Establishing readiness for milk feeds 11

2. Establishing readiness for milk feeds Establishing feeding is fundamental to the management of infants NICUs and special care

nurseries. Feeding impacts on many aspects of an infant‘s health, development and ability to

go home safely from Monash Newborn. The question of how to decide when an infant is ready

to feed is of central importance in establishing feeding.

To emphasise the importance of planning and monitoring establishment of feeding, the GDG

agreed that:

An infant‘s readiness for feeds and progress in establishing feeding should be

evaluated at least once every 24 hours.

A clear plan for introduction and establishment of feeding should be documented for

all infants.

While many articles address the issue of readiness to feed in infants in NICUs and special care

nurseries, there is little high quality research available on which to base a set of criteria for

determining readiness for feeding. The literature in this area is complex and often

methodologically weak. The varying definitions of commonly used terms add further

complexity. It is clear from the literature that the process of determining readiness to feed is

complicated and not yet well understood.

Readiness for milk feeds has two components; readiness for any enteral feeds, that is

provision of milk into the gastrointestinal tract by any method; and readiness for sucking

feeds, either from breast or bottle.

2.1 Readiness for enteral feeds

Most studies investigating readiness for enteral feeds focus on factors which impact on the

development of feed intolerance or incidence of necrotising enterocolitis (NEC). Some studies

have also investigated factors which delay onset of enteral feeding.

The majority of studies assume that infants should be haemodynamically or physiologically

stable and have no signs of sepsis before enteral feeds are considered. These assumptions

have not been tested in high quality comparative studies and may not be appropriate in clinical

practice.

In the absence of evidence the GDG agreed that clinicians should:

D Consider the possibility of enteral feeds for all infants.

Many factors have been suggested as impacting on readiness for enteral feeds, these include:

age

availability of breastmilk

intra-uterine growth restriction

absence or reversal of end-diastolic flow

perinatal asphyxia

presence of umbilical artery or venous catheters

respiratory support

intestinal motility

The evidence for most of these is weak and guidelines and protocols from other hospitals use a

variety of combinations of these criteria.



2.1.1 Age

Several, mostly low quality, studies have investigated the impact of ―early‖ versus ―late‖

introduction of enteral feeds. Definitions of ―early‖ and ―late‖ differ widely between studies,

and even overlap. ―Early‖ may refer to introduction of enteral feeds as soon as the first hour

after birth or up to 14 days after birth. ―Late‖ may mean delaying introduction of enteral feeds

for anywhere between 5 and 28 days after birth.

A Cochrane systematic review3 investigating early versus delayed initiation of

progressive enteral feeds for parenterally fed low birth weight or preterm infants

found only two small randomised controlled trials (RCTs) with a total of 72

patients which met inclusion criteria. In one of these studies, early feeding was

initiated at a median age of 2 days, and late feeding at a median of 5 days. In the

other study early feeding was commenced within 4 days of birth and late feeding

was commenced at or after 10 days of age. Feeds were then progressively

increased.

The small number of infants included in these studies limits the ability to detect

clinically important differences between the groups. Early feeds had no significant

effect on weight gain, NEC, mortality, or age at discharge. In one trial some

benefits of early feeding were found, such as fewer days on parenteral nutrition,

fewer infants who were treated with gastric suction and interruption of feeds,

fewer infants with sepsis evaluations, and fewer infants with percutaneous central

venous catheters.

The authors of the review conclude ―For such a fundamental issue in the care of

sick preterm infants, we have embarrassingly little data on which to base decisions

about when to begin enteral feedings.‖

1+

A number of other less well conducted studies have suggested that delaying

enteral feeds may have potentially detrimental effects.

In one prospective cohort study4 examining the risk of feed intolerance in 132 very

low birth weight (VLBW) infants following initiation of enteral feeds within the first

72 hours of life, every hour delay in enteral feeds increased the risk of feed

intolerance. In this study after controlling for confounders, age at initiation of

enteral feeds was the only factor significantly associated with feed intolerance. No

difference was seen in the rate of NEC.

A retrospective cohort study7 in 385 infants (<1500g) found that early enteral

feeds initiated in the first or second day of life was associated with decreased risk

of sepsis without increased risk of NEC. Enteral feeding was started at an earlier

age in infants who did not develop sepsis compared to those who did (2.8 vs 4.8

days, p=0.0001). The findings were similar for subgroups of infants <1000g and

those with respiratory distress syndrome.

A second retrospective cohort study5 in 36 premature infants (<37 weeks) with

respiratory distress in two hospitals found that infants in the hospital where

enteral nutrition was commenced at a mean of 0.7 days after birth compared with

infants in a hospital where enteral nutrition was commenced at a mean of 4.9 days

after birth, reached full enteral feeds sooner, required fewer days of respiratory

support, took fewer days to regain birth weight, and had fewer total days in

hospital. There was no difference in the rate of NEC between the groups. While

there were no differences between the groups for gestational age, birth weight,

Apgar score or respiratory distress syndrome severity, these results may represent

other underlying differences between the groups.

2-



Trophic feeding of small, non-nutritional volumes of milk, is another way in which

enteral feeds can be introduced ―early‖. A second Cochrane systematic review6

investigating trophic feeds for parenterally fed infants included ten RCTs

comparing early trophic feeds to no enteral feeding. In these studies trophic feeds

were commenced as early as the first day and as late as the eighth day after birth

and was continued without progression for 5-10 days.

In infants given trophic feeds there was an overall reduction in:

mean days to full enteral feeding (weighted mean difference

(WMD)=2.6 days),

total days that feeds were withheld (WMD=3.1 days),

total hospital stay (WMD=11.4 days).

There was no significant effect on NEC (relative risk (RR)=1.16 (95%CI 0.75,

1.79). There was inconsistency in the treatment effect between studies for days to

full enteral feeding, days to regain birth weight, days of phototherapy, and days of

hospital stay. The explanation for this inconsistency is unclear.

On the basis of these findings and given the intermediate quality of the included

trials the authors concluded that there was a ―plausible rationale‖ for trophic feeds

and that the review ―suggested benefits‖ however they urged cautious

interpretation as the results do not rule out an increase in NEC.

1++

On the basis of the little evidence available and the consensus of the GDG, members of the

GDG agreed that:

C Enteral feeds should not be delayed until an infant reaches a certain corrected age.

While exact figures vary, there is agreement in guidelines and protocols from other neonatal

units suggesting that unless there are particular concerns, enteral feeding (either trophic or

nutritional) can begin as early as the first day of life and should begin in the first 2-8 days of

life.13,14,15

In the absence of evidence the GDG agreed that:

D Enteral feeds should be commenced in the first 48 hours of life unless precluded by

other criteria.

2.1.2 Availability of breastmilk

In light of the many benefits of breastmilk (discussed further in Chapter 7), some authors have

suggested that for those infants whose mothers wish to provide breastmilk, enteral feeds be

delayed until breastmilk is available. No studies were identified which evaluated the impact of

this approach.

While there is evidence from prospective cohort studies10,11 that mother‘s breastmilk may

reduce NEC compared to formula milk, and from a systematic review of RCTs12 that donor

breastmilk may reduce NEC compared to formula milk, these studies do not examine the

impact of withholding enteral feeds until breastmilk is available. Also, the protective effect of

breastmilk in reducing NEC appears to be maintained where supplementary formula milk is

also supplied.10

One study10 examining the effect of breastmilk and formula in the development of

NEC found that in formula fed babies delaying the onset of the first enteral feed 2+



was associated with a marginally statistically significant lower incidence of NEC

(size of effect not provided, p=0.05), however for infants fed breastmilk (alone or

with formula) there was no relationship.

As discussed in 2.1.1, delaying enteral feeds may have negative impacts on the infant‘s

development.


D While enteral feeds should usually be started within 48 hours, they may be withheld

for up to 72 hours after birth while waiting for breastmilk to be available if the

mother wishes.

See section 7.2 for recommended practice if breastmilk is not available 72 hours after birth.

2.1.3 Intra-uterine growth restriction

While some authors suggest that enteral feeding protocols should be altered for infants with

intra-uterine growth restriction (IUGR), no studies were identified that examined this. Studies

have been conducted in infants with IUGR and absent or reversed diastolic flow, and these are

discussed in the following section. The GDG noted that it can be difficult to establish feeding in

infants with IUGR and that caution should be exercised with these infants.


D Enteral feeds should not be delayed in infants with intra-uterine growth restriction if

they meet all other criteria.

Infants with intra-uterine growth restriction may take longer to establish feeding.

Take special care when introducing feeds in these infants.

2.1.4 Absence or reversal of end-diastolic flow

No studies were identified which investigated the benefit of delaying introduction of enteral

feeds for infants with absent or reversed end-diastolic flow (AREDF). Five studies26-30 were

identified which investigated the relationship between AREDF and NEC or feed tolerance. Most

of these studies included only small numbers of infants, limiting their ability to detect

differences between the AREDF and non-AREDF groups.

A retrospective matched cohort study28 was identified that compared 36 infants

with AREDF to 36 infants without AREDF matched for gestational age and birth

weight. The study found no difference in rates of NEC between groups. There were

no differences in the proportion of infants who had their first enteral feed within

their first day of life or the proportion of infants with feed intolerance. The mean

length of stay (LOS) and proportion of infants receiving parenteral nutrition were

also similar between groups.

2+

In a retrospective cohort study26 including 578 infants with IUGR, with and without

AREDF, no differences were found in rates of NEC between groups. Infants with

IUGR and AREDF had higher rates of intrauterine death, neonatal death and

respiratory distress syndrome (RDS) and longer LOS than those without AREDF. It

is possible that these associations are confounded by the higher birth weight and

gestational age in the group of infants without AREDF.

Another retrospective matched cohort study27 compared 40 infants with AREDF to

2-



40 infants of matched gestational age without AREDF. The authors did not find a

significant difference in rates of NEC between groups. The time taken until

tolerating 100ml/kg/day of milk was higher in the AREDF group (18 vs 11 days,

p=0.001), however this may be confounded by the fact that control infants had a

substantially higher birth weight, and it is also not clear what feeding protocol was

followed, or if introduction of enteral feeds was routinely delayed in infants with

AREDF.

A third retrospective matched cohort study29 compared 61 case infants with AREDF

to 61 control infants of matched sex and gestational age without AREDF. The

infants with AREDF had substantially lower median birth weight than infants

without AREDF (1014g vs 1483g). The study found no differences between groups

in mortality, need for mechanical ventilation or arterial pH. Case infants were more

likely to receive parenteral feeds and had enteral feeds introduced later. The

difference in incidence of NEC between the groups was not statistically significant,

7 (11%) cases had suspected NEC compared to 1 (2%) control, p=0.07. If

analysis was restricted to NEC confirmed by radiology there was no difference

between the incidence of NEC in cases (3%) and controls (2%). An analysis not

planned in the original study design including only the 58 case infants with normal

karyotype and their controls, found suspected NEC was more likely in the case

group, (10% vs 0%, p=0.03). Any relationship is likely to be confounded by the

substantial variation in birth weight between the two groups.

A fourth retrospective matched cohort study30 included 15 infants with AREDF

compared with 15 controls matched for gestational age and birth weight. The

study found no significant difference between groups for proven NEC (33% cases

compared to 7% controls), but a significant difference in NEC with 53% cases and

7% of controls having proved or suspected NEC, RR=8 (95%CI 1.14, 56.33).

These studies vary too much in methodology, inclusion and exclusion criteria for their results

to be meaningfully combined. While the evidence suggests that infants with AREDF may be

more likely to develop NEC, there is no evidence to evaluate whether delaying introduction of

enteral feeds is beneficial in these infants.

In light of the small amount of available evidence the GDG agreed that:

D Enteral feeds should not be delayed in infants with absent or reversed end-diastolic

flow if they meet all other criteria.

D Clinicians should take special care to monitor for feed intolerance and signs of NEC

when initiating enteral feeds in infants with absent or reversed end-diastolic flow.

2.1.5 Perinatal asphyxia

It has been suggested that the damage caused by asphyxia to the infant‘s neural system may

have a detrimental effect on gastrointestinal motor activity and, therefore, feeding tolerance.8

Three studies were identified which investigated the impact of perinatal asphyxia on readiness

to feed.4,8,9 We were unable to retrieve full text of one of these papers.9

A prospective cohort study including 132 VLBW infants4 found a non-significant

association between perinatal asphyxia and feed intolerance. Thirty-six percent of

infants with feeding intolerance had perinatal asphyxia compared with 21% of

infants without feeding intolerance.

A second prospective cohort study included 25 infants8 requiring ventilator support

2+



with gestational ages between 36 and 44 weeks and birth weights between 1670

and 4510g. Seven of the 11 asphyxiated infants were feed intolerant in the first

week compared to none of the 14 non-asphyxiated infants, p<0.001.

According to the authors of a third study,9 a case-control study of 42 infants with

NEC and 303 infants without NEC, asphyxia was strongly associated with NEC.

However the methodology of this article is very poorly described and some of the

data presented is contradictory. The authors were contacted to request

clarification, however no response was received.

In light of this relatively low quality evidence the GDG agreed that

C Clinicians should take special care to monitor feed intolerance and signs of NEC

when initiating enteral feeds in infants with perinatal asphyxia.

2.1.6 Presence of umbilical arterial or venous catheters

Umbilical catheters can be placed in the umbilical vein or either ‗low‘ at the 4th-5th lumbar

vertebrae, or ‗high‘ at the 8th-10th thoracic vertebrae, in the umbilical artery. The placement of

an umbilical catheter may potentially impact on blood flow to the gastrointestinal tract.

One RCT was found which examined enteral feeding of premature infants with low

umbilical artery catheters in place.20 Twenty-nine infants received enteral feeds at

a median of 2 days of age while the low umbilical arterial catheter (UAC) was in

place, 31 received enteral feeds at a median of 5 days of age 24 hours after the

low UAC was removed.

Infants in the early feeding group had fewer evaluations for sepsis (17 vs 52%,

p=0.005) and fewer episodes of nothing by mouth while gastric suction tubes were

in place (21 vs 45%, p=0.044). There were no significant differences in RDS (100

vs 94%), patent ductus arteriosus (55% in both groups), haemorrhage (17 vs

32%) or NEC (7 vs 13%) between the groups.

1+

In a prospective cohort study4 examining the risk of feed intolerance in 132 VLBW

infants following initiation of enteral feeds within the first 72 hours of life, neither

umbilical arterial or umbilical venous catheterisation were significantly associated

with feeding intolerance. It was not clear in this study whether the arterial

catheters were sited low or high.

2+

In light of this evidence the GDG agreed that:

B Infants with low umbilical artery catheters should still receive enteral feeds if they

meet all other criteria.

C Infants with high umbilical artery catheters or umbilical venous catheters should still

receive enteral feeds if they meet all other criteria.

2.1.7 Respiratory support

The impact on readiness for enteral feeds of respiratory distress, mechanical ventilation and

continuous positive airway pressure (CPAP) is especially important as many infants in Monash

Newborn need assistance to maintain respiratory stability. There is, however, little high quality

evidence and no published guidelines to direct decision making in this area.



One RCT was identified which examined the use of trophic feeds in 29 infants

requiring ventilation.18 Infants in the trophic feeds group received up to 1ml/hr of

formula or breastmilk (dependent on weight) and infants in the control group

received intravenous feeds only until 12 hours after removal of the UAC. After

removal of the UAC feeding was advanced according to the same standard regime

in both groups.

Episodes of feeding intolerance and days to regain birth weight did not differ

between the groups. On average, infants in the trophic feeds group took fewer

days to reach full enteral feeding (10 vs 13 days, p<0.05) and had an increased

weight above birth weight at day 30 (223 vs 95g, p<0.05).

1+

Two studies were identified which investigated the impact of respiratory support

on gastric emptying.16,17 In one cohort study,16 gastric emptying was found to be

delayed in infants with acute respiratory distress (ARD) with gastric retention at

24/39 in infants with ARD and 20/39 in infants without ARD at 48 hours of life.

There was no difference between the groups at 168 hours of life. The clinical

importance of this is unclear as incidence of feeding intolerance was similar in both

groups (16/39 infants with ARD vs 11/39 infants without).

In the second study including 36 infants,17 gastric emptying was found to be

shorter for infants treated with CPAP compared to a control group not receiving

CPAP (28 vs 40 min, p<0.05). There was no difference between the groups in

feeding intolerance or weight gain.

2+

In a prospective cohort study4 examining the risk of feed intolerance in 132 VLBW

infants following initiation of enteral feeds within the first 72 hours of life, neither

ventilatory support nor CPAP therapy were significantly associated with feed

intolerance.

2+

A survey of all NICUs in the United Kingdom undertaken in 199419 found that 71% of units

routinely give enteral feeds to infants less than 33 weeks of gestation during ventilation. A

further 19% of units would feed enterally if ventilation continued for more than 4 weeks.

In light of the small amount of evidence available and their consensus opinion, the GDG

agreed that:

C Infants receiving respiratory support, including those with respiratory distress

syndrome, should be given enteral feeds if they meet all other criteria.

2.1.8 Intestinal motility

2.1.8.1 Intestinal manometry

One study was identified which investigated the ability of intestinal manometry to

predict feed intolerance in infants with respiratory distress in NICU.25 Forty-eight

infants, 40 feed tolerant and 8 intolerant, underwent manometry during a period

of fasting and period of enteral feeding. Feeding intolerance was defined as

vomiting, an increase in abdominal girth greater than 1.5cm, or milk present in

the stomach 2 hours after a feed.

In this small study, abnormal manometry (defined as the combination of motor

quiescence <20min/hour/lead and change in motility index ≤10% between feeding

and fasting manometry responses) had a sensitivity of 1.0 and a specificity of 0.13

in predicting feed intolerance.

2+



In light of the limited evidence and their clinical experience, the GDG agreed that:

Intestinal manometry cannot predict which infants will be feed intolerant with

enough accuracy to be of clinical value.

2.1.8.2 Passing of meconium/first stool

It has been suggested that passage of meconium or the first stool is an important indicator of

gastrointestinal maturity and is a potential indicator of readiness to feed.

One study was identified which examined the association between meconium

passage and feeding tolerance.21 In this study 41 extremely low birth weight

(ELBW) infants were prospectively followed to establish timing of meconium

passage and measures of feeding tolerance. Two-hourly enteral feeding was

commenced at 48 hours of age, starting at 12ml/kg/day and advancing by

12ml/kg/day. Median time to first meconium passage was 31 hours and to the last

meconium passage was at 5.9 days. Time to first meconium passage was not

related to feed volume on day 14, however time to last meconium passage was

inversely correlated with feed volume on day 14, p<0.001. Infants who passed

their last meconium early had a larger volume of enteral feed on day 14.

2+

Three studies were identified that reported time to stool in premature

infants.22,23,24 In the two studies in VLBW infants23,24 delay in passage of first stool

beyond 48 hours was noted in 20.4% and 6.3% of cases. In the study in infants

≤36 weeks, the median age at time of passage of first stool was 18 hours, and 5

infants passed their first stool after 10 days.

2+

These studies suggest that there is wide variability in passage of meconium and/or first stool

and it should not be relied upon as a criterion to establish readiness to feed.

In light of this evidence the GDG agreed that:

B Infants should not be required to pass meconium or first stool before introduction of

enteral feeds.

2.1.8.3 Intestinal sounds

Presence of intestinal sounds has been suggested as an indicator of readiness to feed however

we were unable to identify studies in which it has been evaluated.


D Infants should not be required to have intestinal sounds before introduction of

enteral feeds.

No studies were identified which examined the relationship between intestinal permeability and

readiness for enteral feeds.

2.2 Readiness for introduction of sucking feeds

This section reviews assessment of readiness for introduction of sucking feeds. The term

‗readiness for sucking feeds‘ is also used in the literature in reference to appropriate timing for



each subsequent feed in infants in whom sucking is already established. Assessment of infant

readiness before individual feeds is discussed in Chapter 3.

There are no accepted criteria on which to assess readiness for introduction of sucking feeds,

or to evaluate the accuracy of such assessment. Most studies in this area focus on factors

which impact on the time to transition to full sucking feeds and feeding efficiency.

Several key factors have been suggested as being useful for assessment of readiness for

sucking feeds, these include:

age

weight

gender

demonstrated sucking behaviours

respiratory stability

ability to maintain oxygenation

respiratory rate

apnoeas

work of breathing

continuous positive airway pressure

tolerance of gavage feeds

No systematic reviews or RCTS specifically evaluating the effectiveness of methods of

assessment of readiness for introduction of sucking feeds were identified. The research

available in this area is generally explorative and hypothesis generating.

2.2.1 Gestational age

Gestational age is a commonly used indicator of infant maturity and development. Many

papers discuss the importance of physiological maturation, and therefore age, in development

of the suck, swallow and breathe sequence. It has also been suggested that low gestational

age at birth is an indicator of increased frailty which has an effect on readiness for sucking

feeds, independent of the age after birth at which sucking feeds are commenced.

2.2.1.1 Gestational age at birth

It has been suggested that infants with low gestational age at birth do not have the same level

of maturity at later ages as infants born at higher gestational age. For example, an infant born

at 26 weeks of age may not have the same level of maturity at 36 weeks of age as an infant

born at 34 weeks. This differing pattern of maturation has led some authors to suggest that

gestational age at birth should be considered when assessing an infant‘s readiness for sucking

feeds.33

No studies were found which compared the effectiveness of introducing sucking feeds at times

determined by gestational age at birth.


D Introduction of sucking feeds should not be delayed in infants with low gestational

age at birth if they meet all other criteria.

2.2.1.2 Corrected gestational age at commencement of sucking feeds

Corrected gestational age at commencement of sucking feeds is a very commonly used

indicator of readiness.

Our searching to address this question did not identify any systematic reviews, RCTs or other

comparative studies which sought to compare the effectiveness of different age thresholds in



determining readiness for sucking feeds. One RCT including 29 infants was identified which

compared the effectiveness of early introduction of a standard bottle feeding protocol (48

hours after establishment of full tube feeds) to usual care.31 This RCT was not appraised as the

effects of early introduction of sucking feeds could not be distinguished from the effects of the

standard feeding protocol, however bottle feeds were successfully introduced at a mean age of

31.1 weeks in the experimental group.

There is a very large body of observational literature which includes measures of age and

descriptions of infant feeding practice. This literature has been used to inform practice

development however it does not provide comparisons between different age thresholds for

introduction of sucking feeds and therefore does not provide definitive evidence to direct

decision making. This research uses a wide variety of outcome measures and reports a range

of conflicting results; systematically synthesising it is beyond the resources of this guideline

development project. A selection of the concepts raised in this literature are discussed in Box

1.

When assessing an infant‘s readiness for and progress in establishing feeding use

corrected gestational age.



Box 1. Gestational age and introduction of sucking feeds – some background

information in the absence of high quality evidence

A selection of the concepts raised in the literature regarding introduction of sucking feeds and

gestational age are discussed here to provide information for clinicians. The references

provided are illustrative, and we do not suggest that the arguments they present are proven

or that they present all relevant arguments – only that they raise issues that may be relevant

when making decisions about the age at which to introduce sucking feeds.

Several issues are raised in the literature as potentially affecting the most appropriate age at

which to introduce sucking feeds. These include:

ability to coordinate the suck, swallow, breathe sequence

development of sucking skills

negative consequences of extended tube feeding

development of oro-facial structure to support sucking.

One of the primary issues in successful establishment of sucking feeds is thought to be the

coordination of the suck, swallow, breathe sequence. The age at which the ability to

coordinate this complex sequence is achieved varies between infants. One recent study has

suggested that sucking feeds can be successfully established in infants who cannot coordinate

the suck, swallow, breathe sequence if milk flow rate is restricted.36

In deciding when to introduce sucking feeds there is a tension between delaying feeds until

the suck, swallow, breathe sequence can be coordinated, and therefore physiologic stability

maintained during feeding,35 and introducing sucking feeds before coordination is

demonstrated in order to give the infant an opportunity to practice and improve sucking

skills.

Studies provide conflicting views on whether sucking skills improve with age. Some articles

suggest that more mature infants have improved sucking intensity, duration and frequency38-

39 and others have found no relationship.32 The association between gestational age and

feeding skills is confounded in many studies by the fact that more mature infants have also

had more sucking practice, and studies have suggested that feeding skills in infants improve

with practice.33-34

Delaying the introduction of sucking feeds, and therefore prolonging tube feeds, may

negatively impact on an infant‘s development of feeding skills40 and also increase the risk of

complications such as aspiration, perforation and infection that may result from the use of

feeding tubes.37 On the other hand, delaying the introduction of sucking feeds may have the

benefit of giving infants time to develop the oro-facial structure required to support effective

sucking.

2.2.1.3 Variation in age of introduction with breast or bottle feeds

It has been suggested that the most appropriate age for introduction of sucking feeds might

depend on whether the infant is being introduced to bottle feeds or to the breast. Other than

the RCT of a standard bottle feeding protocol mentioned above,31 no studies of the introduction

of sucking feeds which focused on either bottle or breastfeeding exclusively, or which

compared timing of introduction of breastfeeds to bottle feeds, were found. In all the other

identified studies infants were allowed to both breast and bottle feed.

In the past it was thought that when an infant was breastfeeding, milk was only provided in

response to active sucking and that when an infant was bottle feeding, milk flowed from the

bottle as soon as and for as long as the bottle was inverted; regardless of whether the infant

was sucking or resting.36 It is now understood that milk may be ejected from the breast even

when an infant is not sucking, and that variation between bottles and teats means that in

many cases milk is only provided in response to sucking.



In light of the lack of evidence about differences in timing of introduction of breast and bottle

feeds, the fact that milk may be provided either only in response to sucking or in the absence

of sucking for both breast and bottle-fed infants, and the central importance of supporting the

mother‘s decision to breast or bottle feed, the GDG agreed to recommend the same age

threshold for either breast or bottle feeding.

There is no consensus in the literature about the most appropriate corrected age to commence

sucking feeds. Although many authors suggest 32 or 34 weeks, a survey of NICUs in the

United Kingdom found that age thresholds for introduction of sucking feeds ranging between

29 and 36 weeks were being used.37

In the absence of high quality evidence the GDG agreed to recommend:

D Consider introducing sucking feeds to infants from a corrected gestational age of 32

weeks.

Methods for introducing sucking feeds are discussed in Chapter 5.

2.2.2 Weight

No studies were identified which evaluated the usefulness of weight (either birth weight or

weight at introduction of sucking feeds) as an indicator of readiness for sucking feeds. Authors

have suggested weight thresholds for introduction of sucking feeds ranging from 1000g to

2000g.37

While some studies have found a relationship between birth weight and measures of sucking

efficiency the importance of this relationship in deciding whether or not to introduce sucking

feeds is unclear.

No studies were identified which provided evidence of effectiveness or a plausible biological

basis for use of a particular weight threshold in deciding when to introduce sucking feeds.


D Introduction of sucking feeds should not be delayed until infants achieve a particular

weight if they meet all other criteria.

2.2.3 Gender

It has been suggested that there may be a relationship between an infant‘s gender and the

point at which they are ready to commence sucking feeds. Through references of other studies

we identified one study32 which examined the relationship between gender and sucking feeds.

This study was not identified through our focused searches as it includes no mention of gender

in the title, abstract or key words.

This cohort study investigating development of sucking included 72 infants (34

female, 38 male) born at 26 to 29 weeks gestation. This study was not designed

to specifically investigate the relationship between gender and sucking feeds, but

rather to examine the role of a range of factors in development of sucking skills.

No relationship was found between gender and any of: overall volume of milk

transferred, rate of transfer or number of sucking feeds per day. It is not clear

from the study whether enough infants were included to identify important

variations in sucking skill between genders.

2-

In light of the limited evidence the GDG agreed that:



D Introduction of sucking feeds should not be delayed on the basis of gender if infants

meet all other criteria.

2.2.4 Demonstrated sucking behaviours

It has been suggested that behaviours which demonstrate an infant‘s ability to suck, such as

non-nutritive sucking on gavage tubes, dummies or fingers, and other behaviours such as

bringing the hand to the mouth, the rooting reflex and hunger cues, are useful indicators of

readiness for introduction of sucking feeds.

No research was identified which examined the effect of delaying feeds until one or more of

these sucking behaviours were present, or introducing sucking feeds on the basis of one or

more of these behaviours.

It might be reasonable to suggest that the presence of these behaviours indicates an ability to

suck, but this does necessarily mean the infants will be able to maintain nutritive sucking, nor

does it rule out the possibility that sucking feeds might be successfully introduced in the

absence of these behaviours.


Demonstration of sucking behaviours may suggest readiness for introduction of

sucking feeds.

D Introduction of sucking feeds should not be delayed until sucking behaviours are

demonstrated if infants meet all other criteria.

In infants who are not demonstrating sucking behaviours, plans to introduce sucking

feeds should be discussed with a senior clinician.

2.2.5 Respiratory stability

2.2.5.1 Ability to maintain oxygen saturation

No research was identified which examined the impact of delaying introduction of sucking

feeds until infants can maintain oxygen saturation at a certain level.

Two small prospective cohort studies were identified that investigated the

relationship between oxygen saturation and sucking feeds. 35,42

The first study35 included 22 infants who were receiving 50% or more of their

feeds from bottle or breast. The authors found that infants who were receiving

supplementary oxygen and those with higher pre-feed oxygen saturation were less

likely to experience oxygen desaturation. These two variables explained 44% of

the variation in number of desaturations. The authors of this study noted that

even infants who are able to maintain oxygen saturation in room air may not be

able to maintain oxygenation when physiologic demand is increased by the activity

of feeding.35

The second study42 included 15 infants who experienced desaturation during

sucking feeds and five who did not. (It is not clear from the study whether infants

were bottle or breast-fed.) The authors found that lower oxygen saturations before

2-



and after feeding were associated with desaturation during feeding and that

infants who desaturated had a significantly longer transition time from tube to oral

feeds (11.2 vs 6.6 days, p<0.05).


D Introduction of sucking feeds should not be delayed until an infant can maintain

their oxygenation at a certain level if they meet all other criteria.

2.2.5.2 Respiratory rate


feeds until infants can maintain their respiratory rate at a certain level. The GDG noted that it

many infants high respiratory rate is transitory and may resolve over 24-48 hours. They also

noted that the decision to introduce feeds or otherwise in an infant with respiratory rate above

70 breaths per minute is complex and relies on the interplay of many clinical factors including

the infant‘s weight, gestational age at birth, number of days since birth, etc.


D Introduction of sucking feeds should not be delayed until an infant can maintain

their respiratory rate at a certain level if they meet all other criteria.

D If an infant has a respiratory rate above 70 breaths per minute, consult medical

staff before introducing sucking feeds.

D If an infant has a respiratory rate above 70 breaths per minute, the decision to

introduce sucking feeds should be made with consideration of other factors including

their weight, gestational age at birth, number of days since birth, etc.

2.2.5.3 Apnoeas


feeds on the basis of presence of apnoeas.

Two prospective cohort studies were identified that investigated the relationship

between apnoea and sucking feeds. 43,44

The first study43 examined the medical records of 435 infants admitted to NICU

with an age at birth of 30-34 weeks and no significant medical or surgical

complications. The study found that there was a strong correlation between the

infants‘ post-menstrual age at their last apnoeic episode and their post-menstrual

age at their last tube feed (r2=0.38, p<0.001). That is infants who were older at

their last apnoeic episode were also older when full sucking feeds were

established. The study did not report data on the infants‘ post-menstrual age at

their first sucking feed, so it is not clear whether the delay in establishment of

sucking feeds was due to increased transition time or delayed introduction of

sucking feeds.

The second study44 included 65 infants in 3 NICUs with a gestational age at birth

of 28-34 weeks, who required less than 48 hours of mechanical ventilation and

who had no congenital abnormalities. The authors found that mean transition time

to full sucking feeds was shorter for infants with no apnoea than for those with

apnoea (6.2 vs 11.0 days, p<0.01, standard deviations (SD) not provided). In this

study apnoea accounted for more variation in transition time than any other

2-



variable – including birth weight, gestational age at birth and age at introduction

of sucking feeds. The authors note that apnoea is also associated with decreased

gestational age at birth and decreased birth weight, and these factors may also

impact on the transition time to full sucking feeds.


D Introduction of sucking feeds should not be delayed in infants who have mild, self-

limiting apnoeic episodes.

D Sucking feeds should not usually be introduced in infants who have apnoeic

episodes requiring intervention.

Infants with apnoea may take longer to establish full sucking feeds. Take special

care when introducing sucking feeds in these infants.

2.2.6 Work of breathing

No research was identified which investigated the most appropriate timing for introduction of

sucking feeds for infants with increased work of breathing – including accessory muscle use,

intercostal recession or retraction. The GDG noted that infants with increased work of

breathing may have difficulty maintaining oxygenation and physiologic stability while feeding.


D Introduction of sucking feeds should not be delayed in infants who have mildly

increased work of breathing, if they meet all other criteria.

D If an infant has moderately or severely increased work of breathing, consult medical

staff before introducing sucking feeds.

D If an infant has moderately or severely increased work of breathing, the decision to

introduce sucking feeds should be made with consideration of other factors including

their weight, gestational age at birth, number of days since birth, respiratory rate,

etc.

2.2.7 Continuous positive airway pressure

No research was identified which investigated the possibility of sucking feeds for infants who

were on continuous positive airway pressure (CPAP). There has been a concern that if infants

were to attempt sucking feeds while on CPAP, the open airway maintained by CPAP might

increase the risk of aspiration of milk.

However, under some circumstances, infants are removed from CPAP for regular brief periods.

It may be appropriate for sucking feeds to be introduced during these periods off CPAP or while

CPAP continues, if all other criteria are met.

Infants removed from CPAP for sucking feeds may require oxygen and should have their

oxygen saturation, heart rate and respiratory rate monitored closely.


D Infants on CPAP may be introduced to sucking feeds either while CPAP continues or



in a transient period off CPAP.

D The decision to introduce sucking feeds to an infant on CPAP should be made by a

multidisciplinary team, in light of the mother‘s wishes.

When an infant is on CPAP, provision of sucking feeds should be reassessed before

each feed in light of the infant‘s clinical state.

2.2.8 Tolerance of gavage feeds

Some protocols suggest that introduction of sucking feeds should be delayed until infants can

tolerate milk feeds provided via gavage. No research was identified which examined the effect

of this approach. It seems unnecessary to introduce gavage feeds in an infant who might

effectively establish sucking feeds without this step.


D Introduction of sucking feeds should not be delayed until infants can tolerate milk

feeds provided via gavage.


Trophic feeds 27

3. Trophic feeds (minimal enteral nutrition) Trophic feeds (also known as minimal enteral nutrition) are small volume, non-nutritional

enteral feeds given to enhance maturation of the gastrointestinal tract and prepare the infant

for the transition to enteral feeds. Trophic feeds can be given as small, constant volume feeds

for a certain period, or as small, advancing volume feeds, which eventually reach nutritional

volumes.

Our searching identified one Cochrane systematic review6 which evaluated the impact of

trophic feeds in parenterally fed infants. The review included articles published up to June

2004, and our searching did not identify any further relevant studies published after this date.

The review included 11 RCTs in infants at high risk of feeding intolerance or NEC,

10 of which compared trophic feeding to no enteral feeding and 1 of which

compared constant volume trophic feeds to advancing feeds. Most infants included

in the studies were less than 1500g, or less than 33 weeks gestation. The trophic

feeding groups received between 12 ml/kg/day and 24 ml/kg/day of breastmilk or

dilute or full strength formula given as either continuous or bolus feeds. Trophic

feeds were started as soon as the first day or as late as the eighth day after birth

and were continued at the same rate for 5-10 days. The control groups received

no enteral nutrition for between 6 and 18 days after birth, although some received

water by the enteral route.

In the meta-analysis comparing trophic feeding to no enteral feeding, infants given

trophic feeds had an overall reduction in mean days to full enteral feeding (n=617,

WMD=2.6 days), total days that feeds were withheld when nutritional enteral

feeds were introduced (n=129, WMD=3.1 days), and total hospital stay (n=370,

WMD=11.4 days). There was no significant effect on NEC (n=650, RR=1.16,

95%CI=0.75, 1.79).

On the basis of these findings and given the intermediate quality of the included

trials the review authors concluded that there was a ―plausible rationale‖ for

trophic feeding and that the review ―suggested benefits‖ however they urged

cautious interpretation as the results do not rule out an increase in NEC. The

variety of rates and methods of trophic feeding employed and varying quality of

the included trials mean it is not possible to determine the most effective trophic

feeding regime.

The RCT comparing constant volume trophic feeds to advancing feeds included

144 infants and compared trophic feeds of breastmilk or formula at 20ml/kg/day

for ten days, to advancing feeds beginning at 20ml/kg/day and increasing by

20ml/kg/day over a ten day period to a maximum of 140ml/kg/day. In the

advancing feeds group parenteral feeds were tapered as enteral feeds increased.

In this trial, infants given constant volume trophic feeds took longer to reach full

enteral feeds (WMD=13.4 days, 95%CI 8.2, 18.6) than infants given advancing

feeds.

This RCT was stopped early because of concerns about NEC. Incidence of NEC was

lower in the constant volume feeds group (1/74, 1.4%) compared to the

advancing feeds group (7/70, 10%) however the difference was not statistically

significant (RR=0.14, 95%CI 0.02, 1.07). In addition, infants given constant

volume trophic feeds had longer hospital stays although this difference was also

non-significant (WMD=11.0 days, 95%CI -1.40, 23.40).

This suggests that while advancing feeds have a positive effect on time to full

feeding and a potentially positive effect on hospital stay, they may also have a

1+


Trophic feeds 28

potentially detrimental effect on rates of NEC. The benefits of advancing trophic

feeds must therefore be weighed up against the potential risks.

The GDG noted that trophic feeds are part of a spectrum of feeding approaches used in

Monash Newborn. While some infants may have small volume non-nutritional feeds for a

significant period of time, for others trophic feeds are a method of gradual introduction of

enteral feeds, which soon become small volume nutritional feeds and are then graded up to full

enteral feeds.

In accordance with the evidence that, compared to no enteral feeds, trophic feeds reduce the

time to full enteral feeds, reduce the number of days feeds are withheld and reduce hospital

stay; and in line with the consensus of the GDG in section 2.1.1 that enteral feeding (either

trophic or nutritional) should be commenced in the first 48 hours of life unless precluded by

other criteria, the consensus of the GDG was to recommend the following:

B Consider trophic feeding in infants who meet the criteria for readiness for enteral

feeds but for whom there are concerns about capacity to tolerate normal feed

volumes, i.e.:

≤25 weeks gestation at birth

birth weight <750 g

absent or reversed end-diastolic flow

intra-uterine growth restriction

significant respiratory distress.

As mentioned above, there is little evidence available to determine the most effective method

of delivering trophic feeds.

In the absence of evidence, and in line with current practice, the GDG agreed that:

Trophic feeds should usually be of breastmilk.

D Trophic feeds should usually be introduced at a rate of 1ml every 4 hours and may

be continued at this rate for several days.

D Clinicians should take special care to monitor for feed intolerance and signs of

necrotising enterocolitis in infants receiving trophic feeds.

D When an infant is ready for increased volumes, trophic feeds may be increased at 24

hourly intervals, first to 1ml every 2 hours and then 1ml every hour if infants have

no signs of feed intolerance.

D If an infant has signs of significant feed intolerance after a trophic feed, the next

scheduled feed should be withheld.

D If significant signs of feed intolerance persist, all enteral feeds should be withheld

and senior medical review sought.

D If trophic feeds are withheld for less than 24 hours, they should be reintroduced at

the last rate given prior to signs of feed intolerance arising.


Trophic feeds 29

D If trophic feeds are withheld for more than 24 hours, they should be reintroduced at

an initial rate of 1ml every 4 hours, and increased as per usual.

Refer to Table 3 in section 4.10.5 for classification of feed intolerance.


Tube feed methods 30

4. Tube feed methods

4.1 Transpyloric or gastric tube feeds

Preterm infants may have enteral feeding tubes placed in the stomach (gastric tubes) or in the

upper small bowel (transpyloric tubes). There are potential advantages and disadvantages of

both methods.

Transpyloric tubes deliver enteral feeds directly to the main sites of nutrient absorption and

may also reduce the risk of reflux and aspiration of milk into the lungs. However, by definition,

transpyloric feeding avoids the gastric phase of digestion which might reduce nutrient

availability and impact on secretion of intestinal hormones and growth factors. This route

might also lead to increased risk of NEC by delivering potentially pathogenic organisms, which

would have been removed in the acidic environment of the stomach, to the upper small bowel.

Transpyloric feeding tubes are difficult to position and their position must be confirmed by

imaging. The transpyloric tube may subsequently migrate back to the stomach and serious

adverse events such as intestinal perforation and pyloric stenosis may also occur.

Our searching identified a Cochrane systematic review,47 and a subsequent update of this

review published by the same authors as a journal article, comparing the effectiveness of

enteral feeds delivered by gastric or transpyloric tubes. Both reviews included the same eight

RCTs, however more detail was included in the initial Cochrane systematic review,47 so the

results of both reviews are included here. No RCTs published after the review search date

(2003) were identified by our searching.

The systematic reviews included 8 RCTs, and a total of 340 infants. The authors

note that the studies were generally small, few concealed allocation, some were

quasi-randomised, and many did not achieve complete follow-up of all infants.

The review found that there were no statistically significant differences between

gastric and transpyloric feeding in weight gain prior to discharge (3 RCTs), rate of

increase of crown-heel length (5 RCTs), rate of increase of occipito-frontal head

circumference (5 RCTs), rate of gain of subscapular skinfold thickness (1 RCT),

time to full enteral feeds (2 RCTs).

Mortality was significantly higher in the group receiving transpyloric feeds (RR=2.5

95%CI 1.4, 4.5) than in the group receiving gastric feeds in the meta-analysis of

results from 6 studies reporting this outcome. After removal of one study from the

meta-analysis because of concerns about differences between the two groups in

baseline measures, the effect was not quite statistically significant (RR=2.2,

95%CI 0.9, 5.4).

In the meta-analysis of 7 trials reporting gastrointestinal disturbance (such as

diarrhoea or feeding intolerance), incidence was higher in the group receiving

transpyloric feeds (RR=1.5, 95%CI 1.1, 2.1). After removal of one study from the

meta-analysis because of concerns about differences between the 2 groups in

baseline measures, the effect was still statistically significant (RR=1.4, 95%CI 1.1,

2.1).

None of the 7 trials reporting NEC, nor the meta-analysis, found a significant

difference in risk of NEC between infants fed via gastric tube and those fed via

transpyloric tube.

1+

In light of the evidence for lack of benefit of transpyloric feeding, and increased rates of

adverse effects, including mortality, when compared to gastric feeding, the GDG agreed to the

recommendation:



A In most situations, gastric tubes should be used in preference to transpyloric tubes.

Transpyloric tubes may be appropriate for some infants such as those with severe

gastro-oesophageal reflux and chronic lung disease.

The management of infants being fed via transpyloric tubes is beyond the scope of this

guideline.

4.2 Nasogastric or orogastric tube feeds

Gastric feeding tubes may be placed via either the nose or mouth. There are potential

advantages and disadvantages of both methods. Nasal placement of feeding tubes may

negatively affect respiration by blocking a nostril, and orally placed tubes may be more prone

to displacement and local irritation.

A Cochrane systematic review45 was found which sought to compare the effectiveness of nasal

and oral placement of feeding tubes in preterm or low birth weight (LBW) infants. The review

included only 1 RCT including 44 infants, published in 1984. This trial is not included in this

discussion as orthodontic palatal appliances were used to secure the oral tubes, making the

results unlikely to be applicable to local practice.

One further RCT published after the systematic review search date was

identified.46 In two of the arms of this RCT, infants were randomised to receive

either intermittent nasogastric (n=22) or intermittent orogastric feeds (n=24). In

the third arm infants received continuous nasogastric feeds.

The study found that the difference in time to full enteral feeds in the intermittent

nasogastric (26.1 days) and intermittent orogastric (28.8 days) groups was not

statistically significant. However, the study only included enough infants to detect

a 40% difference between groups in time to full enteral feeds and smaller

differences may well be clinically important. No significant differences were found

in total energy intake, total protein intake or growth rate of the lower leg.

The orogastric group had non-significantly increased rates of NEC (1/24 vs 0/22),

septicaemia (83.3 vs 72.7%), antibiotic treatment (27 vs 24.5 days), episodes of

vomiting (median of 4.5 vs 2), episodes of gastric residuals >50% of previous feed

(median of 6 vs 3) and patent ductus arteriosus (58.3 vs 50%).

The nasogastric group had non-significantly increased rates of respiratory distress

syndrome (77.3 vs 70.8%), mechanical ventilation (81.8 vs 62.5%) and need for

supplemental oxygen (44.8 vs 32.4%).

The study did not include enough infants to reliably detect clinically important

differences between the groups for these secondary outcomes.

1-

This evidence is of limited use in deciding whether to use nasogastric or orogastric feeding

tubes.

In summary , the one relevant study did not detect a statistically significant difference

between nasogastric and orogastric tubes in time to introduce enteral feeds; however the

study‘s small sample size limits its ability to detect differences between groups of less than

40% which may still be clinically important. The study suggested that orogastric and



nasogastric tubes may be associated with different patterns of adverse effects, but again did

not include enough infants to reliably detect clinically important differences.

In light of the limited evidence and in line with current local practice to use nasogastric tubes

in most infants, the GDG agreed that:

D Nasogastric tubes should be used in preference to orogastric tubes in most infants.

Orogastric tubes may be considered when the infant is receiving CPAP, or there is

significant concern about the possibility of compromising respiration. Orogastric

tubes should also be used when infants are having gavage tubes placed

intermittently.

4.3 Intermittent or continuous feeds

Premature infants can be given milk feeds by gavage tube either continuously or

intermittently, usually over a 10-20 minute period every two or three hours. Both methods

have potential benefits and risks.

While gavage feeds can be given via either nasogastric or orogastric tube (as discussed above)

all of the identified studies compared intermittent or continuous feeds in infants fed by

nasogastric tube.

A Cochrane systematic review49 was found which compared the effectiveness of continuous

nasogastric milk feeding to intermittent nasogastric milk feeding in premature infants with

birth weight <1500g. We identified one further RCT46 published after the most recent search

conducted by the Cochrane review authors.

The Cochrane systematic review included 6 studies and a total of 443 infants with

birth weight <1500g, no prior history of feeding and no congenital anomalies

which might interfere with establishing enteral feeds.

Four trials (n=356) assessed the number of days to full feeds. Two of these trials

found that it took significantly longer for infants fed by the continuous method to

reach full feeds, the other 2 trials found a non-statistically significant difference in

the same direction. The meta-analysis found that it took infants significantly

longer to reach full enteral feeds when fed by the continuous method (WMD=3.0

days, 95%CI 0.7, 5.2).

Four trials reported weight gain. One trial found that infants fed by the continuous

method gained weight more slowly than infants fed by the intermittent method.

The other 3 trials reported conflicting results, none of which were statistically

significant and the meta-analysis did not find a difference in weight gain between

the two groups.

The effect of intermittent or continuous feeding on feed intolerance is not clear.

Four trials assessed feeding intolerance. No differences were found between

groups in any of the trials. A meta-analysis could not be performed because

different measures of feeding intolerance were used. Two trials reported the

number of infants who failed to complete protocol (for reasons which included, but

were not exclusively related to feed intolerance). One trial found that infants fed

by the continuous method were significantly less likely to be withdrawn from the

assigned feeding protocol than infants fed intermittently. The other trial and the

meta-analysis did not find a difference between the groups. Three trials reported a

greater incidence of gastric residuals in infants fed by continuous tube feeding,

however varying methods were used to classify ‗significant‘ residuals (greater than

1+



a certain proportion of the previous feed, or greater than a certain volume, at

differing time points). The clinical importance of these results is therefore unclear.

Neither the trials nor the meta-analysis found any evidence of a difference

between feeding methods in: time to regain birth weight, length gain, head

circumference gain, change in triceps skinfold thickness, days to discharge, days

to discharge weight of 2040g, days on mechanical ventilation, proven or probable

NEC or apnoeas.

The RCT46 published after the systematic review included 44 infants with

gestational age between 24 and 29 weeks and birth weight <1200g. In contrast to

the 4 RCTs included in the Cochrane review, this fifth RCT found that continuously

nasogastric fed infants reached full feeds more quickly than intermittently

nasogastric fed infants. The result, while not quite statistically significant by usual

(arbitrary) thresholds, is potentially still clinically important (20.1 vs 26.1 days,

p=0.055). The study also found that the rate of growth of the lower leg was higher

in the continuously fed infants. Other outcomes did not differ significantly between

the groups. There were 2 cases of NEC in the continuously fed group and none in

the intermittently fed group, this difference was not statistically significant.

Addition of the more recent RCT to the Cochrane meta-analysis of time to

establish full enteral feeds meant the result of the meta-analysis was no longer

statistically significantly in favour of intermittent feeds (WMD=2.2 days; 95%CI -

1.1, 5.6), and that there was also a substantial amount of unexplained variation in

the data, with this newer RCT producing results that are in conflict with the

previous studies.

The reason for this variation is unclear. Infants included in the more recent RCT were not

substantially different from infants included in the earlier studies, either in terms of gestation

or birth weight, although they tended to be towards the smaller and younger end of the

spectrum.

The feeding protocols varied between all studies, however in the earlier studies it appears that

infants were maintained on the method to which they were randomised until they reached full

enteral feeds, whereas in the more recent RCT infants were transitioned to intermittent feeds

at 32 weeks of age, regardless of the feeding method to which they were randomised.

It may be that continuous, or more frequent feeds are more suitable for younger infants and

intermittent, or less frequent feeds are more suitable for older infants, however there is not

enough evidence available to test this theory.

The GDG noted that all of the evidence identified related to infants weighing less than 1500g,

and that there was no evidence to guide decision making in infants weighing more than 1500g.

In the absence of clear evidence and in line with current local practice the GDG agreed that:

D Gavage feeds should usually be given as intermittent boluses.

4.4 Indwelling or intermittently placed tubes

In infants receiving gavage feeds, tubes can be indwelling or placed intermittently (placed

before each feed, or for a certain period of time and replaced regularly). Both methods have

potential benefits and risks.



Indwelling tubes may reduce the risk of misplacement inherent when tubes are frequently

removed and replaced, however the ongoing presence of a tube through the oesophageal

sphincter, which is therefore not completely closed, might increase the risk of reflux and/or

aspiration. The continuous presence of a tube in the nasal passage also increases airway

resistance, potentially leading to increased energy expenditure.

One RCT was identified which investigated the impact of indwelling or intermittently placed

feeding tubes in premature neonates.

The study included 93 neonates with gestational age of 24-34 weeks, who were

‗medically stable‘, on full enteral tube feeds, did not require ventilation and had no

major medical problems. Forty-nine infants were randomised to indwelling

nasogastric feeding tubes and 44 to intermittently placed orogastric tubes. Nine

infants (1 in the indwelling and 8 in the intermittent group) did not complete the

six-day study as they were transferred to other units or became ‗medically

unstable‘. Medical stability or instability was not further defined.

No statistically significant differences were found between the indwelling and

intermittent groups in weight gain (127.5 vs 116.3g), or proportion of infants with

apnoeic (56 vs 43%) or bradycardic (60 vs 45%) episodes. However these results

are very difficult to interpret because the study did not include enough infants to

demonstrate a statistically significant difference between the groups on the

primary outcome, weight gain, or either of the secondary outcomes. Prior to

commencing the study it was estimated that 80 infants would be required in each

group to demonstrate statistical significance at 20g weight difference between

groups.

1-

In the absence of evidence and in line with current local practice the GDG agreed that:

D Feeding tubes should usually be indwelling. Only consider intermittent placement of

feeding tubes if the infant is receiving ≤2 feeds per day by gavage tube

4.5 Rate of introduction and advancement of milk feeds

The rate at which enteral feeds should be increased in infants being introduced to milk feeds

has not been firmly established. Like many areas in establishment of feeding in neonates,

there is a need to balance the concerns about the potential of rapid introduction of feeds to

increase NEC with quickly establishing independent feeding so that infants‘ hospital stays are

not prolonged.

Our searching identified a Cochrane systematic review63 investigating rapid versus slow rate of

advancement of feeds, and one further RCT64 published after the systematic review search

date.

The Cochrane review63 included 372 premature or LBW infants from 3 RCTs at high

risk of NEC and feeding problems, receiving supplemental parenteral nutrition. The

review aimed to examine the impact of different rates of advancement of feeds on

NEC, length of stay (LOS), days to full feeding, days feeds withheld and days to

regain birth weight.

The criteria for including studies in the review and the search terms were not well

described, making it difficult to assess how comprehensive the review was. The

studies included in the review varied in their definition of slow advancement (from

10 to 20ml/kg/day) and rapid advancement of feeds (20 to 35ml/kg/day). Two of

the included RCTs used formula only, and the third used either formula or

1+



breastmilk.

The review found that a rapid rate of feeding advancement was associated with a

significant reduction in mean days to full enteral feeding (WMD=

-3.2 days, 95%CI -4.87, -1.62) and in days to regain birth weight (WMD=-2.1

days, 95%CI -3.26, -0.96). No significant effect was found on NEC or LOS.

The additional RCT64 included 53 infants with birth weight <1250g, no gross

congenital abnormalities, no abdominal distension, no recurrent apnoea, who were

not requiring pressor support or oxygen by hood. All infants were fed expressed

breastmilk. Infants in the slow advancement group received 2 hourly bolus

nasogastric feeds started at 15ml/kg/day and advanced by 15ml/kg/day until the

volume reached 180ml/kg/day. Infants in the rapid advancement group received 2

hourly bolus nasogastric feeds started at 15ml/kg/day and advanced by

30ml/kg/day until the volume reached 180ml/kg/day.

Twenty-six percent of infants in the rapid and 46% of infants in the slow groups

died before completing the study protocol. These infants had significantly lower

birthweights than those who completed the study.

The RCT found that time to full enteral feeds was reduced in the rapid feeding

group (10 vs 15 days, p<0.001), as was time to regain birth weight (18 vs 23

days, p value not provided). Seven infants in the rapid feeding group died, causes

of death were sepsis (3), NEC and sepsis (2), RDS (1) and prematurity (1). Twelve

infants in the slow feeding group died, causes of death were sepsis (10), inborn

error of metabolism (1) and pulmonary haemorrhage (1).

The GDG noted that preterm infants often have considerably smaller energy stores than term

infants, and therefore preterm infants may require a higher initial rate of milk feeds. In light of

this and the evidence that in infants being introduced to enteral feeds, rapid advancement of

feeds reduces days to full enteral feeding and days to regain birth weight and does not

increase the rate of NEC, the GDG agreed that:

B In infants being transitioned from trophic feeds to nutritional enteral feeds, feeds

may be increased once or twice per day in steps of 1ml/hr.

B In healthy infants (weight <1000g) being introduced to nutritional enteral feeds,

feeds should usually be commenced at an initial rate of 1ml/kg/hr.

B In healthy infants (weight ≥1000g) being introduced to nutritional enteral feeds,

feeds should usually be commenced at an initial rate of 2ml/kg/hr.

B In preterm infants with hypoglycaemia, a higher initial rate may be considered in

consultation with a senior clinician.

B Nutritional enteral feeds should usually be increased by 1ml/kg/hr every 24 hours,

unless signs of feed intolerance or NEC are observed.



4.6 Frequency of feeds

There is very little evidence available to inform decisions about the most appropriate frequency

at which enteral tube feeds should be provided. There is some concern that feed frequency

might be related with feed intolerance, however there is no evidence to evaluate this.

One small study68 was identified which investigated the impact of varying the

frequency of nasogastric milk feeds in preterm babies. The study included 15

preterm babies who ―survived the early neonatal period‖ and whose mothers

declined to breastfeed. All infants received formula milk.

Infants served as their own controls and were randomised to differing order of

continuous, 1, 2 or 3 hourly feeds, each for a 3 day period. After each three day

period weight, weight gain, transit time, milk intake and fat absorption were

assessed.

The study provides only a very limited description of the methods used. The

authors do not adequately describe randomisation, allocation concealment,

baseline characteristics of the infants, blinding or analysis methods. No reason is

given for choosing a comparison period of 3 days.

No differences were identified between the feeding methods for any outcome

measure.

1-

In some nurseries feeds are provided at 1 hourly, 2 hourly, 3 hourly, 4 hourly, or, less

commonly, 5 or 6 hourly intervals. The decision as to which is used is often based on

gestational age or weight, with frequency decreasing with increasing age and weight, however

there is no evidence available to evaluate the effectiveness of these methods.

In light of the little evidence available and in line with current local practice, the GDG agreed

that:

D In infants receiving gavage feeds, frequency of feeds should usually be:

1 hourly for infants weighing <1000g

2 hourly for infants weighing 1000-1500g

3 hourly for infants weighing 1501-2000g

4 hourly for infants weighing >2000g.

4.7 Maximum tube feed volume per day

The maximum feed volume per day to be provided to infants can be defined by establishing

the goal weight gain to be achieved per day and increasing the volume fed until an infant

achieves the goal daily weight gain, or by setting a certain maximum volume to be fed per kg

of the infant‘s body weight. No evidence was identified which evaluated the effectiveness of

defining the maximum volume of milk tube fed to infants on the basis of a certain goal weight

gain as compared to using a certain volume per kg body weight.

The ideal maximum volume of milk fed per day will depend on the caloric density of the milk,

which varies between formulas and breastmilk, and the energy required by the infant for

appropriate maintenance of bodily functions, physiological and neurological development and

growth.

Linking maximum feed volume to rate of weight gain requires that the ideal rate of weight gain

can be established. This is commonly estimated by referring to charts which plot birth weight

against gestational age, and thus give an indication of the expected rate of weight gain in

utero. Several different charts of this type are available, and while they differ to some degree,

most show a similar pattern, with average weight gain of approximately 10-15 g/day from



gestational ages of 22 to 27 weeks, accelerating to 25-30 g/day from 28-40 weeks, and then

levelling off.65-66

These rates may be affected by a range of factors including gender, ethnicity, intrauterine

growth restriction, physiological stability, illness after birth and whether the infant is the

firstborn child.

One RCT67 was identified which examined the impact of differing maximum feed volumes.

The study reported data from 54 infants with gestational age (GA) <30 weeks at

birth, on full enteral feeds, with no congenital abnormalities likely to interfere with

feeding. Infants received enteral feeds of breastmilk or formula, were initially tube

fed, and over the course of the study made the transition to sucking feeds. Infants

were randomised to receive a maximum of 150ml/kg/day (the 150 group) or

200ml/kg/day (the 200 group).

Infants weighing less than 1500g at birth who received expressed breastmilk had

breastmilk fortifier (Enfamil) whereas those who did not receive breastmilk were

given preterm formula. Enfalac 24 was used during the first half of the study

period, after which PreNan was used. Breastmilk fortification or preterm formula

was ceased at the discretion of the attending specialist once the infant reached a

weight of approximately 1800–2000g.

Feeding volumes in the 150 group were increased at the physician‘s discretion if

minimum weight gain of 8g/kg/day was not achieved and reduced in the 200

group if infants developed feed intolerance or clinically significant fluid retention.

Fully breast-fed infants were fed on demand and fully bottle-fed infants received

their prescribed volume.

Infants in the 200 group were on average slightly heavier (1020 vs 895g) and

older (27.5 vs 26 weeks GA) at baseline. Five infants of the 59 originally

randomised were withdrawn from the study. Two developed NEC (both in 150

group) and 3 had surgical problems requiring transfer to another hospital (2 in the

150 group and 1 in the 200 group).

Forty-three percent of infants in the 150 group required feed volumes above

150ml/kg/day to maintain minimum weight gains and 54% of infants in the 200

group required permanent reduction in feed volumes below 200ml/kg/day to avoid

feed intolerance or fluid retention.

At 35 weeks infants in the 200 group were on average heavier than infants in the

150 group (2020 vs 1885g, p=0.014), however it is difficult to interpret this

difference in light of the apparent difference in birth weight between the groups at

baseline. At 35 weeks infants in the 200 group also had more arm fat area (282 vs

218mm2, p=0.009) and had a higher rate of daily weight gain (18.6 vs 16.5 g/kg,

p=0.047). Length, head circumference, arm area and arm muscle area showed

similar, but non-statistically significant trends.

Follow-up at 12 months corrected age did not demonstrate any statistically

significant differences between the groups, though infants in the 150 group

appeared to be more likely to be below the 10th percentile for weight and length

(33 vs 17%, p=0.21 and 25 vs 9%, p=0.10 respectively).

1+

This study suggests that there are potential benefits of increased maximum feed volumes. The

substantial variation in feed volume actually received by the infants in this study highlights the

importance of adjusting the maximum feed volume to meet the individual needs of each infant.



In light of the limited evidence and in line with current local practice, the GDG agreed that:

B In infants receiving tube feeds the maximum feed volume (fortified as appropriate)

should be adjusted within the range of 6 to 8ml/kg/hr to a volume that supports an

average weight gain of 10g/kg/day, while avoiding feed intolerance.

The GDG noted that it was not necessary that infants be weighed every day, and that in unwell

infants overly frequent weighing can be detrimental.

In line with current practice, the GDG agreed that:

D Infants who are physiologically stable and receiving tube feeds should usually be

weighed every 2 days.

D In infants who are physiologically unstable and receiving tube feeds, frequency of

weighing should be decided in consultation with a senior clinician.

4.8 Size of feeding tubes

No evidence was identified to inform decisions about the most appropriate size of feeding

tubes to be used in infants receiving gavage feeds.

In the absence of evidence and in line with current local practice, the GDG agreed that:

D In infants receiving gavage feeds use:

size 8 French Gauge tubes for intermittent feeds

size 6 French Gauge tubes for indwelling tubes in infants ≥1500g

size 5 French Gauge tubes for indwelling tubes in infants <1500g.

Consider using a size 8 French Gauge tube for indwelling tubes in infants >3000g.

Consider using a smaller tube if an infant displays signs of not tolerating a larger

tube.

4.9 Positioning for tube feeds

It has been suggested that the position in which an infant is placed during tube feeds may

influence the success of the feed by impacting on respiratory stability, development of gastro-

oesophageal reflux or other factors.

A Cochrane systematic review69 was identified which investigated the impact of positioning in

hospitalised infants and children with acute respiratory distress, however the role of position

during feeding was not examined, and the review was therefore excluded.

Two randomised cross-over trials70-71 and two controlled trials72,75 were found which

investigated the impact of body position during feeds on a range of outcome measures. A third

controlled trial was identified89 but excluded as it reported the same comparisons and

outcomes as the randomised cross-over trials.

Two very small randomised cross-over trials were found that investigated the

impact of body position during feeds of breastmilk in ventilated VLBW infants with 1-



―radiographic and clinical evidence of chronic lung disease‖.70,71 Both of these

studies used a cross-over design in which infants randomly received a feed in

either prone or supine position, and their next feed was then in the other position.

The studies included 7 and 6 infants respectively, did not describe the method of

randomisation and were not blinded. Infants received nasogastric feeds over 1

hour, at 3 hourly intervals. At entry to the study all infants were receiving milk

feeds of more than 100ml/kg/day.

The first study70 examined the impact of body position during the feeding period.

Infants had a mean postnatal age of 56.4 days (range 28-107 days) and weight of

1018g (range 806-1368g) when the study was performed. The study found that

oxygen saturation was statistically significantly higher in infants placed in the

prone position than in those who were supine at all time points (before feeding:

97.4 vs 93.1%, 20 minutes into feed: 97.1 vs 92.4%, 40 minutes into feed: 96.7

vs 90.3%, immediately post-feeding: 94.1 vs 88.1%, all p<0.05). The study also

found that transcutaneous carbon dioxide tension (TcCO2) was statistically

significantly lower in the prone position than supine before feeding and 20 minutes

into the feed (45.5 vs 49.9mmHg, 46.6 vs 50.9mmHg respectively, both p<0.05)

and that work of breathing for spontaneous breaths was significantly less in the

prone position compared with the supine position at the 40 minute interval of

feeding and immediately post-feeding (only graphical data provided, p<0.03). The

study further reported that pulmonary resistance was not different in either

position and all values returned to the pre-feeding level by 30 minutes prior to the

next feed regardless of the body position.

In the second cross-over study71, impact of body position on outcomes before and

after feeds was assessed. Infants were placed in the randomly allocated position 1

hour before a feed, were fed in that position and then remained in that position

after the feed until all values returned to baseline levels. Infants had a mean

postnatal age of 47.5 days (range 21-85 days) and weight of 956g (range 602-

1262g) when the study was performed.

The study found that oxygen saturation was statistically significantly higher in

infants in the prone position than supine both pre- and post-feeding (pre-feed:

98.2 vs 95.8%, post-feed: 98.2 vs 94.8%, both p<0.01) and a similar non-

significant trend was seen in TcCO2 levels. The study also found that tidal volume

was greater in the prone position pre-feeding (7.1 vs 5.2ml/kg, p<0.05) but there

was no difference post-feeding, and that heart rate (HR) and respiratory rate were

lower in infants in the prone position both pre- and post-feeding (HR pre-feed:

145.8 vs 153.3 beats/min, HR post-feed: 141.7 vs 155.7 beats/min, respiratory

rate pre-feed: 25.3 vs 31.5 breaths/min, respiratory rate post-feed: 21.2 vs 31.7

breaths/min, all p<0.05).

In the first controlled trial, reported in a Doctor of Philosophy thesis,72 24 preterm

infants in NICU who were in room air and not yet on sucking feeds received 1

nasogastric feed of breastmilk or formula in a prone position and then the 2

subsequent feeds while held upright in chest-to-chest position.

No randomisation was undertaken, little description is provided of inclusion and

exclusion criteria and infants served as their own controls. Infants were evaluated

at 2 minute intervals during the feeds, and then 10 minutes after the conclusion of

the feed.

The study found no differences between the positions in oxygen saturation,

episodes of apnoea or bradycardia, heart rate, or respiratory rate. However in the

held position infants had significantly higher skin temperatures compared to when

2+



they were prone (36.3°C vs 35.8°C, p<0.05), and spent a greater proportion of

time in quiet irregular sleep during gavage feeding (66 vs 42%, p<0.05). Held

infants also spent significantly smaller proportions of time in the active sleep and

active awake states (4 vs 19%, 1 vs 10%, both p<0.05). Proportions of time spent

in other behavioural states was similar between held and prone positions.

The second controlled trial75 included 18 healthy LBW infants requiring tube feeds.

Infants received orogastric breastmilk or formula feeding in prone, supine, left and

right lateral positions. Each infant was fed in each of the 4 positions, with the

order ―varied so as to obtain a balanced distribution‖. The study used dyed milk

feed and a nasogastric tube containing white cotton to measure regurgitation and

reflux height (in 2.5 cm intervals from <2.5 to >10cm), up to an hour after each

feed. There is no description provided of how the placement of the nasogastric

tube into the stomach was confirmed. This is particularly important because the

method of measurement of reflux height assumes that all tubes were placed

identically in the infants‘ stomachs.

The study found that the incidence of reflux was similar in each position (range

13-15) however compared to supine infants, infants in the prone position were

much less likely to reflux to higher heights. For example there was a significantly

lower number of feeds refluxed to a height of 7.5cm or more in the prone group

compared to the supine group (4/18 vs 12/18, p=0.007). There were no

significant differences in comparisons of other positions.

The GDG highlighted that, particularly in small, unwell infants, unnecessary handling should be

minimised so as not to cause stress to the infant. The limited evidence for benefit of prone

positioning during feeds needs to be balanced against the potential for frequent repositioning

to cause harm.

In light of the evidence available which only compares a small range of positions on a limited

number of outcomes in restricted populations, and in line with current local practice, the GDG

agreed that:

B Physiologically unstable infants and infants receiving trophic feeds should be fed in

the position which minimises unnecessary repositioning.

B Physiologically stable infants should preferably be fed either in a prone position or

held upright in a chest-to-chest position, however other positions may be used at

the clinician‘s discretion, avoiding unnecessary handling.


Sucking feed methods 41

5. Sucking feed methods

5.1 Breast or bottle feeding

5.1.1 Benefits of breastfeeding

Breastfeeding has a range of established benefits for both mothers and infants. These benefits

include effects on infant and maternal health as well as mother-infant bonding. A review of all

of these is beyond the scope of this guideline, however it is worth noting that most of the

research investigating the benefits of breastfeeding has been undertaken in term infants.

National and international health organisations advocate breastfeeding as the best method of

infant feeding. The National Health and Medical Research Council (NHRMC) Dietary Guidelines

for Children and Adolescents in Australia87 state that ―Breastfeeding is the normal and most

appropriate method for feeding infants and is closely related to immediate and long-term

health outcomes‖. Box 2 below lists the benefits of breastfeeding highlighted in the NHMRC

guideline.

Box 2 Health advantages of breastfeeding87

Child Mother

reduced incidence and duration of

diarrhoeal illnesses

protection against respiratory

infection and reduced prevalence of asthma

reduced occurrence of otitis media and recurrent otitis media

possible protection against neonatal

necrotising enterocolitis,

bacteraemia, meningitis, botulism and urinary tract infection

possible reduced risk of auto-

immune disease, such as type 1

diabetes and inflammatory bowel

disease

reduced risk of developing cows‘ milk allergy

possible reduced risk of adiposity later in childhood

improved visual acuity and psychomotor development

higher IQ scores

reduced malocclusion as a result of better jaw shape and development.

promotion of maternal recovery from

childbirth—accelerated uterine

involution and reduced risk of

haemorrhaging (thus reducing

maternal mortality) and preservation

of maternal haemoglobin stores

through reduced blood loss, leading to improved iron status

prolonged period of post-partum

infertility, leading to increased spacing between pregnancies

possible accelerated weight loss and

return to pre-pregnancy body weight

reduced risk of pre-menopausal breast cancer

possible reduced risk of ovarian cancer

possible improved bone

mineralisation and thereby

decreased risk of post-menopausal hip fracture.

The World Health Organization (WHO) states ―Breastfeeding is an unequalled way of providing

ideal food for the healthy growth and development of infants; it is also an integral part of the

reproductive process with important implications for the health of mothers.‖86

In the International Code of Marketing of Breastmilk Substitutes, to which Australia is a

signatory, the WHO states that ―breastfeeding … forms a unique biological and emotional basis

for the health of both mother and child‖.88

Both the activity of breastfeeding and the breastmilk itself contribute to the benefits of

breastfeeding described above.



No studies were identified that compared the effect of breastfeeding with bottle feeds of

breastmilk on long-term outcomes in preterm infants. Five controlled studies were identified

that compared the effect of breast and bottle feeds of breastmilk on short-term outcomes such

as oxygen saturation and body temperature.91-96 Three of these studies compare breastfeeding

to bottle feeds of breastmilk only and two of the studies compare to bottle feeds of breastmilk

and formula. One study was reported in abstract form only and an unsuccessful attempt was

made to contact the author for more information.91 Two of the included papers are from the

same research group.95,96 It seems likely from details of the populations and methodologies

that these two papers report results from studies including different infants, but this could not

be firmly established.

The first controlled trial92 included 25 preterm infants with a birth weight <1800g

whose mothers were willing to express breastmilk. The study compared the

physiological stability of the infants when breastfeeding (a total of 80 sessions)

with that when given bottle feeds of breastmilk (a total of 80 sessions). Infants

served as their own controls and breast and bottle feeds were paired; 54/80 on

same day, 26/80 within 3 days.

The infants‘ mean GA at birth was 29.2 weeks, (range 27.7-33.0) and mean birth

weight was 1199g (range 760-1759). At commencement of the study their mean

chronological age was 9.3 weeks (range 1.1-25.3) and mean weight was 1703g

(range 1120-2080). The study found that body temperature and oxygen saturation

were higher during breastfeeds (data provided graphically only). There were 2

episodes of apnoea and 20 episodes of desaturation (oxygen saturations <90%)

during bottle feeding and none during breastfeeding.

The second controlled trial93 included 8 preterm infants with weight <1500g at first

oral feed, weight appropriate for gestational age (AGA), whose mothers decided to

breastfeed. The study compared the physiological stability of the infants when

breastfeeding (a total of 24 sessions) with that during bottle feeds of breastmilk (a

total of 16 sessions). Infants served as their own controls. One infant ceased

breastfeeding, and their data was not included in quantitative analysis. Eleven

feeding sessions (10 breast and 1 bottle) were excluded due to no milk intake or

equipment malfunction, and 13 within-infant comparisons were included.

The infants‘ mean GA at birth was 30.2 weeks (range 25-33) and mean birth

weight was 1370g (range 813-1950). At commencement of the study mean GA

was 34 weeks (range 33-36) and mean weight was 1700g (range 1320-1880).

This is confusing given that weight <1500g at first oral feed was an inclusion

criterion.

Five infants had desaturation events (oxygen saturations <90%) during feeding.

Three of these infants had 1 desaturation during breastfeeding and 2

desaturations during bottle feeding. One infant had a desaturation event during

bottle feeding but no breastfeeding session recorded and the fifth infant had

events in all 4 bottle feeding sessions and 4 of the 5 breastfeeding sessions. The

authors state that ―preterm infants have greater stability of oxygenation during

breast-feeding when compared to bottle-feeding at the same gestational age‖

however the data does not seem to be strong enough to support this.

The third controlled trial94 included 5 preterm infants with weight <1500g at first

oral feed, weight AGA, whose mothers decided to breastfeed. The study compared

the physiological stability of the infants when breastfeeding (a total of 39 sessions)

with that during bottle feeds of breastmilk (a total of 32 sessions). Infants served

as their own controls. Breastfeeds were initiated within 2 days of bottle feeds.

2-



The infants‘ mean GA at birth was 30.5 weeks, (standard deviation (SD) 1.2) and

mean birth weight was 1296g (SD 106). At first bottle feed mean age was 33.8

weeks (SD 1.1) and mean weight was 1328g (SD 70). The study found that the

transcutaneous oxygen pressure decrease between baseline and post-feeding was

non-significantly greater in infants receiving bottle feeds (data not provided,

p=0.145). Post-feeding, temperature increased in the breast-fed infants but

continued to decrease in the bottle-fed infants (data not provided, p<0.001). Mean

maximal temperature change was +0.20oC in bottle-fed and +0.49oC in breast-fed

infants (p<0.001).

The fourth controlled trial95 included 12 singleton infants with birth weight ≤800g,

whose mothers decided to breastfeed. The study compared the physiological

stability of the infants when breastfeeding (10 daily sessions) with that when

bottle-fed breastmilk or formula (31% of feeds)(10 daily sessions). Infants served

as their own controls. Five infants did not have data for all feeds sessions (4 were

discharged, 1 became ill).

The infants‘ mean GA at birth was 26 weeks (SD 2) and mean birth weight was

672g (SD 95). At onset of study mean age was 35 weeks (SD 2) and mean weight

was 1460g (SD 210). Mean weight gain was greater following bottle feeds (31 ±

6g) than breastfeeds (9 ± 6g), p<0.001. There were no differences in respiratory

rate or HR. Oxygen saturation and temperature were significantly higher during

breastfeeds than during bottle feeds (data presented graphically, differences of 1-

2%, p<0.05 and up to 0.5oC, p<0.001 respectively). Nine percent of oxygen

saturations recorded during breastfeeds were desaturated compared to 20% of

oxygen saturations recorded during bottle feeds.

The fifth controlled trial96 included 20 infants with birth weight <1500g, whose

mothers decided to breastfeed. The study compared the physiological stability of

the infants when breastfeeding (5 daily sessions) with that during bottle feeds of

breastmilk or formula (10% of feeds) (5 daily sessions). Infants served as their

own controls. Four infants initially included in the study were subsequently

excluded (3 mothers could not express adequate volumes of breastmilk and 1

could not meet with staff for social reasons). Eleven scheduled sessions were

missed.

The infants‘ mean GA at birth was 28 weeks (SD 3) and mean birth weight was

1065g (SD 255). At onset of study mean age was 35 weeks (SD 1), weight data

was not provided.

The study found that median weight gain was greater following bottle feeds than

breastfeeds (30 vs 0g, p<0.001). There was no difference in temperature change.

Oxygen saturations were higher during breastfeeds than bottle feeds but the

difference was marginally non-significant (differences of 1-3%, data presented

graphically, p=0.056). Desaturation events (oxygen saturation <90%) occurred in

21% of breastfeeds and 38% of bottle feeds (p<0.025).

In light of the evidence of improved physiological stability in breast-fed preterm infants, the

well established long- and short-term benefits of breastfeeding in full term infants and the

unanimous opinion of the GDG members based on clinical experience that breastfeeding has

substantial benefits for preterm infants, the GDG agreed that:

C Mothers should be encouraged and supported to breastfeed.

D The benefits of breast feeding should be explained to mothers, education material

promoting breast feeding should be prominently displayed and practical assistance



in establishing and maintaining lactation should be readily accessible

D A mother‘s informed decision to breast or bottle feed should be respected.

5.1.2 Consistency of practice and communication

There is substantial discussion in the literature about the importance of standardised practice

and consistent communication between clinicians and parents during the establishment of

sucking feeds. The GDG and the Monash Newborn Parent Group both emphasised the vital

importance of consistent clinical practice, communication, information and education about

establishing sucking feeds in Monash Newborn:

D Parent education materials explaining the process involved in establishment of

sucking feeds, the likely stages and outcomes and the role of parents and clinicians

in this process should be provided to parents, and opportunities given to discuss any

questions parents have.

D Clinician education materials explaining the process involved in establishment of

sucking feeds, the likely stages and outcomes and the role of parents and clinicians

in this process should be provided to clinicians working in Monash Newborn, and

opportunities provided to discuss any questions clinicians have.

5.1.2.1 Impact of a single caregiver

No research evidence was identified that examined the impact of multiple caregivers compared

with a single caregiver on the transition to sucking feeds for premature infants. The GDG noted

that while it was not possible that every infant receive care from only one or two clinicians,

decreasing the number of caregivers and standardising techniques used would lead to

decreased variation in care. The GDG also noted that for some infants, particularly those

having difficulty establishing feeds, it may be useful to encourage the mother to be the

primary caregiver for as many feeds as possible.

D In infants being introduced to sucking feeds, use as few caregivers as possible to

ensure consistency of practice.

5.2 Non-nutritive sucking

Non-nutritive sucking (NNS) is often used in preterm infants as part of the process of

establishing full sucking feeds. Non-nutritive sucking is thought to support establishment of a

calm, organised state in the infant and promote development of sucking behaviours, however

it has been suggested that non-nutritive sucking might have a negative impact on the

development of breastfeeding.

A Cochrane systematic review51 was identified which investigated the effect of non-nutritive

sucking in preterm infants. Two further relevant RCTs were identified, one of which was

published after the systematic review search date52 and another which was not identified by

the search strategy.53

The Cochrane systematic review51 included 21 studies of infants <37 weeks GA at

birth who were allocated to receive either NNS or standard care without NNS.

Infants in all studies were either tube or bottle-fed and the impact of NNS on

1-



breastfeeding was not measured in any of the studies. NNS was provided in a

variety of ways in the studies. In many cases NNS on a dummy was offered with,

before or after tube feeds, and in other cases NNS on a dummy was offered before

or after bottle feeds for periods of between 2 and 10 minutes. At least 1 RCT53

appears to have been missed by the search strategy which raises questions about

the rigour of the search, however our broader searching did not identify any further

missed RCTs. The review included 6 studies in which randomisation was not

undertaken or the method of allocation was not described.

The review found a significant reduction in length of stay (WMD=-7.1 days, 95%CI

-12.6, -1.7) in infants receiving NNS (2 RCTs, n=87), a 3 day reduction in the

number of days of tube feeding (1 RCT, n=57) and a 1.6 day reduction in time for

transition to bottle feeds (1 study, n=40), confidence intervals and p-values were

not reported in these 2 studies however the differences are described as

‗significant‘. All of these studies provided NNS during tube feeds.

No clear benefit was found on weight gain. The meta-analysis of 3 RCTs found no

effect, a fourth RCT found benefit during the 6 week study period but couldn‘t be

included in the meta-analysis as standard deviations were not reported, and 2 non-

RCTs found no benefit. No clinically important effect was seen on heart rate (4

RCTs) or oxygen saturation (3 RCTs). There was an unclear effect on feeding

performance with 2 RCTs reporting differing results. One RCT reported that infants

in the NNS group had more intake within the first 5 minutes, more total amount of

feeding, less feeding time and a faster feeding rate, the second RCT did not find a

difference between the groups for similar outcomes. The effect on behavioural state

was also unclear with 4 RCTs reporting mixed results and using varied outcome

measures. One RCT found that infants receiving NNS during tube feeding spent

significantly less time in fussy and active awake states during and after a tube feed,

settled more quickly into a sleep state and exhibited less defensive behaviours.

Another RCT found that preterm infants receiving NNS before bottle feeding spent

significantly more time in a quiet awake state and less time in active sleep,

drowsiness, active awake and crying states. Two RCTs found that NNS had no effect

on behavioural state.

The RCT52 which was published after the Cochrane review search date included a

total of 156 infants in 3 groups – only 2 of which (n=104) are relevant here. All

infants were bottle-fed formula. Infants were 32-34 weeks post conceptual age

(PCA), weight AGA, had no medical problems known to affect feeding, Apgar >3 at

1 minute and 5-6 at 5 minutes after birth, and were not using oxygen during

feeding. Infants were randomised to either 7 days of NNS (5 minutes of sucking on

a dummy before each scheduled feed) or no NNS. The third group were given oral

support, this data is not included here. Outcomes were assessed before,

immediately after and 7 days after the 7 day intervention period. It is not clear

whether allocation to groups was concealed. Assessment of outcomes was not

blinded. On average, infants in the NNS group were 1 week older at commencement

of the study, weighed 120g less at birth, and 40g less at commencement.

Assessment of the two main outcomes, volume taken during a feed and duration of

feed, could both be potentially influenced by researchers.

The study reports that both immediately after the 7 day intervention period and at

1 week follow-up, infants in the NNS group consumed more formula and took less

time to feed than infants in the control group (data not provided, p<0.01). Infants

in the NNS group also had less sucking bursts than the control group (data not

provided, p<0.001) both immediately after intervention and at 1 week follow-up,

however this may be due to the increased time the control infants spent feeding.

There were no differences between the groups in length of sucking bursts, number

or length of pauses.



The RCT53 which was not identified by the Cochrane systematic review included 319

infants <34 weeks gestation whose mothers wished to breastfeed and who had no

abnormalities precluding feeding. The study was undertaken at two Australian

tertiary hospitals. Infants were randomised to four groups: bottle and dummy,

bottle and no dummy, cup and dummy, cup and no dummy. Infants received bottle

or cup feeding as randomised when the mother was unavailable to breastfeed or if

additional volume was required after breastfeeding. Infants randomised to receive a

dummy were encouraged to use them during tube feeds and whenever they were

restless. The intervention lasted for the infant‘s entire hospital stay. For analyses

about the impact of NNS, data was aggregated into two groups on an intention to

treat principle: dummy and no dummy.

The study was not blinded. Groups were similar at baseline however infants

randomised to receive dummies had mothers who were less likely to have breast-

fed previously (34 vs 46%, difference 12% (95%CI 0.3%, 24%)). The method of

assessment of outcomes was not described.

There was substantial non-compliance, 31% of infants who were randomised to no

dummy were introduced to a dummy. The proportion of infants fully or partially

breast-fed at discharge, 3 or 6 months did not differ between infants randomised to

receive a dummy and those randomised to not receive a dummy. Similarly no

difference was seen in length of stay.

Table 2 Rate of breastfeeding (BF) in infants randomised to dummy or no

dummy Number (%)

without dummy

Number (%)

with dummy

OR (95%CI); p-value

Discharge

Fully BF 79/152 (52%) 85/151 (56%) 0.84 (0.51, 1.39); 0.50

Any BF 107/152 (70%) 108/151 (72%) 0.83 (0.45, 1.50); 0.53

Any BF at 3 months 58/142 (41%) 53/141 (38%) 0.99 (0.56, 1.77); 0.98

Any BF at 6 months 43/141 (30%) 34/140 (24%) 1.23 (0.66, 2.30); 0.51

In summary, these studies provide evidence that NNS during tube feeding can substantially

reduce LOS for preterm infants. The impact of NNS during tube feeding on other outcomes is

unclear with studies reporting differing results, however no negative effects were reported in

any of the studies.

The evidence for NNS before or after sucking feeds is less clear with some studies reporting

positive impacts on behavioural state and other studies not finding any effect. There is limited

data suggesting NNS before bottle feeds increases milk intake and reduces feeding time.

Only one study was found which investigated the impact of NNS on breastfeeding outcomes in

preterm infants. There is substantial discussion in the literature about the concept of ‗nipple-

confusion‘ suggesting that the different action infants use when sucking on dummies or bottles

compared to that used when breastfeeding may lead to difficulties in establishing

breastfeeding. The Monash Newborn Parent Group highlighted that mothers are also concerned

about the possibility of ‗nipple confusion‘ and are anxious not to do anything that might

jeopardise the potential of their infants to breastfeed. However no research was identified to

support this suggestion and the RCT described above53 did not find any association between

the use of dummies and difficulties in breastfeeding.

No research was identified which examined the effectiveness of other forms of NNS, such as

sucking on an emptied breast or a carer‘s finger. The GDG agreed that these forms of NNS

may also be useful. Searching did not identify any evidence evaluating the safety of sucking on



clinician‘s gloved finger, however the GDG noted that if a gloved finger was used, that gloves

should be latex-free to avoid latex allergy.

The GDG also noted that there were no studies of NNS for infants born at ≥37 weeks GA and

that these infants would not normally require NNS as they are more likely to have well-

developed sucking skills.

The GDG noted that many clinicians would normally offer only a small number of sucks before

a feed in order to support development of an alert calm and organised state for feeding and

would be concerned that extended periods of NNS before sucking feeds might result in the

infants being too tired to suck nutritively. The GDG also noted that NNS is more often used

before a sucking feed for infants who are agitated, fussy or upset and that infants who are

already alert and organised might be less likely to benefit from NNS.

The GDG noted that recent research in populations including both term and preterm infants,

including a meta-analysis of case-control studies,97 and a subsequent population-based case-

control study98 have suggested that use of dummies may be protective against Sudden Infant

Death Syndrome (SIDS).

In light of the evidence of some potential benefit and no harm in preterm infants, the potential

protective effect against SIDS in the broader infant population, and in line with current local

practice the GDG agreed that:

B Infants <37 weeks corrected age should be offered non-nutritive sucking during

gavage feeds when alert or restless, with their mother‘s permission.

C Infants <37 weeks corrected age may be offered a short period of non-nutritive

sucking to support establishment of a calm, organised state before sucking feeds,

with their mother‘s permission.

D Non-nutritive sucking may be offered on a dummy, emptied breast, mother‘s finger

or carer‘s gloved (non-latex) finger, as per the mother‘s preference as is feasible.

In the absence of evidence for older infants the GDG agreed that:

D Infants ≥37 weeks corrected age who require gavage feeds may be offered non-

nutritive sucking during gavage feeds when alert or restless, with their mother‘s

permission.

D Infants ≥37 weeks corrected age may be offered a short period of non-nutritive

sucking to support establishment of a calm, organised state before sucking feeds,

with their mother‘s permission.

5.3 Demand or scheduled feeds

When infants are receiving enteral feeds, these feeds can be provided either at scheduled

intervals or in response to ‗hunger cues‘, such as sucking on fingers, hand-to-mouth motions,

or inability to settle, demonstrating the infant‘s desire or readiness to feed (referred to as

demand or ad libitum feeding).



Demand feeds are a potentially important part of the transition to ‗normal‘ feeding patterns for

preterm infants. This transition process may include kangaroo care, long periods of

uninterrupted skin-to-skin contact with the mother‘s chest and other interventions designed to

support the successful introduction of breastfeeding.

Feeding at scheduled intervals (often 3 or 4 hourly) is simpler to manage and ensures the

infant receives adequate nutrition. Demand feeding allows the infant to develop skills to self-

regulate their needs for periods of uninterrupted sleep and periods of feeding. It has been

suggested that scheduled feeding might cause increased stress and energy expenditure by

interrupting the infant‘s natural sleep-wake cycle, and that demand feeding might lead to

inadequate energy intake.

A variety of approaches have been trialled to overcome these perceived risks, including semi-

demand, demand and ad libitum feeds. These approaches vary in the extent to which infants

are allowed to regulate the frequency and volume of their feeds. There is no consistency of use

in the literature of the terms ‗demand‘ and ‗ad libitum‘. These terms are sometimes used

interchangeably by authors, and other times with specific (varying) meanings.

Our searching identified a protocol for a Cochrane systematic review which has not yet been

completed54, and several RCTs comparing demand and scheduled feeds in NICU and special

care nursery settings. Two articles were excluded57,59 as the data they reported was also

reported in other included articles.58,60 Another study55 was excluded as it was a cross-over

trial in which infants were demand or schedule-fed for only two days and no patient outcomes

were reported.

All of the studies identified were in populations of infants who were fully enterally fed, or who

were being introduced to sucking feeds.

The quality of the RCTs was generally low, description of the nature of feeding

protocols was often very poor and both the intervention and follow-up periods

were often very short.

The first RCT56 included 36 infants who weighed <2500g at birth and ≥1800g at

study entry, whose weight was AGA. They were required to be in NICU ≥24hrs,

have a coordinated suck, swallow, breathe (SSB) pattern, be on full enteral feeds,

still having at least 1 gavage feed per day, be normo-glycaemic, and have no

major abnormalities. Infants were randomised to demand or scheduled feeds.

Infants in the demand-fed group were considered to be demanding a feed when

they cried, tried to suck their fingers or dummy, moved actively and rooted, or did

not settle with position or nappy change. Infants in the schedule feeding group

were fed 3 or 4 hourly according to weight, gestational and post-natal age, and

were gavage or bottle-fed according to nurse assessment of their ability (no

further details given). Little information was provided about the feeding protocols,

and no details were given as to when demand-fed infants were gavage fed.

Schedule-fed infants were on average slightly heavier (1984 vs 1897g), had a

higher GA at birth (33 vs 32 weeks) and were younger (15.4 vs 20.7 days) at

baseline.

Infants in the demand-fed group had fewer gavage feeds per day (0.1 vs 4.8 per

infant, p<0.0001). This difference in management of demand-fed and schedule-

fed infants means that it is difficult to separate the impact of demand feeding from

the impact of increased sucking feeds on the outcomes.

The study found that demand-fed infants had similar volume intake, fewer feeds

per day (5.1 vs 7.8, p<0.0001) and shorter period of time until feeding well

enough for discharge (2.7 vs 8.9 days, p<0.0001) compared to infants in the

1-



schedule-fed group.

The second RCT58 included 81 infants who were receiving full enteral feeds with a

PCA of 32-34 weeks, whose weight was AGA and who had no major abnormalities.

Infants were randomised to semi-demand or scheduled feeds. Fortified breastmilk

or commercial formula were used.

Infants in the semi-demand-fed group were given 10 minutes of NNS every 3

hours. If the infant was then assessed to be restless or wakeful they were offered

a sucking feed. If they were asleep, the infant was reassessed after 30 minutes. If

still asleep at 30 minutes, or a proportion of their volume remained after oral

feeding, they were gavage fed. Infants in the schedule-fed group were given a

sucking and/or gavage feed at 3 hourly intervals for ≤30 min. Sucking feeds were

started at 1 per day and then increased as per protocol. Any volume remaining

after a sucking feed was gavage fed. Semi-demand-fed infants were slightly

heavier (1630 vs 1569g) than infants in the schedule-fed group at baseline.

The study found that infants fed according to the semi-demand protocol had

shorter time to full sucking feeds (5 vs 10 days, p<0.001), and that there was no

difference in weight gain between the groups.

The third RCT60 included 100 fully breast or bottle-fed preterm infants in 2

different hospitals who were <35 weeks GA at birth, weight AGA, and who were

not transferred to another hospital before discharge. Infants were randomised to

demand or scheduled feeds.

Details of the intervention were minimal. In the demand-fed group, feeds were

given ad libitum (not further described). In the schedule-fed group, feeds were

given 3 or 4 hourly (depending on in which of the 2 hospitals the infant was being

cared for). Follow-up was for only 5 days.

The study found that infants who were demand-fed had a lower caloric intake

(data not provided), but this difference was not quite statistically significant

(p=0.06). Weight gain was not different between the groups.

The fourth RCT61 included 32 infants who were <37 weeks of age at birth, weight

AGA, had no major abnormalities, weighed ≥1550g at study entry, were on full

oral feeds, had coordinated SSB and blood glucose >40mg/dl (~2.2mmol/l).

Infants were randomised to demand or scheduled feeds.

Infants in the demand-fed group were fed if they demonstrated 2 or more of:

rooting, sucking fist or fingers, hand-to-mouth action, crying or inability to settle

after nappy change or positioning. A maximum of 5 hours was allowed between

feeds. The schedule feeding protocol was not detailed.

The study found no clinically important difference in weight gain at 6 days,

hospital stay was shorter in the demand-fed group (7.2 vs 8.4 days) but the

difference was not statistically significant. Gavage feeding was more common in

the schedule-fed group (data not provided) and the number of feeds per day was

less in the demand-fed infants (~3.0 vs 4.3, p value not provided)

The final study62 was a pseudo-RCT which included 10 infants who were ≤34

weeks at birth, had no major abnormalities, weight AGA, weighed ≥1500g at

study entry, had blood glucose ≥40mg/dl (~2.2mmol/l), and were on formula

only. Odd numbered infants (sic) were allocated to demand or scheduled feeds by

coin toss, even numbered infants were automatically allocated to the other group.



Infants in the demand-fed group were fed if they demonstrated 2 or more of:

crying, hand-to-mouth action, sucking, rooting, inability to settle after nappy

change or positioning. Infants were woken for feeds after 5 hours if not self-

woken. Schedule-fed infants received oral and/or gavage feeds at 3-4 hour

intervals.

This study found a lower nutritional intake in the demand-fed infants (statistical

significance not provided) however data on weight gain was not provided.

The low quality of reporting of the available evidence, as well as the variety of infant

populations and the range of feeding protocols followed, makes it difficult to make firm

conclusions. When compared to scheduled feeding, demand feeding in response to hunger

cues appears to reduce time to full sucking feeds and length of hospital stay. None of the

studies reported any adverse effects of demand feeding.

The GDG noted that demand-feeding of breast-fed infants also depends on the mother‘s

availability, and decisions about whether an infant would be demand feed should be made in

conjunction with the mother.

In light of the evidence in infants being introduced to sucking feeds, the GDG agreed that:

B Infants aged ≥35 weeks who meet the criteria for sucking feeds should usually be

demand-fed, if the mother agrees, with consideration given to the infant‘s blood

sugar, sucking ability and weight gain. These infants should be offered a sucking

feed when they demonstrate hunger cues such as sucking on fingers, hand-to-

mouth motions, rooting actions or inability to settle after nappy or position change.

D Infants aged <35 weeks will not necessarily display hunger cues and should

therefore be schedule-fed. Demand feeding may be considered for infants in this

age group who meet the criteria for sucking feeds, display hunger cues, and are

otherwise stable and well.

D An interval should be set after which an infant will be fed if they have not

demonstrated hunger cues. This interval should not usually be more than 5 hours in

infants aged <36 weeks. Longer intervals may be set for older infants.

5.4 Breastfeeding methods

Mothers of both term and preterm infants often need support and advice to successfully

establish breastfeeding. The NHMRC Dietary Guidelines for Children and Adolescents in

Australia state that:

―Most women experience a number of … difficulties while breastfeeding. If

appropriate advice and support are not given, a mother may prematurely terminate

breastfeeding. In studies in Australia and other developed countries, the main

reason for termination cited by women is a perceived insufficient milk supply. The

actual number of mothers who may be physiologically incapable of providing

sufficient milk is, however, extremely low. For the remainder of women who

prematurely terminate breastfeeding, there are numerous causes—both biological

and psychological—the majority of which are temporary and can be resolved with

experienced advice or avoided by better preparation, hospital management or

appropriate support.‖

―All health professionals need to constantly promote the benefits of breastfeeding.

The benefits should be discussed with mothers (or potential mothers) at the earliest

opportunity, such as the first antenatal visit. Health professionals should ensure that



their patients know about the protective properties of breastmilk and the risks

involved when infants do not receive it.‖

―Since breastfeeding should be regarded as ‗best practice‘, mothers wishing to

initiate formula feeding in a health care institution should sign a form indicating

their informed consent.‖87

The process of successfully establishing breastfeeding is more complex in the case of preterm

or LBW infants. The psychological challenge of establishing breastfeeding given the extra

stress and difficulties of having a preterm infant cannot be overestimated. However, given the

beneficial effects of breastfeeding, mothers of preterm and LBW babies should be supported

and encouraged to breastfeed, as well as educated about both the importance of breastfeeding

and the reality that establishment of breastfeeding is a complex and sometimes difficult

process.

5.4.1 Skin-to-skin contact

Skin-to-skin contact and kangaroo care have been suggested as methods of supporting

establishment of breastfeeding in preterm infants.

Skin-to-skin contact refers to contact between the skin of the infant and the skin of the parent

or carer. The term ‗kangaroo care‘ is sometimes used interchangeably with ‗skin-to-skin

contact‘ but is also used in a number of different ways in the literature. Kangaroo care was

originally developed in a resource-poor setting as an alternative to expensive technological

approaches to stabilisation of temperature and respiration in low birth weight infants, but has

now been adapted for wider use. The major components of kangaroo care are:

1) skin-to-skin contact, babies are kept between the mother‘s breasts firmly attached to the

chest in an upright position,

2) frequent and exclusive or nearly exclusive breastfeeding, and

3) early discharge from hospital regardless of weight or gestational age.

The term ‗kangaroo care‘, is used in a number of different ways in the literature, and often in

situations which only involve skin-to-skin contact. In addition, the duration of skin-to-skin

contact undertaken can vary from a few minutes to many hours per day.



Although this research has some weaknesses, it suggests that kangaroo care or skin-to-skin

contact have a range of benefits for infants, both on establishing breastfeeding and other

physiological outcomes.

Kangaroo care or skin-to-skin contact may assist establishment of breastfeeding in a number

of different ways. As the NHMRC guidelines state ―Close mother–child contact immediately

after birth helps to establish lactation, and frequent suckling or feeding on demand helps to

maintain it.‖ 87

Box 3. Kangaroo care and skin-to-skin contact

In parallel with the guideline development process, the Centre for Clinical Effectiveness

undertook a systematic review to answer the question: ―In preterm or low birth weight

infants does kangaroo care or skin-to-skin contact compared to standard care improve

outcomes?” A summary of this review is provided here and the full version of this review is

available from CCE.

Medline, All EBM Reviews, CINAHL and The Cochrane Library databases were searched in

late 2007. Studies assessing the effect of kangaroo care or skin-to-skin contact in low birth

weight or preterm infants were eligible for inclusion in the systematic review. All outcomes

were included.

One Cochrane systematic review and two randomised controlled trials were found which

evaluated kangaroo care. Seventeen randomised controlled trials and two cohort studies

were found which evaluated skin-to-skin contact alone. The studies included were generally

of moderate quality with a range of methodological problems.

The results of these studies suggest kangaroo care has a range of benefits and does not

adversely affect preterm or low birth weight infants. Two studies reported beneficial effects

on breastfeeding including increased likelihood of not exclusively breastfeeding at discharge

(RR=0.41, 95%CI 0.25, 0.68, n=279) and increased number of breastfeeds per day (12.5

vs 2.0, p<0.01, n=8). Other studies reported no impact on exclusive breastfeeding at 41

weeks corrected GA (n=663, 1 RCT), at 1 (n=379, 2 RCTs) or 6 months follow-up (n=146,

1 RCT), or at 12 months corrected age (n=589, 1 RCT). Possible benefits of kangaroo care

were seen for a range of other outcomes including improved temperature control, growth

and development, maternal satisfaction and maternal sense of competence and costs.

The results of the studies examining skin-to-skin contact alone suggest that similarly it has

a range of benefits and does not adversely affect non-inubated preterm or low birth weight

infants. Three studies reported benefits of skin-to-skin contact on breastfeeding including:

Increased likelihood of breastfeeding in all infants (OR=2.8, 95%CI 1.0, 8.3, n=60),

with an increased effect seen in infants who spent more than 30% of their time in

skin-to-skin contact (OR=10, 95%CI 1.8, 57, n=27).

Increased proportion of infants breastfeeding at discharge (90 vs 61%, p<0.05) and

1 month post-discharge (50 vs 11%, p<0.01) (n=50).

Increased duration of lactation (9.2 vs 5.1wks, p=0.017, n=71).

One study reported no impact on breastfeeding at 3 or 6 months follow-up (n=50) and

maternal milk production (n=50). Benefits of skin-to-skin contact were also seen for a

range of other outcomes including hypoglycaemia, cardiorespiratory stability, hypothermia,

growth and development, infant behaviour, maternal psychological outcomes, pain

response, maternal preference, and family interaction.

There is some concern about skin-to-skin contact for intubated infants. In one study the

FiO2 required to maintain oxygen saturation levels increased significantly during skin-to-skin

contact (45% vs 36%, p=0.009, n=14) and these infants also had greater difficulty regulating their body temperature.



See recommendations below.

5.4.2 Breast contact

It is important to distinguish between skin-to-skin contact and breast contact. As discussed

above, skin-to-skin contact appears to be a helpful and safe intervention for any

physiologically stable infant.

Infants, and particularly those with less well developed sucking skills, will often progress from

skin-to-skin contact which is suitable for most infants, to early breast contact which involves

no sucking, to non-nutritive sucking on an emptied breast, to early breastfeeding attempts and

then full breastfeeding.

Breast contact, initiated specifically with the intention of providing infants with an opportunity

to suck at the nipple, without previously emptying the breast, should usually only be offered

for those infants who meet the criteria for readiness for sucking feeds. While contact with the

breast is not synonymous with breastfeeding, even very young infants may suck at the nipple

while having breast contact. Although this sucking may not be nutritive, encouraging breast

contact may mean that infants are introduced to sucking breastfeeds earlier than might

otherwise be the case. As is discussed elsewhere non-nutritive sucking may also be offered on

an emptied breast.

No research was identified that specifically investigated the value of breast contact in preterm

or low birth weight infants.

To obtain the benefits of skin-to-skin contact and breast contact both on establishment of

breastfeeding and on other health outcomes, the GDG agreed that:

B Infants should be introduced to skin-to-skin contact, preferably with their mother,

as soon as they are physiologically stable, unless precluded by other criteria.

B Parents, and particularly mothers, should be encouraged to spend as much time as

possible in skin-to-skin contact with their physiologically stable infant.

D The benefits of skin-to-skin contact should be explained to mothers.

D Infants who meet the criteria for introduction of sucking feeds should be offered the

breast.

5.4.3 Assessing the success of breastfeeds

Assessing the success of breastfeeding is a complex task. In infants being introduced to

breastfeeds, early breastfeeds are successful if the infant and mother have a positive

experience that builds confidence and skills in breastfeeding. As infants gain more experience

and skills in breastfeeding they can begin to move towards full breastfeeding. For infants to be

able to receive all their nutritional requirements through breastfeeding, adequate milk must be

transferred to support growth and development. This is dependent on the combination of

successful attachment, effective sucking, maintenance of physiological stability and adequacy

of the available milk volume.

5.4.3.1 Assessing milk transfer

In early breastfeeds the volume of milk transferred is not of major importance as the emphasis

is on supporting a positive breastfeeding experience. In infants moving towards full



breastfeeds it is important to ascertain that the infant is receiving enough milk via breastfeeds

to support their growth and development. Low milk transfer may result from limitations in

attachment, sucking method and duration or availability of breastmilk.

Three trials were identified which examined the accuracy of different methods of assessment of

milk transfer in infants in special care nurseries or NICUs. Two of these trials investigated test

weighing,109,120 and the other investigated clinical assessment.110

The first study109 included 50 clinically stable infants with GA <38 weeks, on full

enteral feeds with no ventilatory support who were bottle-fed formula. Test

weights before and after feeding on an electronic infant scale were compared with

the volume of formula calculated by subtracting the amount remaining in feeder

from the initial volume. The study reported the mean absolute difference (MAD),

mean percent error (MPE) and Pearson‘s correlation. Description of inclusion

criteria and study subjects was limited. Selection of subjects was non-consecutive

and not further described. It was unclear how formula loss due to drooling or

vomiting was accounted for. Two cases were excluded from the analysis – one

because of an error and the other because volume of intake by test-weight was

recorded as -340g.

The study found that volume of intake varied between infants from 10 to 67ml

(mean 33, SD 11.5). The MAD between test-weights and volume intake was

1.97ml, MPE was 5.96%, the Pearson‘s‘ correlation was r=0.97, and 54.2% of

differences were less than or equal to 5ml.

The second study120 included 100 bottle-fed infants in a special care nursery.

Inclusion criteria were not further described. Infants were weighed before and

after feeding in a Salter ‗Trent‘ baby scale, and bottles were weighed before an

after feeding on a Metler automatic laboratory balance. Four measurements were

excluded on the basis that they were ―obviously wrong‖. There is no mention of

blinding.

The study found that the MPE was 23.5%. In 68% of measurements MPE was

>10%, in 47% it was >20% and in 27% of measurements it was >30%. Test

weighing tended to underestimate small feed volumes and overestimate large feed

volumes. Errors were largest for infants taking in very small or very large

volumes. The authors concluded that ―Routine test weighing using standard baby

scales… [is] a poor indication of feed intake‖.

The third study110 included 39 breastfeeding sessions in 29 infants with GA <38

weeks at birth (mean 31 weeks) who had no congenital abnormalities and were

not oxygen dependent. Milk transfer was assessed by a certified lactation educator

(CLE) using clinical indices (but no specific tool or checklist) and the result was

compared with test-weights before and after feeding on an electronic Smart Model

20 infant scale. The study reported, MAD, MPE and the correlation between

volume as assessed by test-weights and clinical indices. The study also reported

accuracy of specific clinical indices but the methodology is too weak to make this

data useful (indices were taken from medical record, were infrequently and

variably recorded, and subjectively identified and grouped).

The inclusion criteria are not well described, the description of the infants has only

minimal detail and selection is described as ―non-random‖ but this is not further

explained. The appropriateness of comparison with test weighing is unclear in

light of the above studies. It is also unclear how weights assessed on the scales

were transformed into volumes for comparison with clinical assessment.

The study found that only small volumes of milk were consumed (mean 5.1ml) as

2-



assessed by test weighing. Differences between the clinical estimate and the test-

weight were ―large and random‖ with MAD 5.1ml (SD 8.11) and MPE 205.62,

r=0.48. These differences were independent of infant age and weight.

These studies all have substantial methodological flaws which make it difficult to interpret their

results.

A systematic review of assessment of milk transfer in term infants has suggested that test

weighing is an inaccurate assessment method which can substantially under or overestimate

milk transfer.123 It is unclear whether these results might also apply to preterm infants given

the usually smaller milk volumes, increased evaporative water losses and differences in

breastmilk density in preterm infants. The GDG noted that infants are weighed every two days

to track weight gain as part of usual practice. In preterm infants weighing before and after

each feed would require a substantial amount of additional handling that may be stressful to

the infant and also distress the mother.

Several indicators are commonly used to assess whether the infant has received an adequate

volume of milk, these include an adequate duration of sucking, contentment and remaining

asleep after a feed and the mother‘s assessment of breast emptying. The GDG noted that it is

also important to be aware of the volume of milk usually expressed by the mother as an

estimate of the volume available to the infant in one feed. The value of these individual

indicators in preterm infants has not been assessed in any methodologically rigorous study.

No studies were identified that examined the impact on the infant of test weighing compared

to clinical assessment of milk transfer.

It is important to note that even if little milk is transferred, as is likely during early

breastfeeding attempts, these sessions are still very important and useful in establishing

breastfeeding skills and preparing for nutritive breastfeeding. Mothers should be encouraged to

continue to offer breastfeeds to infants regardless of the volume of milk transferred.

A supportive and comfortable environment for breastfeeding should be provided to

encourage breastfeeding.

5.4.3.2 Positioning and attachment

When being breast-fed, the infant should be held unwrapped in a sidelying position, close to

the mother‘s chest. In a ―chest-to-chest, chin-to-breast‖ arrangement, the infant‘s neck will be

slightly extended. No evidence was identified which determined the most accurate way of

assessing attachment.

The GDG agreed that:

When an infant has successfully attached, their

chin touches the breast

nose is not blocked

mouth is wide open

cheeks are round and full

lips are splayed out

tongue is forward over the lower gum

mouth covers much of the areola particularly on the 'chin side'.

Assessing attachment can be difficult and the GDG agreed that where clinicians were unsure

they should check with an experienced clinician.



5.4.3.3 Assessing effectiveness of sucking

Preterm infants may not always suck with adequate strength or coordination to receive enough

milk during a breastfeed. Effective sucking is characterised by a rhythmic alternation of suction

and expression/compression (with visible working of jaw muscles) with swallowing and

breathing for a period of several minutes. This pattern of sucking and ability to maintain

sucking without fatigue may develop over time as the infant has more experience in

breastfeeding.

Assessing effectiveness of sucking can be difficult and the GDG agreed that where clinicians

were unsure they should check with an experienced clinician.

5.4.3.4 Maintenance of physiological stability

In infants being introduced to breastfeeds it is important to ensure that breastfeeding does not

cause substantial physiological instability. Infants who are having their first sucking feeds may

have isolated apnoeic, desaturative or bradycardic events as they learn to coordinate sucking

and swallowing. The GDG agreed that breastfeeding should not be withheld as a result of such

isolated desaturative events but that infants with isolated episodes of oxygen desaturation

should be watched carefully during feeds. If such episodes occur frequently, or an infant has

apnoeic or bradycardic events, breastfeeding should be discontinued and the infant‘s readiness

for breastfeeds discussed with senior clinical staff. The GDG agreed that:

D During early breastfeeding attempts, little milk is likely to be transferred and the

emphasis should be on supporting mothers and infants to have a positive

experience as these oral feeding opportunities are important in establishing

breastfeeding.

D Give regular encouraging, positive feedback to mothers of infants being introduced

to breastfeeding.

D Assess breastfeeding effectiveness on the basis of the combination of successful

attachment, effective sucking, maintenance of physiological stability and availability

of adequate milk volume, in consultation with the infant‘s mother, and weight gain

assessed by ongoing alternate day weights.

C Test weighing should not be routinely used to assess the volume of milk transferred

during a breastfeed.

D Breastfeeding should not be withheld as a result of isolated desaturative events but

infants with isolated episodes of oxygen desaturation should be watched carefully

during feeds.

D If episodes of oxygen desaturation occur frequently or an infant has apnoeic or

bradycardic events during breastfeeds, sucking feeds should be discontinued and

the infant‘s readiness for sucking feeds discussed with senior clinical staff.



5.4.4 Rate of introduction and advancement of breastfeeds

No research evidence was identified which determined the most effective protocol for

introducing or advancing breastfeeds. The GDG emphasised that there is substantial variation

between infants in the process of establishing breastfeeds, and that it was very important to

consider each infant individually.

The GDG noted that during introduction of breastfeeds it is vital to give the mother and infant

every opportunity to successfully establish breastfeeding. With this in mind they noted that it

was important not to overfeed infants via gavage tube (either on scheduled feeds or as a ―top-

up‖ after breastfeeds) as this may diminish the infants motivation to suck during breastfeeds.

They also noted that in early breastfeeds breastmilk volumes may be low and regular sucking

opportunities are important in establishing breastmilk supply.

The opinion of the GDG was that infants aged less than 35 weeks who are ready for

introduction of sucking feeds be ofference sucking feeds when they demonstrate feeding cues,

as these young infants often have less well developed sucking skills. If infants breastfeed

successfully (see below) then the proportion of feeds provided by breastfeeds should gradually

be increased until the infant is receiving all their feeds by breastfeeding.

The GDG agreed that older infants should be introduced to ad libitum breastfeeding, that is,

should be given as many breastfeeds per day as the mother can provide and the infant can

successfully receive. The number of breastfeeds per day would then be increased on the basis

of the infant‘s increased sucking skill and the mother‘s availability and milk supply.

The GDG further agreed that an infant who is scheduled to receive a gavage feed and is

displaying feeding cues, should be offered a breastfeed in preference to the gavage feed if the

mother is available.

D An infant who meets the criteria for readiness for sucking feeds, who is scheduled to

receive a gavage feed and is displaying feeding cues, should be offered a breastfeed

in preference to the gavage feed if the mother is available.

The GDG highlighted that when an infant is being introduced to breastfeeds, it is important to

be flexible in scheduling these feeds and work with the mother‘s availability.

D Infants who successfully breastfeed should have the proportion of their feeds

provided by breastfeeding gradually increased based on the infant‘s ability to

receive the breastfeed, and the mother‘s availability to provide breastfeeds.

D In an infant being introduced to breastfeeding, the need for additional feed to be

provided as a ‗top-up‘ after a breastfeed should be decided in consultation with the

mother, on the basis of the infant‘s weight, gestation and the apparent success of

the feed.

Breastfeeds are usually introduced in the context of replacing a scheduled gavage feed. The

GDG believed that it was important to be flexible in the timing of these feeds to give the

mother every opportunity to breastfeed. That is, if a mother is planning to be present to

breastfeed for a scheduled feed and the mother is not present on time, nursing staff should

wait 20 or 30 minutes before feeding the infant via another method.

D In infants who are being introduced to breastfeeds, timing of feeds should be

flexible to give the mother every opportunity to breastfeed.



If a mother is planning to be present to breastfeed for a scheduled feed and is not

present on time, wait up to 30 minutes before feeding the infant via another

method.

D If the infant is awake up to 30 minutes before a scheduled feed and the mother is

present to breastfeed, offer the infant a breastfeed.

D If the infant is asleep at the time of a scheduled breastfeed and the mother is willing

to wait, wait up to 30 minutes before attempting to wake the infant.

D If the infant cannot be aroused 30 minutes after a scheduled breastfeed, provide the

feed by gavage.

The above discussion relates to infants who are being introduced to breastfeeds and who are

schedule-fed. The GDG noted that during the introduction of breastfeeds many infants will also

move from scheduled feeds to demand feeds. This transition process needs to be managed to

ensure that the infant continues to receive adequate nutrition and gain weight. The GDG

agreed that:

D An infant who is regularly waking for breastfeeds, feeds well and rests well between

feeds, and continues to gain weight appropriately should be introduced to demand

feeds if the mother agrees.

D An interval should be set after which an infant will be fed if they have not

demonstrated hunger cues. This interval should not usually be more than 5 hours in

infants aged <36 weeks. Longer intervals may be set for older infants

5.4.5 Supplementary feeding methods

Infants who are not able to receive all their nutritional requirements through breastfeeds will

need to have further feeding provided by other methods. This may occur because the infant‘s

sucking skills are not yet well developed or because the mother does not wish or is unable to

exclusively breastfeed.

The GDG noted that exclusive breastfeeding is an ideal aim for most infants, and that mothers

should be supported to exclusively breastfeed wherever this is possible and it is their wish. It is

important to provide support for the method of feeding that the mother wishes to continue

once the infant has been discharged home. The GDG further noted that there are several

different situations in which exclusive breastfeeding may not be feasible and clinicians should

be flexible in meeting the varying needs of individual infants and families.

5.4.5.1 Mothers who wish to exclusively breastfeed

If the mother is aiming to fully breastfeed her infant, but is unable to be present for all feeds,

then a decision needs to be made about what method of supplementary feeding will be used.

Potential feeding methods include cup, bottle, gavage, spoon, dropper or finger feeding.

A systematic review111 was identified that compared the effect of bottle feeding to other

supplementary feeding methods in the establishment of breastfeeding. Two RCTs published

since the systematic review search date were also identified.53,112 The data from one of these53

was included in the systematic review (the same authors were responsible for both the review

and the RCT).



The systematic review111 included 5 RCTs with a total of 535 infants born <37

weeks GA whose mother had chosen to breastfeed. The review compared infants

who were randomised to bottle feeding during the transition from tube to

breastfeeds to infants who received supplementary feeds via some other method

(cup, tube, spoon, dropper, finger). The search terms used to identify studies were

provided but there was no information about how these terms were combined.

There was substantial variation between the studies in inclusion criteria, outcome

definition and results.

The review found that time to reach full sucking feeds was shorter in bottle-fed

infants compared to cup fed infants (MD 10.4 days, 95%CI 4.7, 16.1, 1 RCT,

n=290) and likelihood of any breastfeeding was higher in infants who had not

received bottles on discharge (marginally RR 1.1, 95%CI 1.0, 1.3, 3 RCTs,

n=465), at 3 months (RR 1.3, 95%CI 1.1, 1.6, 3 RCTs, n=444) and at 6 months

(RR 1.5, 95%CI 1.1, 2.1, 2 RCTs, n=364). The trend was the same when infants

were analysed in subgroups of GA <28 weeks and GA 28-37 weeks and when cup

feeding was compared with bottle feeding, but the results were no longer

statistically significant. The review also found that the likelihood of any

breastfeeding was higher in infants who had received tubes rather than bottles on

discharge (RR 1.4, 95%CI 1.1, 1.7, 1 RCT, n=84), at 3 months (RR 1.7, 95%CI

1.2, 2.4, 1 RCT n=83) and at 6 months (RR 2.1, 95%CI 1.2, 3.6, 1 RCT, n=83).

Similarly, likelihood of fully breastfeeding on discharge was higher in infants who

had not received bottles (RR 1.4, 95%CI 1.2, 1.7, 3 RCTs, n=401) and the trend

was the same when infants were analysed in subgroups of GA <28 weeks and GA

28-37 weeks and when cup feeding was compared with bottle feeding but the

results were only marginally significant. Likelihood of fully breastfeeding at 3 and 6

months was higher in infants who had received tubes rather than bottles (RR 2.3,

95%CI 1.3, 4.2, 1 RCT, n=84) and (RR 2.9, 95%CI 1.4, 6.3, 1 RCT, n=84).

The review did not find any difference in weight gain between groups. Length of

hospital stay was higher in infants who had not received bottles (MD 6.6 days,

95%CI 1.9, 11.4, 2 RCTs, n=385). This difference remained significant when

infants were analysed in subgroups of GA <28weeks and GA 28-37 weeks. There

was significant heterogeneity between studies with an increase in LOS of 1.6 days

(34.6 vs 33 days, p=0.68, n=84) in the tube fed infants and 10 days in the cup

fed infants (59 vs 48 days, p=0.01, n=301), as compared to bottle-fed infants.

This may result from differences between studies in the characteristics of the

populations or the interventions. Discharge criteria were not provided in either of

the studies.

There was no difference between groups in incidence of apnoea, bradycardia,

mean HR or respiratory rate in 1 RCT. The authors of another RCT reported a

difference in incidence of apnoea and bradycardia, however the data provided does

not support the p-value quoted. We have communicated with the authors of the

review and the statistician of the study, however the correct interpretation is still

unclear. Mean oxygen saturation was higher in cup fed infants compared to bottle-

fed infants in 1 RCT (96.5 vs 94.5%, p=0.02, n=112) but lowest oxygen

saturation did not differ between groups in another RCT. Mean number of

desaturations was lower in cup fed infants compared to bottle feeding in 1 RCT

(0.05 vs 0.13, p=0.02, n=112) but there was no difference in the other RCT.

1-

The RCT published after the systematic review search date112 included 54 infants

born <35 weeks GA and aged >30 weeks at study entry. Infants had no congenital

abnormalities, were on full nasogastric milk feeds, were expected to have a LOS of

at least another 7 days and had mothers who wished to breastfeed. Infants were

randomised to bottle feeding or cup feeding during the transition from tube to

breastfeeds. The authors calculated that 188 infants would be required to

1-



demonstrate a clinically important difference, however only 54 infants were

recruited. The study provided little data for comparison of the groups at baseline.

Outcome assessment was potentially subjective as assessment of post-discharge

breastfeeding was by phone interview. The difference between withdrawal from

the trial and failure to establish breastfeeding is unclear. Fourteen mothers (11

from the cup feeding group) withdrew from the trial before discharge.

The study found a marginally non-statistically significant trend that cup-fed infants

were more likely to be exclusively breast-fed at discharge than bottle-fed infants

(37 vs 15%, 95%CI for difference -1%, 43%). A similar proportion of infants in

both groups were having some breastfeeds at discharge, exclusive breastfeeding

at term and some or all breastfeeds at 6 weeks post-term. These results are

difficult to interpret given the inadequate number of infants included in the study.

Although these studies have some flaws they suggest that when deciding whether or not to

use bottles in infants whose mothers wish to breastfeed there is a tension between the

increased likelihood of breastfeeding in infants who do not receive bottles, and the associated

increased length of stay in these infants. There does not appear to be a strong impact on other

outcome measures.

The GDG noted that no evidence was identified that directly compared cup feeding with tube

feeding in infants whose mothers wish to breastfeed, but that it appears that cup feeding may

result in a larger increase in length of stay than tube feeding when compared to bottle feeding.

The GDG emphasised that support for breastfeeding was vitally important and, while

acknowledging that avoidance of bottlefeeding might to lead to longer length of hospital stay,

suggested that this could be overcome if appropriate plans were made to increase the

proportion of feeds provided by breastfeeding while the infant is in hospital, and provide

appropriate support once the infant was discharged home.

In light of the evidence, the central importance of supporting establishment of breastfeeding,

and in line with current local practice the GDG agreed that

Mothers who wish to exclusively breastfeed should be fully supported and

encouraged to achieve this goal.

B Infants being introduced to breastfeeds whose mothers wish to exclusively

breastfeed should have any supplementary feeds provided by tube. A plan should be

made to increase the proportion of feeds provided by breastfeeding while in hospital

and provide appropriate post-discharge support.

5.4.5.2 Mothers who do not wish to exclusively breastfeed

If the mother is not aiming to exclusively breastfeed her infant, and her infant is having some

breastfeeds, it may be appropriate to replace remaining tube feeds with bottle feeds with the

mother‘s permission. For bottle feeding methods see section 5.5.

There is some concern that using bottle feeds as an adjunct to breastfeeding may reduce the

likelihood of maintaining even partial breastfeeding, however no evidence was found to

examine this. The GDG agreed that:

D If the mother wishes to primarily (but not exclusively) breastfeed her infant, bottle

feeds should not be introduced until breastfeeding is well established and the infant

is >35 weeks old.



D If the mother does not wish to primarily breastfeed her infant, and her infant has

successfully breast-fed, bottle feeds may be introduced with the mother‘s

permission.

5.4.6 Gavage tube placement during feeds

Members of the GDG questioned whether in infants being introduced to breastfeeds, the

gavage tube should be left in place during breastfeeds. The presence of a gavage tube during a

breastfeed might potentially increase the difficulty of breathing or swallowing and might

therefore result in a less successful breastfeed. This is particularly important in the case of

orogastric tubes which may significantly affect the infant‘s ability to latch on and suck.

No studies were found which investigated the impact of presence or removal of orogastric

tubes during sucking feeds. Two articles were identified which examined the impact of

presence of nasogastric tubes during oral feeds.42,114 One was excluded114 as the second

article42 extended on the data reported in this study.

This prospective cohort study included 20 VLBW infants with no severe

neurological or physical anomalies who had a mean age of 49 days. Infants were

observed during two oral feeds, one with the nasogastric tube and one with the

tube removed. The order of the feeds was random – however the method of

randomisation was not described. Eighteen of the infants were randomly sampled

and 2 ―conveniently‖ sampled – this is not further described. A desaturation event

was defined as a fall in oxygen saturation below 90%.

The study found that a total of 15 infants had desaturation events during oral

feeds, 13 without the nasogastric tube and 15 with the nasogastric tube present.

Presence of the nasogastric tube increased the length of desaturation events by a

mean of 8 seconds (26.4 vs 18.7 secs, p<0.05) but did not affect the number of

desaturation events per oral feed (5.33 (SD 5.67) without or 5.53 (SD 5.30) with

the nasogastric tube).

2-

In the light of the limited evidence, and in line with current practice the GDG agreed that:

D When an infant is being introduced to breastfeeds, orogastric tubes should be

removed for the duration of the breastfeed. Nasogastric tubes may be left in place

or removed as the clinicians and mother feel is most appropriate.

5.5 Methods of bottle feeding

As was discussed in section 5.1, breastfeeding has a range of established benefits for both

mothers and infants. Some mothers however are unable to or do not wish to fully breastfeed.

These mothers may choose to partially or fully bottle feed their infant. As discussed in section

5.1.1, a mother‘s informed decision to breast or bottle feed should be respected.

Bottle feeding should, therefore, only be introduced if the mother agrees.

The GDG agreed that:

D Bottle feeds require the mother‘s consent.



5.5.1 How to bottle feed

No research evidence was identified to direct the techniques that should be used for bottle

feeding preterm infants. In the absence of evidence and in line with current local practice the

GDG agreed to the following process:

Position of the infant

Feed in a quiet, calm environment.

Hold the infant in your arm against your chest. Support the infant's head gently on

your arm, so that it rests in a straight line with the infant‘s body, not curled up, and

not rolling back. Look at and talk softly to the infant. If the mother or father is

present, encourage them to feed the infant and suggest they provide skin-to-skin

contact.

Small infants may be held facing the clinician, with the infant‘s head in the

clinician‘s hand, and the infant‘s body supported along the clinician‘s forearm.

If skin-to-skin contact is not being provided, wrap the infant lightly, gently pulling

the infant‘s arms into its body and tucking the blanket underneath the infant‘s legs.

Alternatively, infants who are in a drowsy state may be unwrapped to help bring

them to a more awake and alert state for feeding.

Position of the bottle

Gently stroke the infant‘s cheek or lips to stimulate a rooting or gape reflex and

encourage them to open their mouth and lower their tongue. Place the teat of the

bottle gently into the infant's mouth. Position the teat fully in the infant‘s mouth so

the infant‘s lips come to the base of the teat and their lips form a seal around the

teat.

Tip the bottle up just high enough so that no air gets into the teat and the teat is

full of milk. Use your hand to support the bottle, holding the bottle from

underneath, close to the teat.

If the infant is not sucking

If the infant does not begin to suck on the teat, use a finger under the chin to

provide chin support and prompt sucking, or remove the teat from the infant‘s

mouth, place a drop of milk on the infant‘s lower lip, reintroduce the teat and stroke

the infant‘s tongue or hard palate with the teat, from the back of the mouth to the

front of the mouth.

If the infant still does not suck, support the jaw and cheeks of the infant gently with

your hand to facilitate a seal around the teat. Consult an experienced clinician if you

are unsure about this procedure. Do not jiggle or squeeze the bottle.

If the infant still does not suck, consider the infant‘s state (e.g. drowsy, irritable),

remove the teat and allow the infant to rest. Attempt to rouse the infant to a more

awake and alert state and reintroduce the teat if the infant‘s state changes. If the

infant does not suck on the second attempt, provide the feed by gavage.

If the infant is choking or coughing

If the infant is choking or coughing, remove the teat from the infant‘s mouth and

put the infant in an upright position. Check the flow from the teat by tipping the

bottle upside-down. If the flow is more than a slow series of drops, change the teat.



When the infant is calm, correctly position the infant and gently reintroduce the

teat, ensuring the bottle is not tipped too high (see above). If choking or coughing

continues stop the feed, provide the remainder of the feed by gavage and notify

senior clinical staff.

If the infant is tiring

If the infant appears to be tiring after a period of sucking, allow a rest break by

removing the teat from the infant‘s mouth and reintroducing it a minute or so later.

Do not jiggle or squeeze the bottle. Consider burping the infant.

If the infant tires before they have consumed all of the scheduled volume, allow the

infant to rest. Reintroduce the teat after a 5 minute rest period. If the infant does

not suck, provide the remainder of the feed by gavage. Total oral feeding time

should not be more than 45 minutes.

After the feed

If the infant consumes all of the provided volume and appears content, burp the

infant gently if desired and either continue skin-to-skin contact or cuddling or

replace in the crib. If the infant consumes all of the provided volume but continues

to display feeding cues (such as rooting, hand-to-mouth actions etc) provide further

volume if on demand feeds.

5.5.2 Choice of teat

Three cross-over studies were identified that compared the effect of different teats on

measures of milk transfer, ventilation and physiological stability.

The first trial117 included 10 preterm infants who were completely bottle-fed (mean

GA 32.3 weeks, postnatal age 23.8 days, birthweight 1648g, study weight 2017g).

Infants were given 1 feed each with an Enfamil single hole teat for preterm infants

and an Enfamil single hole teat for term infants. The order of teats was random.

Outcomes assessed were ventilation volume, respiratory rate and tidal volume.

The study provided only limited description of included infants and the method of

randomisation was not described. It was not specified whether investigators were

blind to which teat was being used, but this seems unlikely. The small sample size

may mean the study is unable to detect a clinically important difference.

The study found that although ventilation fell during bottle feeding, there were no

statistically significant differences between teats on any outcome measure.

The second trial by the same authors118 included 12 preterm infants who were

completely bottle-fed and not receiving oxygen (mean GA 30.8 wks, postnatal age

31.3 days, birthweight 1350g). Infants were given 1 feed each with an Enfamil

single hole teat for preterm infants and an Enfamil single hole teat for term

infants. The order of teats was random. Outcomes assessed were total feed time,

total sucking time, amount fed, sucking frequency, sucking pressure, milk flow

rate, proportion of feed time used for sucking.

Like the earlier study, only a limited description of included infants was provided

and the method of randomisation was not described. It was not specified whether

investigators were blind to which teat was being used, but this seems unlikely. The

small sample size may mean the study is unable to detect a clinically important

2-



difference.

The study found that there were no statistically significant differences between

teats on any outcome.

The third trial119 included 10 VLBW infants with no major congenital abnormalities

who were not requiring oxygen (mean birth weight 941g, GA 27 weeks). At 1-2

and 6-8 oral feeds per day, infants received one bottle feed using a Similac

premature teat, one using an Enfamil premature teat and one using a Similac

infant teat. The order of teats was random. The study assessed rate of milk

transfer, percent overall transfer, duration of feed, duration of ‗out‘ times

(burping, coughing, resting), oxygen desaturation, sucking stage, sucking

amplitude and burst length.

As in the other studies, only a limited description of included infants was provided

and the method of randomisation was not described. It was not specified whether

investigators were blind to which teat was being used, but this seems unlikely. The

small sample size may mean the study is unable to detect a clinically important

difference. Duration of ‗out‘ time was at feeder‘s discretion and this may effect

many other outcomes – especially as it is unlikely the feeder was blind to

allocation.

The study found that when infants were receiving 1-2 oral feeds per day there

were no statistically significant differences between teats on rate of transfer,

overall volume transferred, burst duration, behavioural state or total ―out‖ time;

but that sucking stage, amplitude, and duration were different between teats,

however the clinical importance of this is unclear. When infants were receiving 6-8

oral feeds per day there were no statistically significant differences between teats

on any outcome.

In summary, there are only a small number of studies of small sample size and relatively low

quality investigating the impact of choice of teat on bottle feeding outcomes. These studies did

not identify any statistically significant differences between the teats they investigated,

however this may be due to the small number of infants included in each study.

The GDG noted that there is wide variation between the milk flow provided by individual teats,

even within a particular brand or type of teat, to the extent that some teats do not even have

a hole. While there was some suggestions that particularly young infants, should be initially

trialled on a slow flow teat, the difficulty of identifying a teat which met this definition, and the

wide variation within teat types meant the GDG was not able to make a recommendation for a

particular type of teat.

In the absence of evidence supporting a particular type of teat, the GDG agreed that

D If an infant who is being bottle-fed is having difficulty feeding, reassess the choice

of teat and their readiness for bottle feeds.

5.5.3 Assessing the success of bottle feeds

Successful bottle feeding requires adequate milk transfer to support growth and development.

This is dependent on the combination of successful attachment, effective sucking and

maintenance of physiological stability. Usually, infants being introduced to bottle feeds will be

schedule-fed. In these infants, a bottle feeding attempt is successful if the infant receives all or

most of the prescribed volume within 30 minutes. In infants being demand-fed, success is



assessed on the basis of the combination of effective sucking, maintenance of physiological

stability and weight gain assessed by ongoing alternate day weights. The GDG agreed that:

D In infants being bottle fed, assess bottle feeding success on the basis of the

combination of effective sucking, maintenance of physiological stability, weight gain

assessed by ongoing alternate day weights and whether the infant receives all or

most of the prescribed volume within 30 minutes.

D During early bottle feeding attempts, little milk is likely to be transferred and the

emphasis should be on supporting mothers and infants to have a positive

experience as these opportunities are important in establishing oral feeding.

D Bottle feeding should not be withheld as a result of isolated episodes of oxygen

desaturation but infants with isolated episodes of oxygen desaturation should be

watched carefully during feeds.

D If episodes of oxygen desaturation occur frequently or an infant has apnoeic or

bradycardic events during bottle feeds, sucking feeds should be discontinued and

the infant‘s readiness for sucking feeds discussed with senior clinical staff.

5.5.4 Rate of introduction and advancement of bottle feeds

No research evidence was identified which determined the most effective protocol for

introducing or advancing bottle feeds.

The opinion of the GDG was that infants aged less than 35 weeks who were ready for

introduction of sucking feeds should initially receive one bottle feed, as these young infants

often have less well developed sucking skills. If infants bottle feed successfully (see above)

then the proportion of feeds provided by bottle should gradually be increased until the infant is

receiving all their feeds by sucking feeds.

The GDG agreed that older infants should be introduced to ad libitum bottle feeding, that is,

should be given as many bottle feeds per day as they can successfully receive. Similarly to

younger infants, the number of bottle feeds per day would then be increased on the basis of

the infant‘s increased sucking skill.

The GDG further agreed that an infant who is scheduled to receive a gavage feed and is

displaying feeding cues, should be offered a bottle feed.

D Infants who successfully bottle feed should have the proportion of their feeds

provided by sucking feeds gradually increased based on the infant‘s ability to

successfully receive sucking feeds.

D In infants who are being introduced to bottle feeds, timing of feeds should be

flexible to give the mother every opportunity to feed the infant.

If a mother is planning to be present for a scheduled bottle feed and is not present

on time, wait up to 30 minutes before feeding the infant.



D In an infant being introduced to bottle feeding the decision to ‗top-up‘ by gavage

infants who do not receive their prescribed volume should be made in light of the

proportion of the scheduled feed received, the proportion taken at other recent

feeds, weight gain assessed by alternate daily weight and the infant‘s age.

Bottle feeds are usually introduced in the context of replacing a scheduled gavage feed. The

GDG believed that it was important to be flexible in the timing of these feeds to give the infant

every opportunity to have a sucking feed, and also to encourage parents to provide the bottle

feed.

D If the infant is awake up to 30 minutes before a scheduled feed, and the parent

either is present or is not planning to be present for the feed, offer the infant a

bottle feed.

D If the infant is asleep at the time of a scheduled bottle feed, and the parent either

can wait or is not planning to be present for the feed, wait up to 30 minutes before

attempting to wake the infant.

D If the infant cannot be aroused 30 minutes after a scheduled bottle feed, provide

the feed by gavage.

The above discussion relates to infants who are being introduced to bottle feeds and who are

schedule-fed. The GDG noted that during the introduction of bottle feeds many infants will also

move from scheduled feeds to demand feeds. This transition process needs to be managed to

ensure that the infant continues to receive adequate nutrition and gain weight. The GDG

agreed that:

D An infant who is regularly waking for bottle feeds, feeds well and rests well between

feeds, and continues to gain weight appropriately should be introduced to demand

feeds.


Feed intolerance and necrotising enterocolitis 67

6 Feed intolerance and necrotising enterocolitis Awareness of the signs of feed intolerance and necrotising enterocolitis is essential for

successful management of the introduction of enteral feeding.

6.1 Necrotising enterocolitis

Necrotising enterocolitis (NEC) is a major concern for infants in NICUs and special care

nurseries. Rates of NEC vary widely between different nurseries, however NEC is associated

with substantial mortality and both short- and long-term morbidity,73,74 particularly in infants

with low weight or gestational age at birth. The underlying pathogenesis has not yet been

firmly established, and successful management relies on early identification of infants with

signs that might be indicative of NEC.

The Modified Bell's Staging Criteria (see Table 3) are the accepted standard for assessing and

classifying infants with NEC. A full evidence-based review of the accuracy of each of these

criteria, or of the management of NEC, is beyond the scope of this guideline.

The GDG agreed to recommend that management of enteral feeding in infants in Monash

Newborn should follow Modified Bell's Staging Criteria.

D Infants who meet the Modified Bell's Staging Criteria for suspected, definite or

advanced necrotising enterocolitis should not be enterally fed.

The GDG also agreed that enteral feeds may be reintroduced in infants who have been

appropriately treated for necrotising enterocolitis, however the management of feeding in

these infants is complex and beyond the scope of this guideline.

D Enteral feeds may be reintroduced in infants who have been appropriately treated

for necrotising enterocolitis, however the management of feeding in these infants

should be decided by a multi-disciplinary group of senior clinicians.



Table 3. Modified Bell's Staging Criteria

STAGE

I. Suspected II. Definite III. Advanced

A B A: Mildly ill

B: Moderately ill

A: Severely ill, bowel intact

B: Severely ill:

bowel perforated

Syste

mic

sig

ns

Temperature instability, apnoea, bradycardia

Same as IA Same as IA Same as I, plus mild metabolic acidosis, mild thrombocyto-penia

Same as IIB, plus hypotension, respiratory acidosis, metabolic acidosis, disseminated intravascular coagulation, neutropenia

Same as IIIA

Inte

sti

nal sig

ns

Elevated pre-gavage residuals, mild abdominal distension, occult blood in stool

Same as IA, plus gross blood in stool

Same as I, plus absent bowel sounds, abdominal tenderness

Same as I, plus absent bowel sounds, definite abdominal tenderness, abdominal cellulitis, right lower quadrant mass

Same as I and II, plus signs of generalised peritonitis, marked tenderness and distension of abdomen.

Same as IIIA

Rad

iolo

gic

sig

ns

Normal or mild ileus

Same as IA Ileus, pneumatosis intestinalis

Same as IIA, plus portal vein gas, with or without ascites

Same as IIB, plus definite ascites

Same as IIB, plus pneumo-peritoneum

6.2 Gastric residuals

Gastric residuals are aspirated to assess whether the enteral tube is correctly positioned in the

stomach and to assess tolerance of enteral feeds.

While there is a substantial body of literature discussing the role of gastric aspirates in

assessment of feed intolerance, there are few trials comparing different approaches, or

examining the ability of characteristics of the aspirated residuals such as volume, colour, etc,

to predict feed intolerance or NEC.

6.2.1 Timing and frequency of aspiration

The timing and frequency of aspiration of gastric residuals has the potential to impact on both

interpretation of the aspirates; as volume or character of the aspirates may vary with how

close to the previous feed the aspiration in undertaken, and also on the workload of clinicians

responsible for managing feeding. Too frequent aspiration of gastric contents may not allow

adequate time for normal gastric emptying.

No research was identified which investigated the impact of varying the timing and frequency

of aspiration of gastric residuals.

6.2.2 Volume and colour of aspirate

The volume of gastric residual aspirated, and the colour or appearance of the aspirate are

characteristics commonly used to assess whether an infant is tolerating enteral feeds.

Traditionally large volume or green gastric residuals have been thought to be indicative of feed

intolerance and potential NEC.



One study was identified76 which compared characteristics of gastric residuals in infants who

developed NEC with those in infants who did not develop NEC. A second study was identified77

which sought to examine the relationship between characteristics of gastric residuals and feed

tolerance measured by volume of milk fed on day 14 of life.

The first study76 was a case-control study that compared characteristics of gastric

residuals in 51 VLBW infants with proven NEC and 102 matched control infants.

Each case infant was matched with 2 control infants based on gestational age,

birth weight, gender and race.

The study found that several measures were higher in infants that developed NEC,

these included:

mean maximum residual in the 6 days prior to cases being diagnosed with

NEC (4.5 vs 2ml, p<0.001)

mean maximum residual as a proportion of feed volume (40 vs 14%,

p<0.001)

mean total residuals as a proportion of feed volume (1.6 vs 0.4%, p<0.001)

mean maximum residual per day (0.9 vs 0.33ml, p<0.005)

mean proportion of feeds with residuals (12.5 vs 6.3%, p<0.01).

Although these differences were statistically significant, there was a wide overlap

in the results of case and control infants, substantially limiting the usefulness of

these measures in a clinical setting. That is, while mean values in the NEC and

control groups differed significantly, individual infants within each group had a

wide variation in results, and no measure would reliably identify infants who would

go on to develop NEC.

2+

The second study77 was a cohort study nested within a RCT. The study included 99

ELBW infants on an early enteral feeding protocol, in which feeds were started at

12ml/kg birth weight/day and advanced every 24 hours by 12ml/kg birth

weight/day. Feeds were decreased or withheld if gastric residuals were greater

than 2ml in infants weighing ≤750g or greater than 3ml in infants weighing 750 -

1000g. Colour of residuals did not affect the feeding plan.

Infants were followed for the first 2 weeks of life and the primary outcome

measure was the volume of milk fed on day 14. Exclusion criteria were broad and

included ―bad general condition‖. Three infants were excluded from analysis

because they died, in 2 cases as a result of severe RDS and in 1 case as a result of

an intracranial haemorrhage.

NEC occurred in 5 infants. These infants were included in the analysis and all had

zero volume fed on day 14. This small number of infants limits the ability of the

study to identify factors related to the development of NEC. Gastric residuals were

absent 44.4% of the time, clear 3.3%, milky 36.0%, green-clear 8.9%, green-

cloudy 5.2%, blood-stained 0.2% and haemorrhagic 1.8% of the time.

The study used multiple linear regression to identify the factors that affected the

volume fed on day 14. The regression found that volume fed on day 14 increased

with increasing percentage of zero gastric residuals (p=0.02) or milky gastric

residuals (p<0.0001). An increase of 1% in percentage of milky gastric residuals

increased the volume fed on day 14 by 2.04ml. An increase of 1% in percentage of

absent gastric residuals increased the volume fed on day 14 by 0.74ml. Variation

in mean gastric residual volume and aspirate colours other than milky did not

affect volume fed on day 14.

2-

The evidence available to assist decisions about the interpretation of volume and colour of

gastric residuals is limited. The GDG noted that while the evidence available suggests that, on



average, babies who develop NEC tend to have larger volume aspirates, there is not a clear

distinction between what is normal and what is abnormal, either as an absolute volume, or as

a proportion of the volume of milk fed.

The GDG also noted that while the role of green gastric residuals in assessing feed tolerance

has not been clearly established by research, clinical experience suggests that they are an

important clinical marker of a range of pathological conditions.

No evidence was identified that compared the effect of returning aspirated gastric residuals to

the infant‘s stomach with discarding the aspirate.

In light of the limited available evidence and in line with local clinical practice, the GDG agreed

that:

D In infants without signs of feed intolerance, gastric residuals should not be routinely

aspirated to assess feed tolerance.

D In infants with signs of feed intolerance, gastric residuals should be aspirated to assess

feed tolerance before every second feed, not more than once every 4 hours.

Aspiration of gastric residuals may be required for assessment of tube position.

D If the aspirated gastric residual is <3ml/kg body weight or <50% of the volume of the last

feed, return the aspirate and give the full volume of milk scheduled.

D If the aspirated gastric residual is ≥3ml/kg body weight and between 50-100% of the

volume of the last feed, return the aspirate and give the volume of milk scheduled minus

the volume of the aspirate returned.

D If the aspirated gastric residual is ≥3ml/kg body weight and ≥100% of the volume of the

last feed, or if aspirate volumes are markedly increased at one feed, or steadily increasing

over several feeds, consult a senior clinician to determine management.

D If the aspirated gastric residual is green or heavily blood-stained, consult a senior clinician

to determine management.

6.3 Abdominal girth

Abdominal girth measurements have been used to identify infants with NEC or other

gastrointestinal problems. No evidence was identified which assessed the value of routinely

measuring the abdominal girth of enterally fed preterm infants.

Increases in abdominal girth are commonly due to benign gaseous distension but may also,

less commonly, be due to a variety of pathologies such as NEC and obstruction. Abdominal

girth may increase as a result of gaseous distension due to CPAP particularly in very small

infants.81 Some, but not all, infants with NEC will have abdominal distension.79

The GDG noted that measurements of abdominal girth will depend on a range of factors

including how close to a feed or defaecation the girth is measured and the precise position at



which the measurement is taken. It is also unclear what degree of change in abdominal girth

over time is indicative of pathology. In premature infants without signs of gastrointestinal

disease, abdominal girth has been seen to vary by as much as 3.5cm after feeding.78

Rate of increase of abdominal girth can only be assessed by a series of sequential

measurements over the course of several hours, however this repeated measurement may

result in unnecessary handling of the infant, which can be detrimental, particularly for small

unstable infants.

In the absence of evidence and in line with local practice, the GDG agreed that:

D Abdominal girth should not be routinely measured.

D Consider measuring abdominal girth in infants who appear to have abdominal

distension with suspected underlying pathology such as NEC or obstruction,

preferably at a time when the infant is already being handled.

6.4 Vomiting

Most infants will vomit occasionally and in most cases vomiting is not associated with any

underlying pathology. No evidence was identified that examined the role of vomiting in

identification of feed intolerance or other conditions. The GDG noted that effortless milky

vomits, often called possets, are usually innocuous and that some particular characteristics

should be considered to identify the small proportion of infants in whom vomiting may be

associated with a pathological condition.

In the absence of evidence and in line with local practice the GDG agreed that:

D Enteral feeds should not be stopped or slowed in infants who have occasional,

effortless, milky vomits (possets), particularly when they occur soon after feeding.

D In an infant with persistent or worsening effortless vomiting, with other symptoms

suggestive of significant gastro-oesophageal reflux, such as bradycardia or

desaturation events, consult a senior member of the medical staff before continuing

enteral feeds.

Sometimes vomit can contain blood and when this occurs, usually the vomit is finely streaked

with blood. Less commonly an infant may produce a heavily blood-stained vomit. The blood

found in vomit may be swallowed during birth, swallowed in breastmilk from a cracked or

damaged nipple, or less frequently may arise from ulcers of the infant‘s stomach, upper airway

trauma or Haemorrhagic Disease of the Newborn. The source of the blood can usually be

established either from the history or by performing an Apt test which distinguishes foetal from

adult haemoglobin.

In the absence of evidence and in line with local practice the GDG agreed that:

D Enteral feeds should not be stopped or slowed in infants who have occasional vomits

that are finely streaked with blood and no other symptoms of underlying pathology,

however these infants should be observed closely.

D If an infant produces a heavily blood-stained vomit, or consistently produces vomits



that are finely streaked with blood, consult a senior member of the medical staff to

decide management.

Sometimes vomit will appear green or bile-stained. Bile-stained vomit is potentially very

serious as it may suggest the presence of a bowel obstruction. The GDG agreed that:

D If an infant produces green, bile-stained vomit, stop enteral feeds and consult a

senior member of the medical staff.

The GDG noted that sometimes infants will produce vomits that contain mucous and that this

is not normally associated with pathology. However this should be differentiated from an

excess of mucous or saliva which may be due to tracheo-oesophageal fistula or oesophageal

atresia.

Vomits that contain mucous do not usually indicate pathology, however an excess of

mucous or saliva might indicate tracheo-oesophageal fistula or oesophageal atresia.

Consult a senior member of the medical staff if you are concerned.

6.5 Signs of feed intolerance

No evidence was identified which examined the accuracy of classifying the severity of feed

intolerance. In light of the evidence discussed in sections 4.10.1-4.10.4 above, and in line with

local practice, the GDG agreed that:

D The following signs may indicate feed intolerance:

Persistent or worsening vomits

Consistently finely blood-streaked vomits

Occasional projectile vomits

Persistent or worsening abdominal distension

D The following signs may indicate feed intolerance and require medical review:

NEC as per Modified Bell‘s Criteria

Bile-stained aspirates

Bile-stained vomits

Heavily blood-stained vomits

Persistent projectile vomits


Breastmilk or formula 73

7. Breastmilk or formula

7.1 Benefits of breastmilk

There is a great deal of discussion in the community and health research literature about the

value of breastmilk and the role of formula milk. The decision about which milk to feed

premature infants is very important. Two Cochrane systematic reviews were identified that

compared feeding preterm or low birth weight infants preterm breastmilk,82 and term

breastmilk83 with formula milk respectively.

The first systematic review82 included 1 RCT of 46 infants with GA 27-33wks and

birth weight less than 1600g. Infants were randomised to receive either pooled,

unfortified preterm (less than 35weeks) breastmilk or Enfamil preterm formula.

The study was not blinded. NEC and feed intolerance were withdrawal criteria, and

infants who developed either of these were not included in the analysis of growth

rates. Given that all 6 infants who developed feed intolerance were in the formula

group, this is likely to lead to an overestimate of growth rates in the formula

group.

The study found that infants randomised to receive formula had a higher average

daily weight gain (27.0 vs 23.7g/day, WMD=3.3g/day 95%CI 0.64, 5.96, p=0.02)

but there were no significant differences between the groups in length or head

circumference gain. Infants randomised to formula were more likely to develop

feed intolerance (6/26 vs 0/20, p=0.02) however the small numbers make

interpreting this outcome difficult.

The second systematic review83 included 6 RCTs and a total of 481 infants with GA

<27wks or birth weight <2.5kg. Infants were randomised to receive unfortified

term breastmilk or formula (term or preterm). None of the studies were blinded

except for 1 reporting long-term outcomes. Two studies excluded from the

analysis infants who were withdrawn because of feed intolerance or NEC. As in the

first review above, the tendency for formula fed infant to have higher levels of

feed intolerance means this is likely to lead to an overestimate of growth rates in

the formula group.

The review found that infants randomised to formula had a shorter time to regain

birth weight in 3 of the 4 studies in which this outcome was reported. The

difference between groups varied from 2.8 to 8.5 days. In the fourth study the

effect was said to be non-statistically significantly in favour of breastmilk (1.5 vs

2.5 weeks). Infants randomised to formula also had a higher rate of weight gain in

4 of the 5 studies in which this outcome was reported (WMD=6.1g/day, 95%CI

4.9, 7.3). In the fifth study no significant difference was found.

While there were some statistically significant short-term differences in rate of

increase of crown-heel length, crown-rump, femoral length and head

circumference in favour of formula milk (crown-heel length WMD=1.7 mm/week

(95%CI 1.1, 2.4, 4 RCTs, n=210); occipito-frontal head circumference, WMD=1.6

mm/week (95%CI 1.0, 2.2, 4 RCTs n=232); crown-rump length WMD=0.6

mm/week (95%CI 0.1, 1.1, 1 RCT n=106); femoral length WMD=0.4 mm/week

(95%CI 0.2, 0.6, 1 RCT n=106)). There were no statistically significant differences

in growth parameters at 9 months, 18 months, or 8 years post-term.

NEC and feed intolerance were non-statistically significantly higher in the formula

fed infants (RR 2.5 95%CI 0.9, 7.3 and RR 3.3 95%CI 0.7, 14.8 respectively). We

cannot rule out that these differences are due to chance, however they may be

clinically important.

1-



Because fortification of breastmilk is usual practice (see below) and these systematic reviews

did not identify any studies comparing fortified breastmilk to formula, they are of little value in

guiding practice.

Two further studies84,11 were identified that are potentially relevant. One RCT was found that

randomised infants to receive either fortified pooled, donor breastmilk or preterm formula

when their mother‘s milk was not available.84 The comparison of formula to donor breastmilk is

of little relevance as donor milk is not currently available in Victoria, however the study also

included a non-randomised comparison group of infants who received only their mother‘s milk

(fortified as appropriate). The data from this comparison would be useful except that there are

significant baseline differences between the preterm formula group and mother‘s milk only

group which were adjusted for in the RCT analysis and only the unadjusted data were

presented in the paper. The substantial and systematic variation in baseline characteristics

(infants fed only their mother‘s milk had higher birth weight, higher Apgar scores, less

mechanical ventilation and less chronic lung disease) makes interpretation of the unadjusted

comparative data essentially meaningless.

The second study11 is a cohort study nested in a RCT which included 108

premature infants (not further defined) who were taking part in a RCT of early or

delayed initiation of enteral feeds. The cohort study was a secondary analysis

comparing infants who received predominantly fortified breastmilk (average

>50ml/kg/day) with those infants who exclusively received preterm formula. GA

and birth weight were similar between the groups at baseline however, little other

data is provided so it is possible that the groups differed at baseline in a way that

affected the outcomes. For example infants without CLD re more likely to

breastfeed successfully (and therefore to be in the breast-fed group) and are also

more likely to be well enough for discharge earlier. Thus breastfeeding might

appear to be associated with early discharge, when in fact the relationship was

due to differing rates of CLD.

The study found that infants receiving breastmilk had less NEC and late onset

sepsis (1.6 vs 13%, p<0.01 and 31 vs 48% p=0.07 respectively) and fewer

episodes of residuals >2ml/kg (3 vs 7% of any gastric residual volume, p<0.05).

Duration of hospitalisation was also lower in breast-fed infants (73 vs 88 days,

p=0.03). Infants receiving breastmilk had slower weight gain (22 vs 26 g/kg/day,

p<0.01), smaller length increments and lower discharge weight (2428 vs 2998g,

p<0.001). There was no difference in days to regain birth weight. The authors

conclude that predominant breastmilk feeding is associated with less morbidity

and shorter duration of hospitalisation, however this relationship may well be

confounded as discussed above.

1-

Breastmilk has been established to have many benefits for term infants, in both the short and

long-term. These are likely to also apply to preterm infants, although the research has yet to

be carried out in this population. As well as providing nutrients and energy, breastmilk

provides a range of other substances that help to develop an infant‘s immune system, promote

development and protect the infant during its vulnerable early life. In the International Code of

Marketing of Breastmilk Substitutes, to which Australia is a signatory, the WHO states that

―the anti-infective properties of breastmilk help to protect infants against disease‖.88 Many of

the beneficial properties of breastmilk are still being discovered and these properties are not

provided in any breastmilk substitute.

The NHMRC Dietary Guidelines for Children and Adolescents in Australia87 state that

―Breastmilk is a living tissue that cannot be duplicated by any other means. It is very

important for pre-term infants.‖ and further that ―feeding of fortified breastmilk should be

actively promoted in premature infants.‖



Breastmilk also contributes to a range of benefits to both child and mother arising from

breastfeeding as described in section 5.1.

In light of this the GDG agreed that:

D Breastmilk should be provided to infants in preference to formula.

D The benefits of breastmilk should be explained to mothers.

D Mothers should be encouraged and supported to provide breastmilk for their infants

whenever possible, either through breastfeeding infants who are having sucking

feeds, or by expressing breastmilk for infants receiving tube or bottle feeds.

7.2 Availability of breastmilk

As discussed above, breastmilk has many benefits over formula, and should be provided in

preference to formula whenever possible. However there are some circumstances in which

enough breastmilk to supply all of an infant‘s nutritional requirements may not be available

and in these circumstances it may be appropriate to supplement breastmilk with formula.

For most mothers, of both term and preterm infants, breastmilk is not available in large

quantities until a few days after birth. This is normal and should not be used as a basis for

starting formula or other feeds. As discussed in section 2.1.3, while enteral feeds should

usually be started within 48 hours of birth, they may be withheld for up to 72 hours after birth

while waiting for breastmilk to be available if the mother wishes.

After normal birth, initiation of lactation can take a few days. As a result, milk volumes are

initially low, and healthy, term babies receive little milk for the first few days of life. As an

example, some studies have reported milk transfer rates of 6-10ml/kg bodyweight/day on day

one and 13-25ml/kg bodyweight/day on day two.121-2

Authors have suggested that these low, normal milk transfer rates challenge current fluid

management protocols,121-2 some of which recommend provision of much higher rates of fluids

of 60-65ml/kg/day from day one.

In light of this, the GDG agreed that:

Decisions about the need for provision of supplementary fluids for full or near-term

infants in the first few days of life should be made in the light of low fluid intake in

healthy term infants.

This is particularly important when deciding whether or not to provide intravenous fluids, which

introduce potential for infection and may reduce the infant‘s drive to feed.

For some mothers, the quantity of breastmilk available after the initial post-birth period will

still be limited. Mothers should be encouraged to continue to express breastmilk regularly

when volumes of breastmilk are low, as regular expressing may increase breastmilk supply.

Other strategies to increase breastmilk supply should also be considered – however a review of

these is beyond the scope of this guideline.

If breastmilk is not available in sufficient quantities to provide an infant‘s nutritional

requirement, provision of supplementary IV fluids or formula should be discussed with the



mother and a senior clinician. The GDG emphasised that breastmilk should be given in

preference to formula whenever it was available.

In line with current local practice, the GDG agreed that:

D If breastmilk is not available in sufficient quantities to provide an infant‘s nutritional

requirements beyond 72 hours after birth, provision of supplementary IV fluids or

formula should be discussed with the mother and a senior clinician.

D Encourage mothers to express breastmilk regularly when volumes of breastmilk are

low.


Role of parents in establishing feeding 77

8. Role of parents in establishing feeding Parents of infants in Neonatal Intensive Care Units (NICUs) and Special Care Nurseries (SCNs)

are central to the process of establishing feeding, and the partnership between parents and

clinicians is fundamental to ensuring infants receive the best possible care. It is vital that

parents are deeply involved in the establishment of feeding, as well as in the other general

activities of caring for their infant while they are in NICU and SCN.

In order to determine how parents can best be involved in establishment of feeding their

infants in NICU and SCN, GDG consulted with parents who have experience of having infants in

NICU and SCN via surveys and focus groups. Consumer representatives were also included on

the GDG.

Several important themes emerged from this consultation and these are described below.

Essentially parents wanted to be well informed, receive practical education and positive

encouragement, and to have the opportunity to be thoroughly involved in decision making and

care delivery during the establishment of feeding for their infant.

8.1 Importance of parent involvement

Parents are the primary care givers for their child and it is important that this relationship is

supported and respected in NICU and SCN. The early days of an infant‘s life are vitally

important in establishing a strong bond between the parents and their child. Having an infant

in NICU and SCN can be very frightening, disempowering and challenging for parents who

expected to have a healthy baby and to take them home a few days later, so it is especially

important that the parental role is actively encouraged and appreciated.

A positive relationship between parents and clinicians, both individually and as a group, can

make an enormous difference to a family‘s experience in NICU and SCN. A clinician who makes

a special effort to build a relationship with the family is highly valued by parents.

Clinicians in NICU and SCN need to take every opportunity to involve parents in decision

making and then to respect the decisions they make. Parents should also be encouraged and

supported to be as involved as possible in feeding their infant and providing other general

care. As well as involving parents in providing feeds and other forms of care wherever

possible, clinicians can also support parents to be involved by encouraging them to interact

with their infant in creative ways such as providing cuddles, stroking their head, reading

books, singing and playing music.

When involving parents in care, it is also important that clinicians understand the many

difficulties parents face when they have an infant in NICU and SCN. Many parents will have

responsibilities for other children, and some will have to travel long distances to visit NICU and

SCN. The stresses on these families can be enormous. Each infant and each family is unique,

and their unique position needs to be considered in planning for how best to involve parents in

care, as well as in decision making about the infant‘s care.

One important issue for parents was the importance of shared development of a feeding plan

which was then respected and followed. Parents want to be consulted in the development of a

feeding plan for their infant, they want the feeding plan to be clearly communicated with them,

and they want the feeding plan to be followed by all staff. Two reasons why this is important

were highlighted. Firstly a feeding plan that outlines the timing and nature of feeds, means

parents can plan to be present at the scheduled time of a feed, and be involved in delivering

that feed. It is therefore important for clinicians not to deviate from the plan. Secondly,

parents also like feeding plans because they give an indication of the next steps for their

infant, and help ensure a consistency of practice in establishing feeding.



8.2 Communication, information and education

Parents repeatedly highlighted the central importance of communication, information and

education in NICU and SCN.

Parents highlighted that education is very important for first-time parents, particularly if their

infant was born before they had an opportunity to attend antenatal education. Clinicians

should be aware that these parents are likely to be especially in need of information and

support.

Parents value written information and brochures about establishment of feeding. They like

having access to lots of information in one document about the facilities available at NICU and

SCN and the expected process of care, that they can refer to at appropriate times. They would

like to receive this information as early as possible, even though it may not all be immediately

relevant.

Many aspects of care in NICU and SCN are new to parents, and for this reason they value

education about clinical topics like gavage feeding. Parents also would particularly like written

information about the benefits and consequences of use of dummies and bottles, especially in

terms of how these effect breastfeeding. However parents also noted that written information

alone is not enough. Parents value the opportunity to talk to clinicians, particularly one-on-

one, about feeding issues and to ask questions.

Breastfeeding and expressing of breastmilk was a particular area in which many mothers

valued practical, one-on-one, ‗hands-on‘ education. Mothers felt that expressing was a way in

which they could contribute to the health of their infant, and that it was very important to get

help, and feel supported to do it well.

Parents noted that every infant is different, and that while videos and brochures were useful,

the information they provided may not apply to every infant. Parents said that this message

should be clearly communicated to parents – so they understand that their experience may be

different. This variation between families and infants also means that clinicians need to be

sensitive to the level of information different families want and need. Some parents like being

informed of long-term goals, so they understand the expected progression for their infant.

Other parents prefer a more day-by-day approach, so that they are not repeatedly

disappointed when their infant does not meet particular milestones. This means clinicians need

to consider the most appropriate approach for each family.

Parents stated that receiving information once was often not enough – and that given the

amount of information they receive, and the levels of stress they are under, it is important that

key pieces of information be repeated, and that the information given by all clinicians should

be consistent.

Consistency of information and education (as well as of clinical practice) was a very important

theme for parents. Parents noted that it can be very confusing and worrying to receive

conflicting information or advice from different clinicians. Consistency of information helps

parents to feel confident and comfortable about the care their infant is receiving.

Parents are also very aware that many different clinicians are involved in caring for each infant

and they believe that the communication with parents should be clearly documented so that

each clinician is aware of how parents have been and want to be involved in their infant‘s care

and the decisions that have been made about establishment of feeding.



8.3 Encouragement, positive feedback and support

Establishing feeding in infants in NICU and SCN can be a challenge, and often there are

setbacks or unexpected events along the way. Parents need lots of support and

encouragement to help their infant to establish feeding.

Clinicians should actively support parental involvement in establishing feeding at all

stages of the process.

Parents have suggested several ways this can be done, these include:

enabling plenty of one-on-one time with nurses

explaining that there is plenty of support available and though the process of

establishing feeding can be challenging, it is possible

providing information on what to expect that includes personal stories from

other families and infants

encouraging expressing and explaining how important breastmilk is and that

expressing helps to maintain and increase supply of breastmilk

encouraging mothers to provide breast contact during tube feeds and at

other times whenever possible

providing positive encouraging feedback on ALL breastfeeding attempts

supporting parents to provide kangaroo care or skin-to-skin contact

giving parents every opportunity to spend time with their infant when the

parent visits NICU and SCN and making this a priority when planning care

encouraging parents to interact with their infant in creative ways, including

providing cuddles, stroking the infant‘s head, reading books, singing and

playing music

providing a variety of opportunities for parents to share their experiences,

support and interact with each other.

Parents noted that clinicians also face a challenge to appropriately encourage breastfeeding

without putting undue pressure on mothers who make an informed decision not to breastfeed.

Parents emphasised that clinicians should be reminded of the importance of respecting a

mother‘s choice, and not adding guilt to the stresses mothers of infants in NICU and SCN

already face.

Mothers also highlighted the fact that fathers should be involved in planning and providing care

for their infant too. Mothers suggested that fathers might need more encouragement to be

actively involved. Mothers believe that staff should proactively talk to fathers as well as

mothers about their infant‘s care.

8.4 Summary

The partnership between parents and clinicians in NICU and SCN is the foundation to ensuring

infants receive the best possible care. It is vital that parents are thoroughly involved in the

establishment of feeding, as well as in the other general activities of caring for their infant

while they are in NICU and SCN. Parents want to receive frequent, consistent, detailed

information about the process of care and their infant‘s progress, be given practical education

and supportive encouragement, and to be thoroughly involved in decision making and care

delivery during the establishment of feeding for their infant.


Dissemination and implementation 80

9. Consultation, dissemination and implementation

9.1 Consultation

A broad consultation process was undertaken on the draft of this guideline. Copies of the draft

guideline were made available in the Special Care Nursery and Neonatal Intensive Care Unit at

Monash Newborn, Clayton. A website was established for the consultation process which

included the draft guideline and technical document. Invitations to provide feedback on the

guidelines were sent to:

nursing, medical and allied health staff at Monash Newborn, Clayton

senior nursing and medical staff at the Southern Health nurseries at Dandenong and

Casey hospitals

senior nursing and medical staff at the nurseries at the Angliss and Frankston hospitals

senior nursing and medical staff at the other Victorian level 3 nurseries

other experts in the field both within Australian and internationally.

9.2 Dissemination

The GDG identified a wide range of stakeholders who needed to be made aware of the

existence of the guideline. These included all clinicians working in Monash Newborn (senior and

junior nurses, doctors and allied health staff), parents of infants in Monash Newborn, Southern

Health Executive Management, Southern Health Public Relations, other tertiary care health

services within and beyond Australia, the Victorian Department of Human Services, maternal

and child health nurses, general practitioners and private paediatricians.

To provide information on the existence of the guideline, the GDG suggested a wide range of

dissemination approaches, including:

Clinicians working in Monash Newborn

o having copies of the guideline in Monash Newborn

o making the guideline available on the internet and intranet

o providing in-services and workshops about the guideline

o putting messages in the bulletin

o putting information about the guideline in the resource folder, communication

folder and policies and procedures manual

o organising a grand round about the guideline

o putting up posters about the guideline in the unit

Parents

o making the guideline available on the internet

o providing patient information pamphlets

o having staff inform parents about the guideline

o putting up posters about the guideline in the unit

Southern Health Public Relations

o providing a guideline summary

o having a meeting to discuss potential opportunities

Victorian Department of Human Services

o as advised by public relations

Other tertiary care health services

o presenting at conferences and writing journal publications

o running an Australian Neonatal Network highlight seminar

o emailing on Australian Neonatal Network list

Maternal and child health nurses


o putting a notice in Australian Nursing Federation Maternal & Child Health Special

Interest Group newsletter

General practitioners and private paediatricians

o putting a notice in Divisions of General Practice newsletters



Dissemination and implementation 81

9.3 Implementation

Clinicians who are to use the guideline need more than just to be aware of the guideline. The

GDG suggested that a range of methods be employed to help staff to implement the guideline.

These included:

ensuring easy access to guidelines in the unit, in hardcopy and electronic form

providing regular education sessions, workshops and in-services focusing on specific

areas of feeding management addressed in the guideline for medical, nursing and allied

health staff

discussing feeding management according to the guideline in every handover

attaching letters about the guideline to payslips

running workshops/focus groups to identify how the implementation is progressing and

what changes could be made to improve implementation

including awareness of feeding management recommendations in assessment of yearly

competency

developing a clinical tool for use at the bedside to implement the guideline

recommendations

having key staff act as role models in use of the guideline

including information about the guideline in the orientation pack for all new staff

assessing and reporting compliance with the guideline recommendations on a regular

basis.


Outcomes and audit 82

10. Outcomes and audit There are many different frameworks for developing outcome and audit measures. For the

purpose of evaluating the implementation of this guideline we defined two types of measures;

process and outcome. Process measures assess the success of the implementation process and

the degree of implementation achieved. Outcome measures assess whether the

implementation has improved clinical practice and patient outcomes.

10.1 Process measures

10.1.1 Implementation process

• Proportion of target groups that received the interventions outlined in the

implementation and dissemination plan. This information will be collected by

attendance sheets and other similar documentation.

10.1.2 Degree of implementation

• Proportion of eligible patients with feeding plans and proportion of feeding plans

completed correctly. This information will be collected by audit of patient records.

10.2 Outcome measures

10.2.1 Clinical practice

Members of the GDG reviewed the guideline recommendations to identify which of the

recommendations would require a change from existing clinical practice. A subset of these

guideline recommendations were prioritised for implementation by the GDG and a detailed

implementation and evaluation plan has been developed. Details of this plan area available

from the Centre for Clinical Effectiveness.

10.2.2 Patient outcomes

To develop a set of outcome measures we reviewed the original aims of the guideline

development. These aims included reducing length of stay in hospital, enhancing the

establishment of oral feeds and improving the consistency of clinical practice and

communication with parents

In light of this, the GDG agreed that the following measures should be assessed, prior to and

after implementation of the guideline through a feeding plan:

Length of stay

Parents‘ satisfaction with care and communication

Parents‘ expectations on admission and actuality on discharge

Proportion of infants discharged home on full, partial or no oral feeds

Patterns of establishment of breastfeeding

Patterns of establishment of feeding in particular groups (i.e.: 32-34 weeks GA with no

RDS or other issues).

Mean gestational age at which first breastfeed offered

Mean gestational age at which first bottle feed is ordered

Mean gestational age at which first formula feed is ordered

Mean gestational age at which infant has no requirement for tube feeds

Interval between first breastfeed and bottle feed

Proportion of term or near term infants given nasogastric tube

Proportion of term or near term infants receiving intravenous fluids

Proportion of infants receiving formula in the first 72 hours

The GDG agreed that these outcomes could be assessed through a combination of routinely

collected data, medical record audit and surveys of parents.


Guideline Development Process 83

11. Guideline development process

11.1 Guideline Development Group

The development of this guideline was undertaken by the Monash Newborn Feeding Guideline

Development Group at Monash Medical Centre, Southern Health. Kate Rivett, Claire Harris,

Jane Reid and Fiona Wilkinson co-ordinated the process and Tari Turner was responsible for

identifying and appraising the evidence, and drafting the guideline.

Members of the Guideline Development Group included representatives from all clinical areas

in Monash Newborn at Monash Medical Centre, Clayton relevant to the management of feeding

and consumers.

Members of the group were:

Name Role

Sandra Akerman Registered Nurse

Maria Balakas Parent

Dianne Bazzucchini Registered Nurse

Elizabeth Carse Neonatologist

Janella Christie Speech Pathologist

Eleanora Cotroneo Clinical Nurse Specialist

Janet Courtot* Nurse Unit Manager

Trish Cove Senior Nurse, Clinical Nurse Specialist

Nina De-Sainte-Croix Registered Nurse

Kerry Draffin Parent

Samantha Gordon Registered Nurse, Mothercraft Nurse

Kim Grasso Associate Nurse Unit Manager

Claire Harris* Director, Centre for Clinical Effectiveness

Amrit Kleingeld Parent

Helena Klette Registrar, Special Care Nursery

Elli Marsili Registered Nurse, Mothercraft Nurse

Jaqui McGregor Nurse Educator

Alison Medhurst* Nurse Unit Manager – Clinical

Martina Moorkamp Senior Registrar, Neonatal Intensive Care Unit

Mel Newton Associate Nurse Unit Manager

Prue Oshlack Lactation Consultant/Family Resource Nurse

Sue Pearce Parent

Andrew Ramsden* Neonatologist

Jane Reid Research Officer, Centre for Clinical Effectiveness

Nicole Rennick Lactation Consultant/Family Resource Nurse

Kate Rivett* Clinical Scholar in Evidence-based practice

Tracy Shand Parent

Sue Tan Clinical Nurse Specialist

Tari Turner Senior Project Officer, Centre for Clinical Effectiveness

Joanne Vien Clinical Nurse Specialist

Natalie Weekley* Speech Pathologist

Fiona Wilkinson Senior Project Officer, Centre for Clinical Effectiveness

Suzanne Worrall Registered Nurse

Victor Yu Neonatologist

*members of the Steering Group

Declarations of interest were made by all members of the Guideline Development Group, no

conflicts of interest were identified.



11.1.2 Acknowledgements

We thank the following people for their important contribution to the development of this

guideline.

Name Role Contribution

Joanne McKenzie Statistician, The Australasian

Cochrane Centre

Advice on undertaking and

interpreting meta-analysis

Damien Jolley Associate Professor Biostatistics,

MIHSR

Advice on interpretation of

statistical results of studies

Mel Wilson Health for Kids Clinical Scholar,

Centre for Clinical Effectiveness

Proofreading and copyediting

Funding for the development of this guideline was provided by the William Buckland

Foundation. Additional support was provided by the Centre for Clinical Effectiveness and

Monash Newborn, Southern Health.



11.2 Methodology

Figure 1. Guideline Development Process

1. Define and scope the guideline

Is the topic related to clinical decision making? no

yes

Are there suitable existing evidence-based guidelines?

2. Convene a multidisciplinary guideline panel

3. Develop the clinical question(s)

4. Identify and appraise scientific

evidence to answer questions

5b. Develop consensus-

based recommendations

that indicate lack of clear

evidence but acknowledge consensus

5c. Make brief non-consensus

statement (state options and acknowledge uncertainty)

7. Consult and pilot test

8. Disseminate and implement

9. Evaluate and revise

5a. Develop evidence-based

recommendations or update existing recommendations

Is there Level I-IV evidence in respect of each recommendation?

Is there consensus?

yes

no

yes

yes

no

Stop

6. Write guideline versions

Is adaptation for local situation required?

yes

no

no



11.2.1 Define and scope the guideline

The scope of the guideline was determined by the Monash Newborn Feeding Guideline Steering

Group. This group consists of senior medical, nursing and allied health staff from Monash

Newborn at Monash Medical Centre, Clayton, and senior staff from the Centre for Clinical

Effectiveness.

The Monash Newborn Feeding Guideline Steering Group agreed to support the development a

guideline which focused on the process of establishing feeding in infants in both the Neonatal

Intensive Care Unit and the Special Care Nursery at Monash Newborn. The Steering Group

decided to exclude questions about the content of feeds (types of formula, nutritional content,

fat type, etc) and to focus on questions about management of the feeding process.

The scope of the guideline was discussed by the Steering Group with the Guideline

Development Group and was accepted by the GDG.

11.2.2 Search for suitable existing evidence-based guidelines

A broad search was undertaken to identify existing evidence-based guidelines. The search

included:

published literature (Medline, CINAHL, Embase),

websites referring to clinical guidelines (including www.guidelines.gov, www.rch.org.au,

www.nhmrc.gov.au, http://www.leitlinien.de)

established neonatal email lists (including the Vermont Oxford Network, the NICHD

Neonatal Research Network, the Canadian Neonatal Network, the British Association of

Perinatal Medicine, the Swiss Society of Neonatology, NICU-Net and Neonatal Talk)

the internet.

Several existing clinical practice guidelines for the management of feeding were identified. In

most cases it was clear that these guidelines were consensus-based (not evidence-based)

documents. Some guidelines appeared to be potentially evidence-based, in these cases we

attempted to contact the authors, or publishing institutions in order to determine the

methodology used in the guideline development. We were not able to identify any relevant

guidelines which had been developed by a methodologically rigorous, evidence-based process.

11.2.3 Convene a multidisciplinary guideline panel

This guideline was developed by a multidisciplinary guideline development group (GDG) which

included representatives of all clinical areas involved in caring for infants in Monash Newborn

and parents. The membership of the group is outlined above in section 11.1.

In the absence of existing evidence-based guidelines for establishment of feeding in infants in

NICU and SCN, the GDG agreed to develop a new guideline, following an evidence-based

process.

11.2.4 Develop the clinical questions

The GDG developed a list of questions that the guideline should address. The list of questions

was revised and expanded in consultation with colleagues, and surveys were sent to members

of our Parent Group. A full listing of the clinical questions is available on request.

11.2.5 Identify and appraise scientific evidence to answer questions

Searches of published research evidence were then undertaken to identify evidence to answer

these questions. The Cochrane Library (including DARE, CENTRAL, HTA, NHS EED), Medline

(Ovid) and CINAHL (Ovid) were searched. The search was restricted to studies published in

English and undertaken in human subjects.

Evidence identified to answer the clinical questions was appraised using a standard appraisal

format (see section 11.4).

http://www.guidelines/

http://www.rch/

http://www.nhmrc/

http://www/



11.2.6 Develop recommendations

Where evidence was available to answer clinical questions, evidence-based recommendations

were made, with the grade of the recommendation reflecting the quality and generalisability of

the evidence available to answer the question.

Where there was no, or very low quality, evidence to answer a clinical question,

recommendations were made on the basis of the consensus of the GDG.

11.2.7 Write guideline, algorithm and parent information

Each chapter of the guideline was drafted by a member of the GDG with expertise in the

identification and appraisal of research evidence, in consultation with the coordinators of the

GDG process, and then revised and adapted in the light of feedback from GDG meetings.

A systematic process was also followed to develop the parent/carer information. The King‘s

Fund policy and practice guide ‗Producing Patient Information: How to research, develop and

produce effective information resources‘ by Mark Duman (2003) was used to ensure that the

information produced reflected what parents/carers wanted to know and how they wanted to

use it.

Key steps were:

1. Identify need for developing parent/carer brochure

2. Develop aims and objectives of parent/carer brochure

3. Plan project

4. Determine if resource currently exists

5. Clarify audience, content, format

6. Check existing resources against identified needs

7. Develop the parent/carer brochure

8. Pilot the parent/carer brochure

9. Make any required amendments

10. Produce the parent/carer brochure

11. Disseminate

12. Evaluate and review

11.2.8 Consult and pilot test

The final draft was provided to all members of the GDG for comment and revision.

The final document was circulated to key clinical leaders outside the GDG and members of the

Southern Health Executive for review and endorsement.

11.2.9 Disseminate and implement

The guideline will be disseminated as described in Chapter 8. The feeding plan is the primary

method of implementation of the recommendations of the guideline. The feeding plan will form

part of the medical record for all infants who meet the inclusion criteria at Monash Newborn,

Monash Medical Centre.

11.2.10 Evaluate and revise

The impact of the guideline will be evaluated as described in Chapter 6. It is intended that the

content of the guideline be revised in light of further available evidence in 2012.

11.3 Searches

All searches to identify research evidence to answer clinical questions followed a standard

process.

Medline (Ovid 1966-present) and CINAHL (Ovid 1982-present) were searched on each

occasion and The Cochrane Library was searched for all questions regarding management.



The search was restricted to studies published in English and undertaken in human subjects. A

summary of each of the searches undertaken is provided in the technical document, available

from the developers. Full details of the search terms used for each search strategy are

available on request.

11.4 Critical appraisal tables

Standard critical appraisal tables were used to summarise and appraise the quality of each of

the included studies. These tables are provided in the technical document, available from the

developers.


References 89

12. References1. The AGREE Collaboration (2004) The AGREE Instrument,.

http://www.agreecollaboration.org/instrument/

2. Scottish Intercollegiate Guideline Network (2004) SIGN

50: A guideline developers‘ handbook.

3. Kennedy, K. A. and Tyson, J. E. (2005) Early versus

delayed initiation of progressive enteral feedings for

parenterally fed low birth weight or preterm infants. The

Cochrane Database of Systematic Reviews.

4. Boo, N. Y., Soon, C. C. and Lye, M. S. (2000) Risk factors

associated with feed intolerance in very low birthweight infants following initiation of enteral feeds during the first

72 hours of life Journal of Tropical Pediatrics, 46, 272-7.

5. Ho, M. Y., Yen, Y. H., Hsieh, M. C., Chen, H. Y., Chien, S.

C. and Hus-Lee, S. M. (2003) Early versus late nutrition

support in premature neonates with respiratory distress

syndrome Nutrition, 19, 257-60.

6. Tyson, J. and Kennedy, K. (2005) Trophic feedings for

parenterally fed infants. The Cochrane Database of

Systematic Reviews.

7. Flidel-Rimon, O., Friedman, S., Lev, E., Juster-Reicher,

A., Amitay, M. and Shinwell, E. S. (2004) Early enteral

feeding and nosocomial sepsis in very low birthweight

infants Archives of Disease in Childhood Fetal & Neonatal

Edition, 89, F289-92.

8. Berseth, C. L. and McCoy, H. H. (1992) Birth asphyxia

alters neonatal intestinal motility in term neonates

Pediatrics, 90, 669-73.

9. Buch, N. A., Ahmad, S. M., Ali, S. W. and Hassan, H. M.

(2001) An epidemiological study of neonatal necrotizing enterocolitis Saudi Medical Journal, 22, 231-7.

10. Lucas, A. and Cole, T. J. (1990) Breastmilk and neonatal

necrotising enterocolitis Lancet, 336, 1519-23.

11. Schanler, R. J., Shulman, R. J. and Lau, C. (1999)

Feeding strategies for premature infants: beneficial

outcomes of feeding fortified human milk versus preterm

formula Pediatrics, 103, 1150-7.

12. McGuire, W. and Anthony, M. Y. (2003) Donor human

milk versus formula for preventing necrotising enterocolitis in preterm infants: systematic review

Archives of Disease in Childhood Fetal and Neonatal

Edition, 88.

13. Kuzma-O‘Reilly, B., Duenas, M. L., Greecher, C.,

Kimberlin, L., Mujsce, D., Miller, D. and Walker, D. J.

(2003) Evaluation, development, and implementation of

potentially better practices in neonatal intensive care

nutrition Pediatrics, 111, e461-70.

14. Premji, S. S., Paes, B., Jacobson, K. and Chessell, L.

(2002) Evidence-based feeding guidelines for very low-birth-weight infants Advances in Neonatal Care, 2, 5-18.

15. Embleton, N. D. (2004) Enteral feeding in preterm

infants, Newcastle Neonatal Service, Royal Victoria

Infirmary, Newcastle upon Tyne.

16. dos Santos Mezzacappa, M. A. and Collares, E. F. (2005)

Gastric emptying in premature newborns with acute

respiratory distress Journal of Pediatric Gastroenterology

& Nutrition, 40, 339-44.

17. Gounaris, A., Costalos, C., Varchalama, L., Kokori, P., Kolovou, E. and Alexiou, N. (2004) Gastric emptying in

very-low-birth-weight infants treated with nasal

continuous positive airway pressure Journal of Pediatrics,

145, 508-10.

18. Troche, B., Harvey-Wilkes, K., Engle, W. D., Nielsen, H.

C., Frantz, I. D., 3rd, Mitchell, M. L. and Hermos, R. J.

(1995) Early minimal feedings promote growth in

critically ill premature infants Biology of the Neonate, 67,

172-81.

19. McClure, R. J., Chatrath, M. K. and Newell, S. J. (1996) Changing trends in feeding policies for ventilated preterm

infants Acta Paediatrica, 85, 1123-5.

20. Davey, A. M., Wagner, C. L., Cox, C. and Kendig, J. W.

(1994) Feeding premature infants while low umbilical

artery catheters are in place: a prospective, randomized

trial Journal of Pediatrics, 124, 795-9.

21. Mihatsch, W. A., Franz, A. R., Lindner, W. and Pohlandt,

F. (2001) Meconium passage in extremely low

birthweight infants and its relation to very early enteral nutrition Acta Paediatrica, 90, 409-11.

22. Kumar, S. L. and Dhanireddy, R. (1995) Time of first stool in premature infants: effect of gestational age and

illness severity Journal of Pediatrics, 127, 971-4.

23. Jhaveri, M. K. and Kumar, S. P. (1987) Passage of the

first stool in very low birth weight infants Pediatrics, 79,

1005-7.

24. Wang, P. A. and Huang, F. Y. (1994) Time of the first

defaecation and urination in very low birth weight infants

European Journal of Pediatrics, 153, 279-83.

25. Berseth CL, Nordyke CK. (1992) Manometry can predict feeding readiness in preterm infants. Gastroenterology,

103, 1523-8.

26. Soregaroli M, Bonera R, Danti L, Dinolfo D, Taddei F,

Valcamonico A, et al. Prognostic role of umbilical artery

Doppler velocimetry in growth-restricted fetuses. Journal

of Maternal-Fetal & Neonatal Medicine 2002;11(3):199-

203.

27. Vossbeck S, de Camargo OK, Grab D, Bode H, Pohlandt F.

Neonatal and neurodevelopmental outcome in infants born before 30 weeks of gestation with absent or

reversed end-diastolic flow velocities in the umbilical

artery. European Journal of Pediatrics 2001;160(2):128-

34.

28. Adiotomre PN, Johnstone FD, Laing IA. Effect of absent

end diastolic flow velocity in the fetal umbilical artery on

subsequent outcome. Archives of Disease in Childhood

Fetal & Neonatal Edition 1997;76(1):F35-8.

29. McDonnell M, Serra-Serra V, Gaffney G, Redman CW,

Hope PL. Neonatal outcome after pregnancy complicated by abnormal velocity waveforms in the umbilical artery.

Archives of Disease in Childhood Fetal & Neonatal Edition

1994;70(2):F84-9.

30. Malcolm G, Ellwood D, Devonald K, Beilby R, Henderson-

Smart D. Absent or reversed end diastolic flow velocity in

the umbilical artery and necrotising enterocolitis. Archives

of Disease in Childhood 1991;66(7 Spec No):805-7.

31. Simpson C, Schanler RJ, Lau C. Early introduction of oral

feeding in preterm infants. Pediatrics 2002;110(3):517-22.

32. Lau C, Alagugurusamy R, Schanler RJ, Smith EO,

Shulman RJ. Characterization of the developmental

stages of sucking in preterm infants during bottle

feeding. Acta Paediatr 2000;89(7):846-52.

33. Medoff-Cooper B. Nutritive sucking research: from clinical

questions to research answers. J Perinat Neonatal Nurs

2005;19(3):265-72.

34. Bromiker R, Arad I, Loughran B, Netzer D, Kaplan M,

Medoff-Cooper B. Comparison of sucking patterns at introduction of oral feeding and at term in Israeli and

American preterm infants. Acta Paediatrica

2005;94(2):201-4.

35. Thoyre SM, Carlson J. Occurrence of oxygen desaturation

events during preterm infant bottle feeding near

discharge. Early Human Development 2003;72(1):25-36.

36. Lau C, Sheena HR, Shulman RJ, Schanler RJ. Oral feeding

in low birth weight infants. J Pediatr 1997;130(4):561-9.

37. Siddell EP, Froman RD. A national survey of neonatal intensive-care units: criteria used to determine readiness

for oral feedings. JOGNN – Journal of Obstetric,

Gynecologic, & Neonatal Nursing 1994;23(9):783-9.

38. Martell M, Martinez G, Gonzalez M, Diaz Rossello JL.

Suction patterns in preterm infants. Journal of Perinatal

Medicine 1993;21(5):363-9.

39. Medoff-Cooper B, McGrath JM, Shults J. Feeding patterns

of full-term and preterm infants at forty weeks

postconceptional age. J Dev Behav Pediatr

2002;23(4):231-6.

40. Bier JA, Ferguson A, Cho C, Oh W, Vohr BR. The oral

motor development of low-birth-weight infants who

underwent orotracheal intubation during the neonatal

period. American Journal of Diseases of Children

1993;147(8):858-62.

41. Conde-Agudelo A, Diaz-Rossello JL, Belizan JM. Kangaroo

mother care to reduce morbidity and mortality in low

birthweight infants. Cochrane Database Syst Rev.

2003(2):CD002771.

http://www.agreecollaboration.org/instrument/


References 90

42. Shiao SPK, Brooker J, DiFiore T. Desaturation events

during oral feedings with and without a nasogastric tube

in very low birth weight infants. Heart & Lung: The

Journal of Acute and Critical Care. 1996;25(3):236-45.

43. Eichenwald EC, Blackwell M, Lloyd JS, Tran T, Wilker RE,

Richardson DK. Inter-neonatal intensive care unit

variation in discharge timing: influence of apnea and

feeding management. Pediatrics. 2001 Oct;108(4):928-

33.

44. Mandich MB, Ritchie SK, Mullett M. Transition times to oral feeding in premature infants with and without apnea.

J Obstet Gynecol Neonatal Nurs. 1996 Nov-

Dec;25(9):771-6.

45. Hawes J, McEwan P, McGuire W. Nasal versus oral route

for placing feeding tubes in preterm or low birth weight

infants. Cochrane Database of Systematic Reviews.

2004(3):CD003952

46. Dsilna A, Christensson K, Alfredsson L, Lagercrantz H,

Blennow M. Continuous feeding promotes gastrointestinal tolerance and growth in very low birth weight infants.

Journal of Pediatrics. 2005 Jul;147(1):43-9.

47. McGuire W, McEwan P. Transpyloric versus gastric tube

feeding for preterm infants. Cochrane Database Syst Rev.

2002(3):CD003487

48. McGuire W, McEwan P. Systematic review of transpyloric

versus gastric tube feeding for preterm infants. Arch Dis

Child Fetal Neonatal Ed. 2004 May;89(3):F245-8.

49. Premji S, Chessell L. Continuous nasogastric milk feeding

versus intermittent bolus milk feeding for premature infants less than 1500 grams. Cochrane Database Syst

Rev. 2005(3):CD001819.

50. Symington A, Ballantyne M, Pinelli J, Stevens B.

Indwelling versus intermittent feeding tubes in premature

neonates. J Obstet Gynecol Neonatal Nurs. 1995

May;24(4):321-6.

51. Pinelli J, Symington A. Non-nutritive sucking for

promoting physiologic stability and nutrition in preterm

infants. Cochrane Database Syst Rev. 2005(4):CD001071.1.

52. Hill AS. The effects of on-nutritive sucking and oral

support on the feeding efficiency of preterm infants.

Newborn and Infant Nursing Reviews. 2005

Sep;5(3):133-41.

53. Collins CT, Ryan P, Crowther CA, McPhee AJ, Paterson S,

Hiller JE. Effect of bottles, cups, and dummies on

breastfeeding in preterm infants: a randomised controlled

trial. Bmj. 2004 Jul 24;329(7459):193-8.

54. Tosh K, McGuire W. Ad libitum or demand/semi-demand

feeding versus scheduled interval feeding for preterm

infants. (Protocol) Cochrane Database Syst Rev 2005,

(2):CD005255

55. Chang SR, Chen KH. Demand feeding for healthy

premature newborns: a randomized cross-over study.

Journal of the Formosan Medical Association. 2004

Feb;103(2):112-7.

56. Collinge JM, Bradley K, Perks C, Rezny A, Topping P.

Demand vs. scheduled feedings for premature infants. JOGN Nursing. 1982 Nov-Dec;11(6):362-7.

57. McCain GC, Gartside PS. Behavioral responses of preterm

infants to a standard-care and semi-demand feeding

protocol. Newborn Infant Nurs Rev. 2002 Sep;2(3):187-

93.

58. McCain GC, Gartside PS, Greenberg JM, Lott JW. A

feeding protocol for healthy preterm infants that shortens

time to oral feeding.. Journal of Pediatrics. 2001

Sep;139(3):374-9.

59. Pridham K, Kosorok MR, Greer F, Carey P, Kayata S,

Sondel S. The effects of prescribed versus ad libitum

feedings and formula caloric density on premature infant

dietary intake and weight gain. Nursing Research. 1999

Mar-Apr;48(2):86-93.

60. Pridham KF, Kosorok MR, Greer F, Kayata S,

Bhattacharya A, Grunwald P. Comparison of caloric intake

and weight outcomes of an ad lib feeding regimen for

preterm infants in two nurseries. Journal of Advanced

Nursing. 2001 Sep;35(5):751-9.

61. Saunders RB, Friedman CB, Stramoski PR. Feeding

preterm infants. Schedule or demand? JOGNN – Journal

of Obstetric, Gynecologic, & Neonatal Nursing. 1991 May-

Jun;20(3):212-8.

62. Waber B, Hubler EG, Padden ML. Clinical observations. A

comparison of outcomes in demand versus schedule

formula-fed premature infants. Nutr Clin Pract. 1998

Jun;13(3):132-5.

63. Kennedy KA, Tyson JE. Rapid versus slow rate of

advancement of feedings for promoting growth and

preventing necrotizing enterocolitis in parenterally fed

low-birth-weight infants. Cochrane Libr. 2006;1.

64. Salhotra A, Ramji S. Slow versus fast enteral feed

advancement in very low birth weight infants: a randomized control trial. Indian Pediatrics. 2004

May;41(5):435-41.

65. Fenton TR. A new growth chart for preterm babies:

Babson and Benda‘s chart updated with recent data and a

new format. BMC Pediatr. 2003 Dec 16;3:13.

66. Oken E, Kleinman KP, Rich-Edwards J, Gillman MW. A

nearly continuous measure of birth weight for gestational

age using a United States national reference. BMC

Pediatr. 2003 Jul 8;3:6.

67. Kuschel CA, Evans N, Askie L, Bredemeyer S, Nash J,

Polverino J. A randomized trial of enteral feeding volumes

in infants born before 30 weeks‘ gestation. Journal of

Paediatrics & Child Health. 2000 Dec;36(6):581-6.

68. Harrison VC, Peat GM. Feeding intervals and fat

absorption in preterm babies. S Afr Med J. 1992 Feb

1;81(3):160-1.

69. Wells DA, Gillies D, Fitzgerald DA. Positioning for acute

respiratory distress in hospitalised infants and children.

Cochrane Database of Systematic Reviews. 2006(1).

70. Mizuno K, Aizawa M. Effects of body position on blood

gases and lung mechanics of infants with chronic lung

disease during tube feeding. Pediatrics International.

1999 Dec;41(6):609-14.

71. Mizuno K, Itabashi K, Okuyama K. Effect of body position

on the blood gases and ventilation volume of infants with

chronic lung disease before and after feeding. American

Journal of Perinatology. 1995 Jul;12(4):275-7.

72. Mosca NW. Holding premature infants during gavage feeding: effect on apnea, bradycardia, oxygenation,

gastric residual, gastrin, and ehaviour a state; 1995.

73. Blakely ML, Lally KP, McDonald S, Brown RL, Barnhart

DC, Ricketts RR, et al. Postoperative outcomes of

extremely low birth-weight infants with necrotizing

enterocolitis or isolated intestinal perforation: a

prospective cohort study by the NICHD Neonatal

Research Network. Ann Surg. 2005 Jun;241(6):984-9;

discussion 9-94.

74. Hintz SR, Kendrick DE, Stoll BJ, Vohr BR, Fanaroff AA,

Donovan EF, et al. Neurodevelopmental and growth

outcomes of extremely low birth weight infants after

necrotizing enterocolitis. Pediatrics. 2005

Mar;115(3):696-703.

75. Blumenthal I, Lealman GT. Effect of posture on gastro-

oesophageal reflux in the newborn. Archives of Disease in

Childhood. 1982 Jul;57(7):555-6.

76. Cobb BA, Carlo WA, Ambalavanan N. Gastric residuals

and their relationship to necrotizing enterocolitis in very low birth weight infants. Pediatrics. 2004 Jan;113(1 Part

1):50-3.

77. Mihatsch WA, von Schoenaich P, Fahnenstich H, Dehne N,

Ebbecke H, Plath C, et al. The significance of gastric

residuals in the early enteral feeding advancement of

extremely low birth weight infants.. Pediatrics. 2002

Mar;109(3):457-9.

78. Bhatia P, Johnson KJ, Bell EF. Variability of abdominal

circumference of premature infants. J Pediatr Surg. 1990 May;25(5):543-4.

79. Rowe MI, Reblock KK, Kurkchubasche AG, Healey PJ.

Necrotizing enterocolitis in the extremely low birth weight

infant. J Pediatr Surg. 1994 Aug;29(8):987-90;

discussion 90-1.

80. Singh SA, Dutta S, Narang A. Predictive clinical scores for

diagnosis of late onset neonatal ehaviour a. J Trop

Pediatr. 2003 Aug;49(4):235-9.

81. Jaile JC, Levin T, Wung JT, Abramson SJ, Ruzal-Shapiro

C, Berdon WE. Benign gaseous distension of the bowel in premature infants treated with nasal continuous airway

pressure: a study of contributing factors. AJR American

Journal of Roentgenology. 1992 Jan;158(1):125-7.


References 91

82. Henderson G, Anthony MY, McGuire W. Formula milk

versus preterm human milk for feeding preterm or low

birth weight infants. Cochrane Libr. 2006;1.

83. Henderson G, Anthony MY, McGuire W. Formula milk

versus term human milk for feeding preterm or low birth

weight infants. Cochrane Libr. 2006;1.

84. Schanler RJ, Lau C, Hurst NM, Smith EO. Randomized

trial of donor human milk versus preterm formula as

substitutes for mothers‘ own milk in the feeding of

extremely premature infants. Pediatrics. 2005 Aug;116(2):400-6.

85. Kuschel CA, Harding JE. Multicomponent fortified human

milk for promoting growth in preterm infants. Cochrane

Database of Systematic Reviews. 2004(1):CD000343.

86. World Health Organisation. Nutrition Infant and Young

Child: Exclusive Breastfeeding. Available at:

(http://www.who.int/child-adolescent-

health/NUTRITION/infant_exclusive.htm).

87. NHMRC. Dietary Guidelines for Children and Adolescents in Australia incorporating the Infant Feeding Guidelines

for Health Workers. 2003. NHMRC. Available at:

http://www.nhmrc.gov.au/publications/synopses/dietsyn.

htm

88. The International Baby Food Action Network. The

International Code of Marketing of Breastmilk

Substitutes. WHO. Available at:

http://www.ibfan.org/english/resource/who/fullcode.html

89. Herrell N, Martin RJ, Fanaroff A. Arterial oxygen tension

during nasogastric feeding in the preterm infant. J Pediatr. 1980 May;96(5):914-6.

90. Narchi H, Kulaylat N. Is gastric lavage needed in

neonates with meconium-stained amniotic fluid? Eur J

Pediatr. 1999 Apr;158(4):315-7.

91. Snell BJ. Physiologic response of the preterm infant

during the early initiation of breastfeeding versus

bottlefeeding; 1991.

92. Chen CH, Wang TM, Chang HM, Chi CS. The effect of

breast- and bottle-feeding on oxygen saturation and body temperature in preterm infants. Journal of Human

Lactation. 2000 Feb;16(1):21-7.

93. Dowling DA. Physiological responses of preterm infants to

breast-feeding and bottle-feeding with the orthodontic

nipple. Nursing Research. 1999 Mar-Apr;48(2):78-85.

94. Meier P. Bottle- and breast-feeding: effects on

transcutaneous oxygen pressure and temperature in

preterm infants. Nursing Research. 1988 Jan-

Feb;37(1):36-41.

95. Blaymore Bier JA, Ferguson AE, Morales Y, Liebling JA, Oh

W, Vohr BR. Breastfeeding infants who were extremely

low birth weight. Pediatrics. 1997 Dec;100(6):E3.

96. Blaymore Bier J, Ferguson A, Anderson L, Solomon E,

Voltas C, Oh W, et al. Breast-feeding of very low birth

weight infants. J Pediatr. 1993 Nov;123(5):773-8.

97. Hauck FR, Omojokun OO, Siadaty MS. Do pacifiers reduce

the risk of sudden infant death syndrome? A meta-

analysis. Pediatrics. 2005 Nov;116(5):e716-23.

98. Li DK, Willinger M, Petitti DB, Odouli R, Liu L, Hoffman HJ. Use of a dummy (pacifier) during sleep and risk of

sudden infant death syndrome (SIDS): population based

case-control study. Bmj. 2006 Jan 7;332(7532):18-22.

99. Ludington-Hoe SM, Anderson GC, Swinth JY, Thompson

C, Hadeed AJ. Randomized controlled trial of kangaroo

care: cardiorespiratory and thermal effects on healthy

preterm infants. Neonat Netw. 2004 May-Jun;23(3):39-

48.

100. Ibe OE, Austin T, Sullivan K, Fabanwo O, Disu E, Costello AM. A comparison of kangaroo mother care and

conventional incubator care for thermal regulation of

infants < 2000 g in Nigeria using continuous ambulatory

temperature monitoring. Annals of Tropical Paediatrics.

2004 Sep;24(3):245-51.

101. Ludington-Hoe SM, Johnson MW, Morgan K, Lewis T,

Gutman J, Wilson PD, et al. Neurophysiologic assessment

of neonatal sleep organization: preliminary results of a

randomized, controlled trial of skin contact with preterm

infants. Pediatrics. 2006 May;117(5):e909-23.

102. Feldman R, Eidelman AI, Sirota L, Weller A. Comparison

of skin-to-skin (kangaroo) and traditional care: parenting

outcomes and preterm infant development. Pediatrics.

2002 Jul;110(1 Pt 1):16-26.

103. Bergman NJ, Linley LL, Fawcus SR. Randomized

controlled trial of skin-to-skin contact from birth versus

conventional incubator for physiological stabilization in

1200- to 2199-gram newborns.. Acta Paediatr. 2004

Jun;93(6):779-85.

104. McCall EM, Alderdice FA, Halliday HL, Jenkins JG, Vohra

S. Interventions to prevent hypothermia at birth in

preterm and/or low birthweight babies. Cochrane Libr.

2006;1.

105. Chwo MJ, Anderson GC, Good M, Dowling DA, Shiau SH, Chu DM. A randomized controlled trial of early kangaroo

care for preterm infants: effects on temperature, weight,

ehaviour, and acuity. Journal of Nursing Research: JNR.

2002 Jun;10(2):129-42.

106. Dodd VL. Effects of Kangaroo Care in preterm infants;

2003. PhD Thesis University of Connecticut

107. Rojas MA, Kaplan M, Quevedo M, Sherwonit E, Foster LB,

Ehrenkranz RA, et al. Somatic growth of preterm infants

during skin-to-skin care versus traditional holding: a randomized, controlled trial. Journal of Developmental &

Behavioral Pediatrics. 2003 Jun;24(3):163-8.

108. Kadam S, Binoy S, Kanbur W, Mondkar JA, Fernandez A.

Feasibility of kangaroo mother care in Mumbai. Indian

Journal of Pediatrics. 2005 Jan;72(1):35-8.

109. Meier PP, Lysakowski TY, Engstrom JL, Kavanaugh KL,

Mangurten HH. The accuracy of test weighing for preterm

infants. J Pediatr Gastroenterol Nutr. 1990 Jan;10(1):62-

5.

110. Meier PP, Engstrom JL, Fleming BA, Streeter PL, Lawrence PB. Estimating milk intake of hospitalized

preterm infants who breastfeed. Journal of Human

Lactation. 1996 Mar;12(1):21-6.

111. Collins CT. Does the use of artificial teats (dummy or

bottle) affect breastfeeding success in preterm infants? A

randomised controlled trial and systematic review.

Adelaide: University of Adelaide; 2003.

112. Gilks J, Watkinson M. Improving breastfeeding rates in

preterm babies: cup feeding versus bottle feeding. Journal of Neonatal Nursing. 2004 Jul;10(4):118-20.

113. Hurst NM, Valentine CJ, Renfro L, Burns P, Ferlic L. Skin-

to-skin holding in the neonatal intensive care unit

influences maternal milk volume. J Perinatol. 1997 May-

Jun;17(3):213-7.

114. Shiao SY, Youngblut JM, Anderson GC, DiFiore JM, Martin

RJ. Nasogastric tube placement: effects on breathing and

sucking in very-low-birth-weight infants. Nurs Res. 1995

Mar-Apr;44(2):82-8.

115. Bier JA, Ferguson AE, Morales Y, Liebling JA, Archer D,

Oh W, et al. Comparison of skin-to-skin contact with

standard contact in low-birth-weight infants who are

breast-fed. Archives of Pediatrics & Adolescent Medicine.

1996 Dec;150(12):1265-9.

116. Whitelaw A, Heisterkamp G, Sleath K, Acolet D, Richards

M. Skin to skin contact for very low birthweight infants

and their mothers. Archives of Disease in Childhood.

1988 Nov;63(11):1377-81.

117. Mathew OP. Breathing patterns of preterm infants during bottle feeding: role of milk flow. J Pediatr. 1991

Dec;119(6):960-5.

118. Mathew OP, Belan M, Thoppil CK. Sucking patterns of

neonates during bottle feeding: comparison of different

nipple units. American Journal of Perinatology. 1992

Jul;9(4):265-9.

119. Scheel CE, Schanler RJ, Lau C. Does the choice of bottle

nipple affect the oral feeding performance of very-low-

birthweight (VLBW) infants? Acta Paediatr. 2005 Sep;94(9):1266-72.

120. Whitfield MF, Kay R, Stevens S. Validity of routine clinical

test weighing as a measure of the intake of breast-fed

infants. Arch Dis Child. 1981 Dec;56(12):919-21. Also

reported in Stevens S and Whitfield MF. How accurate is

clinical test weighing of the newborn? Midwives Chronicle

& Nursing Notes. 1980 May:148-9

121. Dollberg S, Lahav S, Mimouni FB. A comparison of

intakes of breast-fed and bottle-fed infants during the

first two days of life. J Am Coll Nutr. 2001 Jun;20(3):209-11.

122. Evans KC, Evans RG, Royal R, Esterman AJ, James SL.

Effect of caesarean section on breast milk transfer to the

http://www/

http://www/

http://www.ibfan.org/english/resource/who/fullcode.html


References 92

normal term newborn over the first week of life. Arch Dis

Child Fetal Neonatal Ed. 2003 Sep;88(5):F380-2.

123. Allen, K. Test weighing of healthy, full-term newborns.

Centre for Clinical Effectiveness, 2007.

Annexes 93

Annex 1 Acronyms and Abbreviations

95%CI 95% confidence interval

AGA Appropriate for gestational age (usually refers to weight)

ARD Acute respiratory distress

AREDF Absent or reversed end-diastolic flow

CLD Chronic lung disease

CPAP Continuous positive airway pressure

BPD Bronchopulmonary dysplasia

EBM Expressed breastmilk

EBP Evidence-based practice

ECMO Extracorporeal membrane oxygenation

ELBW Extremely low birth weight (<1000g)

GA Gestational age

GDG Guideline development group

IDM Infants of diabetic mothers

IUGR Intra-uterine growth restriction

IV Intravenous

IVH Intraventricular haemorrhage

KMC Kangaroo mother care

LBW Low birth weight

LOS Length of stay

MD Mean difference

MEN Minimal enteral nutrition (also known as trophic feeding)

NEC Necrotising enterocolitis

NG Nasogastric

NICU Neonatal Intensive Care Unit

NNS Non-nutritive sucking

PCA Post-conceptual age

PMA Post menstrual age

RCT Randomised controlled trial

RDS Respiratory distress syndrome

RR Relative risk

SD Standard deviation

SSB Suck, swallow, breathe

TcCO2 Transcutaneous carbon dioxide tension

VLBW Very low birth weight (<1500g)

WMD Weighted mean difference

Annexes 94

Annex 2 Definitions

Breast contact Early contact between the infant and the mother‘s breast. Aims to

provide early support for breastfeeding. Infant is usually placed in

the mother‘s arms, in the breastfeeding position, against the

mother‘s skin.

Feed(ing) intolerance Intolerance to enteral feeds, assessed by a range of measures

including vomiting (volume and frequency), abdominal distension,

gastric aspirates (volume or appearance). See Chapter 4 for more

details.

Kangaroo (mother)

care

Early contact between the infant and the mother‘s skin. Aims to

improve growth and development. Infant is usually placed on the

mother‘s chest, lying between the mother‘s breasts, head under the

mother‘s chin (also known as skin-to-skin care).

Minimal enteral

nutrition

Small volume, non-nutritional enteral feeds given to enhance

maturation of the gastrointestinal tract (also known as trophic

feeding).

Near-term infants Infants born at 35 weeks gestation or greater.

Necrotis(z)ing

enterocolitis

A condition resulting in intestinal inflammation and necrosis, most

commonly in the terminal ileum and ascending colon.

Non-nutritive sucking Sucking on a dummy or other object in which no milk is transferred

to promote development of sucking skills.

Posset An effortless vomit usually during or immediately after a feed.

Rooting A reflex action in the infant to turn the head towards stimulus on the

cheek or side of the mouth in order to seek out a nipple.

Senior clinician A member of medical staff at fellow level or above, or a member of

nursing staff at Associate Nurse Unit Manager level or above.

Skin-to-skin care Early contact between the infant and the mother‘s skin. Aims to

improve growth and development. Infant is usually placed on the

mother‘s chest, lying between the mother‘s breasts, head under the

mother‘s chin (also known as kangaroo care).

Trophic feeding Small volume, non-nutritional enteral feeds given to enhance

maturation of the gastrointestinal tract (also known as minimal

enteral nutrition).

Date post:	28-Jan-2023
Category:	Documents
Upload:	uba
View:	1 times
Download:	0 times

929 Feeding guideline

Documents