Chapter 7 Review of breastfeeding assessment toolstexastenstep.org/.../12/review-of-breastfeeding-assessment-tools.pdf · 7.2 What should a breastfeeding assessment tool include?

138

Chapter 7 Review of breastfeeding assessment tools

Chapter 7 Review of breastfeeding assessment tools

Chapter 7: Review of breastfeeding assessment tools

Many malnourished infants present with reports of breastfeeding problems (Chapter 6) and exclusive breastfeeding in infants <6m is the treatment goal of most guidelines (Chapter 4). Therefore breastfeedingassessment to identify feeding problems and monitor recovery is an important part of MAMI. Since most current SAM guidelines do not include or refer to speci!c breastfeeding assessment tools, this chapter reviews those tools currently available and their potential application to the MAMI contextiii.

7.1 Scope of reviewExisting reviews, focused on developed countries, have identi!ed a number of breastfeeding assessment tools with none clearly identi!ed as a gold standard.153 This chapter identi!es areas of consensus between tools, gaps in coverage and advantages and disadvantages in di"erent settings. To do so, a framework is used that identi!es three aspects or settings of severe acute malnutrition (SAM) care: community screening/case detection, case !nding in clinical contexts, and diagnosis in clinical contexts.154 We conceive that there is a gradient of importance of di"erent tool characteristics from community to clinical contexts.# For example a community-health worker (CHW) with responsibility for treating malnourished infants would want a highly sensitive tool to identify women experiencing breastfeeding problems, but a clinician diagnosing a reported breastfeeding problem in a hospital-setting would bene!t from using a tool with higher speci!city so that ‘false positives’ did not result in unnecessary inpatient admission.#

Objectives of the review are as follows:

· To identify published breastfeeding assessment tools, and document which breastfeeding domains arecovered and which omitted (adapted from Hall Moran, 2000155).

· To consider tool value in emergency settings and other nutritionally vulnerable contexts, looking at performance characteristics for identifying breastfeeding problems in community and clinical settings (adapted from Myatt et al, 2006156).

· To grade the tools based on the quality and level of evidence underlying their development and validation.

When talking about ‘case-!nding’, we envisage that the tools would be used in combination with anthropometric assessment. Exact use will depend on context. In MAMI, their main importance is identifying the immediate cause of poor anthropometric status. The better the diagnosis, the faster and better the patient can be directed to treatment services. For example, a young infant may be identi!ed in the community as having a simple breastfeeding problem due to poor positioning. He/she might bene!t from early referral to a breastfeeding support group before severe malnutrition and clinical complications develop. Another infant may have a low weight-for-length but no identi!ed breastfeeding problem; in thiscase clinical referral and assessment is needed. Other approaches, such as establishing HIV status, should be considered.# Maternal factors may also determine breastfeeding ‘success’ and other strategies will be needed as a result, e.g. a severely malnourished mother who is exclusively breastfeeding will require nutritional rehabilitation of the mother to enable adequate nourishment of her infant; children of HIV-infected mothers who are ill or die are more likely to die themselves, independent of the HIV status of the infant157.

iii This review was carried out as an additional activity beyond the scope of the original MAMI Project, undertaken by Jenny Saxton and supported by the MAMI Research Team. We acknowledge valuable feedback from Felicity Savage and Ann Ashworth on content, to inform this and future write-up of this work.

139

Management of Acute Malnutrition in Infants (MAMI) Project

7.2 What should a breastfeeding assessment tool include?

7.2 What should a breastfeeding assessment tool include?E"ective breastfeeding can be divided into several essential elements that are important for overall success. One study that critically analysed six breastfeeding assessment tools identi!ed the following elements: baby’s behaviour, mother’s behaviour, positioning, attachment, e"ective feeding (e.g. audible swallow), health of the breast, health of the baby and mothers’ perception of the breastfeeding experience.158 Other elements, such as severe maternal wasting and maternal HIV status, are important considerations in emergency settings. Determining the main issues for an individual mother/infant dyad matters: more precise diagnoses will direct infants and their carers to more tailored and ultimately more e"ective treatment. Simple issues can be pinpointed and rapidly resolved (e.g. attachment di$culty). More complex issues can be referred for further care (e.g. breastfeeding support alone is unlikely to addressmaternal depression). Good assessment can also help rule out breastfeeding problems as the main cause ofinfant SAM (e.g. an infant may be breastfeeding well yet be malnourished due to underlying HIV disease).

Table 36: Importance gradient of key properties of breastfeeding assessment tools fromcommunity to clinical contexts

Tool Property Context (score)a

Reasons for gradingCommunityscreening/casedetection

Case!ndingin clinicalcontexts

Diagnosisin clinicalcontexts

Simplicity (forCHWsb &doctors)

3 2 1 Simplicity is important in community settings as CHWs havelower education/training. Simplicity is less important in clinicalcontexts unless there is a high caseload & limited time/resources.

Acceptability (bymothers)

3 2 1 BFc assessment by definition is intimate. Male CHWs orinsufficient privacy/comfort are likely to be unacceptable. Sometools rely on questions alone, not assessment, and are likely tobe more acceptable to mothers.

Cost/Time 3 2 1 Resource-poor settings would demand that assessment be cost& time-effective in all contexts, although at diagnosis there ispotentially more time to spend identifying specific problems.

Objectivity 1 2 3 Most tools are observational & will be subjective. Objectivity maybe more crucial in diagnostic settings if decisions aboutadmission are being made.

Quantitativeness 1 2 3 Assessment scores give useful indication of risk (e.g. BFcessation). A total score is less useful for diagnosis as specificproblems would be masked, but may be helpful in makingdecisions about admission. Some scoring systems are arbitrary& could detract from finding ways to overcome specific problems.

Precision(reliability)

1 2 3 Important that a tool can pick out specific BF problems in areliable consistent way, although CHWs may just refer all SAMinfants with a breastfeeding problem to more specialised care, orfor specific diagnosis of a breastfeeding problem.

Accuracy(Validity)

1 2 3 Very important to be able to pinpoint which aspects of BF arecausing problems in order to provide simple solutions. Timing offeed & emotional state could affect the ability to pinpoint aproblem if only one assessment is made.

Sensitivity(Identifying truecases)

3 2 1 Sensitivity is less important in clinical contexts because theinfant is already present at hospital.

Specificity(identifying truenegatives)

1 2 3 False positives' could overload the health system; make mothersfeel inadequate & unlikely to resume BF. Unnecessary inpatientcare could increase risk of infection, & may cause mother toleave other children unattended, or lose money through notbeing able to go to work.

Predictive value 1 2 3 Important that assessments predict infant nutritional outcomes &preventable mortality. For a CHW involved in referral to morespecialist BF services, predictive value is less important, betterthat they identify as many potential true cases as possible torefer for further help.

a Scoring: 1=least important 2=medium importance 3=most importantb Community Health Worker; c Breastfeeding

Table and scoring system adapted from Myatt et al, 2006159

140


7.3 Methods

7.3 Methods

7.3.1 Literature search strategyPubmed was systematically searched for articles with no date parameters set using AND/OR combinationsof the following key words and MeSH (medical subject headings) terms:

• Breastfeeding, breastfeeding problems, breastfeeding technique, breastfeeding performance assessment,breastfeeding skill, breastfeeding ability, breastfeeding assessment tool, breast milk, human, actual or perceived milk insu$ciency, insu$cient milk supply, latch, root, suck, swallow, position, milk transfer.

• Needs assessment, risk assessment, nutrition assessment, psychometrics, nursing diagnosis, index, checklist, questionnaire, diagnostic tools, guidelines.#

• Infant nutrition disorders, infant nutritional physiological phenomena, infant very low or low birth weight, infant welfare, infant behaviour, infant premature, infant postmature, infant newborn, infant small for gestational age, infant diseases, infant care, infant growth,# developing countries.

Reference lists of breastfeeding assessment tools !tting the inclusion criteria were searched for further relevant tools.# Due to the small number of tools identi!ed that were designed for use in developing countries and emergency contexts, a purposive search of WHO/UNICEF and initiatives on infant and young child feeding in emergencies (IFE) in the emergency nutrition sector was conducted.# A secondary search was conducted to identify any studies attempting to validate or examine the underlying propertiesand predictive value of these tools.

7.3.2 Tool inclusion/exclusion criteriaCriteria for inclusion in the current review were articles written in English that described the development of breastfeeding tools applied in clinical or community contexts that were designed to assess breastfeeding performance of the mother-infant dyad through observations or questionnaires. Tools were excluded if they were designed to assess bottle-feeding or breastfeeding problems of women who have undergone breast surgery. Tools using complex assessment methods (e.g. analysis of feeding behaviour with direct linear transformation160) were also excluded on the basis that these would be unavailable in developing country settings. Single item and survey questions seeking population-level data about infant feeding were not considered to function as diagnostic tools and were also excluded. Tools assessing frenulum function were considered beyond the scope of this review. Maternal self-reported questionnaires assessing maternal perceptions of breastfeeding are summarised in this paper but not explored in detail.

7.4 Results and DiscussionFifteen tools were identi!ed in the literature search.# The Lactation Assessment Tool (LAT™) !tted our inclusion criteria, however we were unable to obtain su$cient information about the development and validation of the tool to include it in the !nal review. Generally, the methodological quality of studies was suboptimal. At best, tools were tested in observational studies; at worst tools were not tested at all. There was some consensus between tools about breastfeeding domains for inclusion in assessment, but few tools achieved comprehensive coverage and particular domains were frequently neglected. Most were designed for newborn infants, often preterm, whose breastfeeding problems are likely to di"er from thoseof SAM infants. In terms of suitability of tools for di"erent contexts, no single tool would be suitable for use without adaptation.

7.4.1 Development and validation of toolsThe majority of tools (9/15) have not been tested against short or long-term nutritional or breastfeeding outcomes and it is therefore di$cult to make recommendations for their use. Five were developed purely from existing literature and clinical experience, whilst a further three have been internally assessed for inter-rate consistency and test re-test reliability with results ranging from poor to excellent (Table 37). Overall, just three tools were designed for use by CHWs in developing settings and the majority were developed for North American neonatal populations for use by highly trained health professionals (e.g. NOMAS), although some were also used by mothers (e.g. LATCH).# None of the tools was developed for or tested on acutely malnourished infants.141


7.4 Results and Discussion

Tab

le 3

7: T

heo

ry-b

ased

bre

astf

eed

ing

ass

essm

ent

too

ls

BF

To

ol*

Au

tho

r(s)

& D

ate

Too

l Des

crip

tio

nC

ou

ntr

y,se

ttin

gS

amp

lesi

ze &

par

tici

-p

ants

Infa

nt

age

(mea

n,

sd)

Infa

nt/

mat

ern

alo

utc

om

es

Sta

tist

ical

An

alys

isR

esu

lts

Su

mm

ary

Au

tho

rC

om

men

ts/n

ote

s

Ou

r co

mm

ents

IFE

Mo

du

le 2

-(S

RA

a & F

Ab

of

BF

c ).

IFE

Co

reG

rou

p(2

004)

d

SR

A: A

ge a

ppro

pria

te fe

edin

g, B

Fea

se, b

aby'

s co

nditi

on. R

efer

prob

lem

s fo

r FA

: obs

erve

:at

tach

men

t, su

cklin

g, m

othe

r’sco

nfid

ence

, fee

d en

d. L

iste

n/le

arn

from

mot

her

re B

Fpr

actic

es/b

elie

fs/w

orrie

s. O

bser

vear

tific

ial f

eed

if re

leva

nt.

Trai

ning

mat

eria

l dev

elop

ed o

n th

e ba

sis

of 'e

xist

ing

best

pra

ctic

e &

publ

ishe

d ev

iden

ce w

here

it e

xist

s. W

here

it d

oes

not i

t dra

ws

upon

exte

nsiv

e ex

perie

nce

& a

bro

ad b

ase

of e

xper

t opi

nion

.' B

Fas

sess

men

t sec

tions

bas

ed e

xten

sive

ly o

n th

e W

HO

40

hour

brea

stfe

edin

g co

unse

lling

cou

rse

(200

4).

Fie

ld w

orke

rssh

ould

syst

emat

ical

lyre

cord

&id

entif

y m

ost

effe

ctiv

em

etho

ds.

Req

uire

s m

odul

e 1

trai

ning

(1-

3 ho

urs)

&m

odul

e 2

(5 h

ours

).D

evel

oped

for

CH

Wse

in e

mer

genc

y/de

velo

ping

cou

ntry

setti

ngs.

Mo

ther

-in

fan

tB

F a

sses

smen

tto

ol

Joh

nso

net

al,

1999

161

Mot

her

& in

fant

sco

red

on 8

item

sto

indi

cate

BF

failu

re r

isk:

latc

h (2

item

s), s

uck,

nip

ple

type

,fr

eque

ncy

of n

ursi

ng/w

et n

appi

es,

prev

ious

suc

cess

w/ B

F,su

ppor

tive

part

ner.

US

A,

hom

evi

sit

n=98

1;no

tst

ated

Not

stat

edR

e-ad

mis

sion

rate

w/ B

Fpr

oble

ms.

Com

pare

d %

re

adm

issi

ons

w/ B

Fpr

oble

ms

ofw

omen

rece

ivin

g ho

me

visi

ts v

s no

t.

Re-

adm

issi

onra

te h

ighe

rif

no h

ome

visi

t was

mad

e.

Tool

is a

usef

ul g

uide

for

novi

cenu

rses

.

A B

F E

valu

atio

n&

Ed

uca

tio

nTo

ol (

BE

ET

)

Tob

in,

1996

162

8 su

b-sc

ales

: Fee

ding

s,po

sitio

ning

, lat

ch, s

uck,

milk

flow

,in

take

, out

put,

wei

ght g

ain.

Tool

bas

ed o

n 2

publ

ishe

d &

4 u

npub

lishe

d B

F r

efer

ence

s. N

ore

fere

nce

to c

linic

al e

xper

ienc

e or

atte

mpt

ed v

alid

atio

n of

tool

aga

inst

mea

sura

ble

outc

omes

, or

inte

rnal

rel

iabi

lity.

Can

use

tool

pren

atal

ly, b

yph

one

or in

hosp

ital.

Rat

iona

le fo

r ite

ms

disc

usse

d bu

t not

test

ed a

gain

stm

easu

rabl

eou

tcom

es.

Mo

ther

-Bab

yA

sses

smen

t(M

BA

)f

Mu

lfo

rd,

1992

163

5 st

eps

in B

F a

sses

sed:

sig

nalli

ng,

posi

tioni

ng; f

ixin

g; m

ilk tr

ansf

er;

endi

ng.

Sco

res

for

mot

her

&in

fant

sep

arat

ely.

US

A, n

otst

ated

Lim

ited

info

rmat

ion

give

n re

evi

denc

e M

BA

item

s ba

sed

on.

Pos

ition

ing,

fixi

ng &

milk

tran

sfer

item

s fr

om p

ublis

hed

wor

kde

scrib

ing

com

mon

sig

ns.

Tool

not

val

idat

ed in

this

stu

dy,

alth

ough

was

in a

furt

her

stud

y.

Tool

trac

ksB

F in

hos

pita

ltr

iage

/ref

erra

lor

res

earc

h

Sys

tem

atic

Ass

essm

ent

of

the

Infa

nt

at t

he

Bre

ast

(SA

IB)

Sh

rag

o &

Bo

car,

1990

164

Obs

erva

tion

of: a

lignm

ent;

areo

lar

gras

p; a

reol

ar c

ompr

essi

on;

audi

ble

swal

low

ing.

No

scor

ing

syst

em.

US

A, n

otst

ated

Inst

rum

ent b

ased

on

scie

ntifi

c un

ders

tand

ing

of la

ctat

ion

&cl

inic

al e

xper

ienc

e. N

ot p

sych

omet

rical

ly te

sted

or

valid

ated

aga

inst

mat

erna

l or

infa

nt n

utrit

iona

l or

BF

outc

omes

.

142



BF

To

ol*

Au

tho

r(s)

& D

ate

Too

l Des

crip

tio

nC

ou

ntr

y,se

ttin

gS

amp

lesi

ze &

par

tici

-p

ants

Infa

nt

age

(mea

n,

sd)

Infa

nt/

mat

ern

alo

utc

om

es

Sta

tist

ical

An

alys

isR

esu

lts

Su

mm

ary

Au

tho

rC

om

men

ts/n

ote

s

Ou

rco

mm

ents

UN

ICE

F b

-r-e

-a-

s-t

ob

serv

atio

nal

chec

klis

t

WH

O/

UN

ICE

F,19

9316

5

27 it

ems/

6 sc

ales

: sig

nsth

at B

F g

oing

wel

l vs

poss

ible

diff

icul

ty: B

ody

posi

tion,

Res

pons

es,

Em

otio

nal B

ondi

ng,

Ana

tom

y, S

uckl

ing,

Tim

esu

cklin

g.

Not

sta

ted,

alth

ough

des

igne

d fo

r C

HW

w/ W

HO

/UN

ICE

F B

F tr

aini

ng.

Non

e st

ated

Tool

cov

ers

wid

e ra

nge

ofB

F d

omai

ns,

but n

o ev

iden

ceof

dev

elop

men

t/va

lidat

ion

agai

nst e

xter

nal

outc

omes

.

Mo

ther

-In

fan

tB

reas

tfee

din

gP

rog

ress

To

ol

(MIB

PT

)

Joh

nso

net

al,

2007

166

8 ite

ms

obse

rve:

resp

onsi

vene

ss to

feed

ing

cues

; tim

ing

of fe

eds;

nutr

itive

suc

klin

g;po

sitio

ning

/latc

hing

fact

ors;

nipp

le tr

aum

a, in

fant

beha

viou

r st

ate

& m

othe

r/pa

rent

res

pons

e to

infa

nt.

US

A,

hosp

ital

n=62

;H

ealth

ym

othe

r&

baby

; 35-

42 w

eeks

GA

7.

2 ho

urs

to 5

days

old

n/a

% a

gree

men

tbe

twee

n ra

ters

Inte

r-ra

ter

agre

emen

t 79-

95%

.

Cau

tion

ifre

lyin

g up

onsi

ngle

sco

res

beca

use

mot

her/

infa

ntpr

ogre

ss w

ithB

F o

ver

time.

Rat

iona

le fo

rite

ms

disc

usse

dbu

t not

test

edag

ains

tnu

triti

onal

/BF

outc

omes

.

Pre

-ter

m In

fan

tB

reas

tfee

din

gB

ehav

iou

rS

cale

(P

IBB

S)

Nyq

vist

et

al, 1

99616

76

item

s as

sess

pre

-ter

min

fant

s on

: roo

ting,

are

olar

gras

p; d

urat

ion

of la

tch;

suck

ing;

long

est s

ucki

ngbu

rst;

swal

low

ing.

Sw

eden

,ho

spita

ln=

24;

Ful

l/pre

term

infa

nts

inne

onat

alin

tens

ive

care

,tr

ansi

tiona

l/m

ater

nity

units

.

Not

stat

edn/

aIn

ter-

rate

r re

liabi

lity

of o

bser

vers

, &ob

serv

ers/

mot

hers

. Unc

lear

anal

ysis

test

ing

tool

abi

lity

to d

etec

tdi

ffere

ntG

A/m

atur

ity o

fin

fant

BF.

Goo

d in

ter-

rate

rre

liabi

lity

for

obse

rver

s(0

.64-

1.00

), b

utpo

or fo

rob

serv

ers

vsm

othe

rs (

0.27

-0.

86).

Poo

rer

item

s re

vise

d.

BF

Su

pp

ort

gu

idel

ines

fo

r a

Bab

y- F

rien

dly

Ho

spit

al:

BF

Ob

serv

atio

nA

id

UN

ICE

F,20

0616

821

dic

hoto

mou

s ite

ms:

sig

nsth

at B

F g

oing

wel

l vs

poss

ible

diff

icul

ty. G

ener

alite

ms

and

item

s as

sess

ing:

brea

st c

ondi

tion,

posi

tioni

ng, a

ttach

men

t,su

cklin

g.

Mat

eria

lfie

ld te

sted

in Zim

babw

e

Not

sta

ted

Non

e st

ated

Tool

is r

evis

edU

NIC

EF

chec

klis

t.P

ilotin

g in

Sie

rra

Leon

eea

rly 2

009.

Table 37 cont’d

* I

ncl

udes

too

ls d

evel

oped

on t

he

bas

is o

f ex

per

t op

inio

n a

nd liter

ature

sea

rchin

g;

tool

s m

ay h

ave

bee

n ‘in

tern

ally

’ va

lidat

ed

143



Tab

le 3

8: B

reas

tfee

din

g a

sses

smen

t to

ols

val

idat

ed a

gai

nst

mat

ern

al a

nd

infa

nt

ou

tco

mes

BF

To

ol*

Au

tho

r(s)

& D

ate

Too

l Des

crip

tio

nC

ou

ntr

y,se

ttin

gS

amp

le s

ize

& par

tici

pan

ts

Infa

nt

age

(mea

n,

sd)

Infa

nt/

mat

ern

alo

utc

om

es

Sta

tist

ical

An

alys

isR

esu

lts

Su

mm

ary

Au

tho

rC

om

men

ts/n

ote

s

BF

a

Ass

essm

ent

Sco

re (

BA

S)

Hal

l et

al,

2002

169

Ris

k of

BF

ces

satio

n(f

irst 7

-10

days

). 5

item

s:m

othe

r’s a

ge, p

revi

ous

BF

exp

erie

nce,

latc

hing

prob

lem

s, B

F in

terv

al,

form

ula

bottl

es. 2

ext

raite

ms:

bre

ast s

urge

ry,

hype

rten

sion

, vac

uum

vagi

nal d

eliv

ery.

US

A,

hosp

ital

n=11

08; N

oin

form

atio

n on

mat

erna

l or

infa

nt h

ealth

indi

cato

rs.

Mea

n 40

hour

s(s

d=13

)

BF

cess

atio

nra

te 7

-10

days

afte

rbi

rth

Sin

gle

varia

ble

logi

stic

regr

essi

ons

iden

tifie

dca

ndid

ate

pred

icto

rs.

Pre

dict

ors

ente

red

into

mul

tiple

logi

stic

mod

el;

eval

uate

d w

ith O

Rb

& R

Rc

to g

ive

optim

al R

OC

d

10.5

% o

f mot

hers

repo

rted

ces

satio

n of

BF

;A

ll B

AS

item

s si

ge

pred

icte

d B

F c

essa

tion.

Gia

nn

i et

al, 2

00617

0

Use

fuln

ess

of a

nas

sess

men

t sco

re to

pred

ict e

arly

EB

Ff

cess

atio

n.

Italy

,ho

spita

ln=

175;

Mot

hers

of

heal

thy

EB

Fin

fant

s. B

irth

wei

ght

!250

0g, G

Ag

37-4

2 w

eeks

.

Not

repo

rted

.M

ean

hosp

ital

stay

2.2

days

BF

cess

atio

n,in

trod

uctio

nof

com

plem

-en

tary

feed

ing,

cont

inue

dE

BF

at 1

mon

th

Chi

2&

Man

n-W

hitn

ey U

Test

s to

test

for

diffe

renc

es in

BA

S s

core

sby

BF

beh

avio

ur 1

mon

thpo

stpa

rtum

.

Bas

elin

e sc

ores

of E

BF

mot

hers

at 1

mon

th s

iglo

wer

than

non

-EB

F. N

Sh

diffe

renc

e if

com

plem

-en

ting

BF

at 1

mon

th v

sno

BF.

Lat

chin

g pr

oble

ms

& n

o pr

ior

BF

suc

cess

nega

tivel

y as

soci

ated

w/

BF

dur

atio

n.

Infa

nt

Bre

astf

eed

ing

Ass

essm

ent

Too

l (IB

FAT

)

Mat

thew

s,19

8817

16

item

s m

easu

re 4

infa

nt b

ehav

iour

s:re

adin

ess

to fe

ed,

root

ing,

fixi

ng &

suc

king

.Tw

o no

n-sc

orin

g ite

ms:

infa

nt s

tate

& m

ater

nal

satis

fact

ion

w/ B

F.

Can

ada,

hosp

ital

n=60

;S

pont

aneo

usde

liver

y.A

PG

AR

i sco

re!8

afte

r 5

min

s;ap

prop

riate

wei

ght f

or G

A.

Ear

lyne

onat

alpe

riod'

BF

sta

tus

at4

wee

ksin

ter-

rate

r re

liabi

lity

(%ag

reem

ent)

IBFA

T s

core

s di

d no

tpr

edic

t BF

sta

tus

at 4

wee

ks. L

imite

d va

riabi

lity

as 8

0% s

till B

F. In

ter-

rate

r ag

reem

ent 9

1%.

IBFA

T s

core

s di

dno

t pre

dict

BF

stat

us a

t 4w

eeks

. Lim

ited

varia

bilit

y as

80%

still

BF.

Inte

r-ra

ter

agre

emen

t 91%

.

Fu

rman

,20

0617

2E

valu

atio

n of

VLB

Wj :

can

we

use

the

IBFA

T?

US

A,

hosp

ital

n=34

; Mot

hers

of V

LBW

infa

nts

35 w

eeks

corr

ecte

dG

A

Milk

inta

ke(t

est-

wei

ghin

g)

Cor

rela

tion

of IB

FAT

scor

es w

/ milk

inta

ke,

com

pare

d w

ith a

ssoc

iatio

nof

a fe

edin

g ob

serv

atio

nfo

rm d

esig

ned

to a

sses

sef

ficie

ncy

of B

F v

s bo

ttle

feed

ing.

IBFA

T p

ositi

vely

corr

elat

ed w

/ fee

ding

obse

rvat

ion

& m

ilkin

take

; IB

FAT

suc

king

scor

e si

g co

rrel

ated

w/ %

time

suck

ing/

suck

bur

sts.

IBFA

T d

oes

not

disc

rimin

ate

betw

een

adeq

uate

/in

adeq

uate

milk

inta

ke.

144



BF

To

ol*

Au

tho

r(s)

& D

ate

Too

l Des

crip

tio

nC

ou

ntr

y,se

ttin

gS

amp

le s

ize

& par

tici

pan

ts

Infa

nt

age

(mea

n,

sd)

Infa

nt/

mat

ern

alo

utc

om

es

Sta

tist

ical

An

alys

isR

esu

lts

Su

mm

ary

Au

tho

rC

om

men

ts/n

ote

s

Infa

nt

Bre

ast-

feed

ing

Ass

essm

ent

Too

l (IB

FAT

)(c

ont’d

)

Sch

lom

er,

1999

173

Eva

luat

ing

the

asso

ciat

ion

of tw

o B

Fas

sess

men

t too

ls w

ithB

F p

robl

ems

&sa

tisfa

ctio

n.

US

A,

hosp

ital

n=30

; Firs

ttim

e B

Fm

othe

rs o

fte

rm in

fant

s

12 h

ours

&1

wee

kpo

stpa

rtum

Mat

erna

lsa

tisfa

ctio

n&

BF

prob

lem

s

Pai

red

t-te

sts

to a

sses

s sc

ore

impr

ovem

ent 1

2 ho

urs-

1w

eek

post

part

um. C

ompa

red

corr

elat

ion

stre

ngth

of I

BFA

T&

LA

TC

H w

/ mat

erna

lsa

tisfa

ctio

n &

BF

pro

blem

s

Low

pre

dict

ive

valid

ity fo

rm

ater

nal s

atis

fact

ion

& B

Fpr

oble

ms

(ns=

0.06

-0.5

0) w

/bo

th to

ols.

Rio

rdan

,19

9717

4R

elia

bilit

y &

val

idity

test

ing

of 3

BF

asse

ssm

ent t

ools

.

US

A,

hosp

ital

& h

ome

visi

t

n=28

; Mot

hers

of h

ealth

y, te

rmne

w b

orns

.

1st w

eek

post

part

umn/

aC

ompa

red

valid

ity &

relia

bilit

y (in

ter-

rate

r &

test

/re-

test

) of

IBFA

T, M

BA

&LA

TC

H w

/ Spe

arm

an’s

corr

elat

ions

.

Inte

r-ra

ter

corr

elat

ions

: IB

FAT

(r=0

.27-

0.69

), L

AT

CH

(0.

11-

0.46

), M

BA

(0.

33-0

.66)

.Te

st-r

e-te

st c

orre

latio

ns 0

.64-

0.88

).

Met

hodo

-lo

gica

l iss

ues

usin

g ta

ped

BF

ses

sion

s.

LA

TC

Hk

Ass

essm

ent

Jen

sen

et

al, 1

99417

55

item

s: L

atch

; Aud

ible

swal

low

ing;

Typ

e of

nipp

le; C

omfo

rt o

fm

othe

r's b

reas

ts/

nipp

les;

Hel

p ne

eded

to h

old

baby

to b

reas

t.

US

AT

his

inst

rum

ent w

as b

ased

on

scie

ntifi

c un

ders

tand

ing

of la

ctat

ion

& c

linic

al e

xper

ienc

e. I

t was

not

psyc

hom

etric

ally

test

ed o

r va

lidat

ed a

gain

st a

ny m

ater

nal o

r in

fant

nut

ritio

nal o

r B

F o

utco

mes

,al

thou

gh it

has

sin

ce b

een

test

ed.

Rio

rdan

,20

01117

6P

redi

ctin

g B

F d

urat

ion

usin

g th

e LA

TC

H B

Fas

sess

men

t too

l.

US

A,

hosp

ital

n=13

3; M

othe

rsof

hea

lthy

sing

leto

ns (

38-

42 w

eeks

GA

)in

tend

ing

to B

F!6

wee

ks

39 h

ours

Not

BF

at 8

wee

ks (

noB

F in

pas

t24

hou

rs)

Fis

her’s

Exa

ct te

st: L

AT

CH

scor

es o

f wom

en B

F v

s no

tB

F a

t 6 w

eeks

. Ste

pwis

ere

gres

sion

to id

var

iabl

esas

soci

ated

w/ B

F d

urat

ion.

Reg

ress

ion

mod

el: L

AT

CH

scor

e, m

othe

r’s a

ge, i

nten

ded

BF

dur

atio

n &

del

iver

y ty

pe a

llsi

g re

late

d to

BF

at 6

wee

ks.

Aud

ible

swal

low

ing

not a

via

ble

varia

ble

until

day

4.

Ku

mar

,20

0617

7LA

TC

H s

corin

gsy

stem

& p

redi

ctio

n of

BF

dur

atio

n.

US

A,

hosp

ital

n=18

2 (4

day

s)n=

188

(6w

eeks

);M

othe

rs o

fhe

alth

y te

rmne

w b

orns

New

born

Mai

nten

-an

ce o

f BF

at 4

day

s &

6 w

eeks

Wilc

oxon

Ran

k Te

st to

com

pare

LA

TC

H s

core

s of

BF

vs n

ot B

F w

omen

at 6

wee

ks.

RO

C a

naly

sis

& Y

oude

n's

J to

eval

uate

spe

cific

ity/s

ensi

tivity

.C

orre

latio

n of

mot

her/

nurs

era

tings

to a

sses

s va

lidity

/su

bsca

le in

depe

nden

ce.

Sig

diff

LA

TC

H s

core

s 0-

48ho

urs

post

part

um in

BF

vs

not

BF

at 6

wee

ks.

RO

C a

naly

sis:

scor

e of

9 a

t 16-

24 h

ours

=1.

7*m

ore

likel

y to

BF

at 6

wee

ks.

Nur

se/m

othe

rs s

core

s si

gco

rrel

ated

& w

/ BF

dur

atio

n.S

ubsc

ales

inde

pend

ent e

xcep

tau

dibl

e sw

allo

w/la

tch

on.

Ad

ams,

1997

178

Mat

erna

l &pr

ofes

sion

alas

sess

men

t of B

F

US

A,

hosp

ital

n=35

; Firs

ttim

e B

Fm

othe

rs

Firs

t 2w

eeks

post

-pa

rtum

n/a

Inte

r-ra

ter

relia

bilit

y of

lact

atio

n co

nsul

tant

s us

ing

LAT

CH

& c

orre

latio

n w

/m

othe

rs L

AT

CH

sco

res.

85-1

00%

agr

eem

ent o

fla

ctat

ion

cons

ulta

nts.

Cor

rela

tion

w/ m

ater

nal

repo

rts

very

low

-mod

erat

e

Mot

hers

eval

uate

BF

expe

rienc

eso

mat

ical

ly,

not b

yob

serv

atio

n.

Table 38 cont’d

145



BF

To

ol*

Au

tho

r(s)

& D

ate

Too

l Des

crip

tio

nC

ou

ntr

y,se

ttin

gS

amp

le s

ize

& par

tici

pan

ts

Infa

nt

age

(mea

n,

sd)

Infa

nt/

mat

ern

alo

utc

om

es

Sta

tist

ical

An

alys

isR

esu

lts

Su

mm

ary

Au

tho

rC

om

men

ts/n

ote

s

Neo

nat

alO

ral-

Mo

tor

Ass

essm

ent

Sca

le(N

OM

AS

)

Pal

mer

et

al, 1

99317

928

item

s: n

utrit

ive/

non-

nutr

itive

suc

king

.O

utco

mes

: nor

mal

,di

sorg

anis

ed o

rdy

sfun

ctio

nal.

Latte

r tw

ogr

aded

by

seve

rity

(mild

, mod

erat

e, s

ever

e)

US

A,

hosp

ital

n=35

; Inf

ants

35-4

9 w

eeks

GA

, ! 1

900g

.

35-4

9w

eeks

PM

Al ;

term

& p

rete

rm

n/a

% p

erce

ntag

e ag

reem

ent o

f3

code

rsIn

ter-

rate

r re

liabi

lity:

80%

agr

eem

ent.

Ho

we,

2007

180

Psy

chom

etric

char

acte

ristic

s of

NO

MA

S in

hea

lthy

pre-

term

infa

nts

US

A,

med

ical

cent

re

n=14

7;m

othe

rs o

f pre

-te

rm, b

utot

herw

ise

heal

thy

infa

nts

32-3

6w

eeks

PM

A

Infa

nt fe

edin

gpe

rfor

man

ce:

tran

sitio

nal

rate

& v

olum

eof

milk

cons

umed

from

bot

tle.

Rel

iabi

lity

ofno

rmal

/dis

orga

nise

d sc

ales

& c

orre

latio

n w

/ inf

ant

feed

ing

skill

s. N

OM

AS

abi

lity

to d

etec

t cha

nge

in fe

edin

gsk

ills

over

tim

e (S

RM

Sm)

&si

g of

cha

nge

scor

es.

Acc

epta

ble

relia

bilit

y of

norm

al c

ateg

ory

(32-

35P

MA

) &

dis

orga

nise

dca

tego

ry (

32 w

eeks

PM

A).

All

cate

gorie

sm

oder

atel

y co

rrel

ated

w/ t

rans

ition

al m

ilk r

ate.

Mod

erat

eco

nver

gent

valid

ity fo

rno

rmal

&di

sorg

anis

edca

tego

ries.

da

Co

sta,

2008

181

The

rel

iabi

lity

ofN

OM

AS

Hol

land

,no

tst

ated

n=75

; Hea

lthy

& V

LBW

infa

nts.

Som

eris

k of

oth

erhe

alth

prob

lem

s

26-3

6P

MA

.In

ter-

rate

rag

reem

ent.

Coh

ens

Kap

pa r

elia

bilit

yco

effic

ient

Test

-ret

est o

f NO

MA

Sw

ith 4

rate

rs=

mod

erat

e-ne

arpe

rfec

t (0.

33-0

.94)

Adj

ust t

ool t

oim

prov

e re

liabi

lity

& in

corp

orat

ene

w k

now

ledg

eof

infa

ntsu

ck/s

wal

low

Ch

eckl

ist

fro

m p

aper

:'B

F &

th

eu

se o

fp

acif

iers

'

Rig

har

d,

1997

182

16 o

bser

vatio

ns: b

reas

tof

ferin

g (3

item

s),

suck

ing

at th

e br

east

(8

item

s), a

fter

feed

ing

(2ite

ms)

, & c

oncl

usio

ns (

2ite

ms)

Sw

eden

,ho

spita

ln=

82; E

BF

mot

hers

w/

inte

ntio

n to

BF

!6 m

onth

s.H

ealth

ym

othe

rs/in

fant

s, n

orm

alde

liver

y/bi

rth

wei

ghts

.

4-5

days

post

part

umB

F r

ate

&pa

cifie

r us

e(h

ours

/day

) at

2 w

eeks

, 1, 2

,3

& 4

mon

ths

Gro

up c

ompa

rison

s (C

hi2

orF

ishe

r's e

xact

test

) of

BF

tech

niqu

e &

dur

atio

n.

Pac

ifier

-use

rs s

ig h

ighe

rB

F a

t 4 m

onth

s in

corr

ect v

s in

corr

ect

suck

ing

grou

ps b

ut s

iglo

wer

than

non

-use

rs.

Non

-use

rs n

o di

ffere

nce

in B

F a

t 4 m

onth

s in

corr

ect v

s in

corr

ect

suck

ing

grou

ps.

Inco

rrec

t suc

king

tech

niqu

e m

ayno

t be

corr

ecte

dif

paci

fiers

use

d.

Han

ds

off

tech

niq

ue

(HO

T)

Ing

ram

,20

0218

38

guid

elin

es to

gui

dem

othe

rs in

'han

ds o

ff'w

ay to

pos

ition

& a

ttach

baby

. In

clud

es le

afle

tw

/ pic

ture

s &

expl

anat

ions

abo

ut B

F.

UK

,ho

spita

ln=

395;

Mot

hers

wer

eB

F o

ndi

scha

rge.

Not

repo

rted

BF

(an

yB

F/E

BF

) 2

&6

wee

ks p

ost

part

um

Chi

2to

iden

tify

fact

ors

sig

asso

ciat

ed w

/ BF

at 6

wee

ks.

Logi

stic

reg

ress

ions

toid

entif

y in

depe

nden

t var

iabl

esas

soci

ated

w/ B

F &

fact

ors

asso

c w

/ BF

at 6

wee

ks &

BF

asse

ssm

ent s

core

s.

Hig

h B

F te

chni

que

scor

e as

soci

ated

w/ B

Fat

6 w

eeks

.

Sho

rt, p

ragm

atic

trai

ning

for

mid

wiv

es to

teac

h go

od B

Fte

chni

que.

Table 38 cont’d

aBF

= B

reas

tfee

din

g;

bO

R =

Odds

Rat

io;

cRR =

Rel

ativ

e Ris

k;

dRO

C =

Rec

eive

r O

per

atin

g C

har

acte

rist

ic;

eSig

=

Sta

tist

ical

sig

nific

ance

(p<

0.0

5);

f

EBF

= E

xclu

sive

Bre

astf

eedin

g;

gG

A =

Ges

tation

al A

ge;

h

NS =

Not

sta

tist

ical

ly s

ignific

ant

(p>

0.0

5);

iAPG

AR =

Appea

rance

, Pu

lse,

Grim

ace,

Act

ivity,

Res

pirat

ion;

scor

e use

d t

o as

sess

the

hea

lth o

f new

bor

n c

hild

ren;

jVLB

W =

Ver

y Lo

w B

irth

Wei

ght;

kLA

TCH

was

ass

esse

d a

gai

nst

ext

ernal

outc

omes

by

2 furt

her

studie

s -

Rio

rdan

1997 a

nd S

chlo

mer

1999.

Thes

e st

udie

s ar

e det

aile

d in t

he

IBFA

T por

tion

of Ta

ble

2b;

lPM

A =

Pos

tmen

stru

al a

ge;

m

SRM

S -

Sta

ndar

dis

ed R

espon

se M

eans.

146



Six tools have been validated against a range of outcomes, although none of these include infant nutritional status (Table 38). Outcomes include immediate assessment of milk transfer through test-weighing, cessation of breastfeeding from the !rst days of life up to four months, introduction of complementary feeding, maternally reported breastfeeding satisfaction and problems. On the basis of consistency and volume of existing evidence, the NOMAS and BAS emerge as the strongest tools. NOMAS categories have been associated with transitional rate of milk, and its key properties (Table 37) are adequate. In two studies, BAS scores predicted early breastfeeding cessation in newborns.# However, thesetools include the least comprehensive coverage of breastfeeding domains and the NOMAS is arguably the most complex and least user-friendly tool overall. The Hands-o" technique checklist technique score was associated with breastfeeding at six weeks, therefore this may be a good way to teach newly delivered mothers good breastfeeding technique. Interventions from this checklist are also more empowering to the mother as she corrects her own technique and there are pictures to help identify problematic aspects of the feed.

The LATCH should be considered a weaker tool because of contradictory !ndings and questionable reliability and validity. Similarly, the IBFAT (Infant Breastfeeding Assessment Tool) lacks predictive validity for a range of outcomes and is unable to adequately discriminate between su$cient and insu$cient milk intake. Righard’s assessment184 of correct vs incorrect sucking was associated with breastfeeding status at four months if paci!ers were used, but not in the non-paci!er group.#

7.4.2 Coverage of breastfeeding domainsIt is striking that only one tool (BEET (Breastfeeding Evaluation and Education Tool)) achieves full coverage of domains (Table 39). The tools that achieve the widest coverage (IFE Module 2, BEET, UNICEF b-r-e-a-s-t and UNICEF Breastfeeding Observation Aid) are generally those that have been developed with developing country settings in mind. However, although these tools are based on extensive clinical and !eld experience, they su"er from a lack of validation, and some miss important domains (e.g. IFE Module 2misses ‘positioning’, UNICEF b-r-e-a-s-t misses ‘health of the baby’). These shortfalls could be addressed with appropriately designed studies and modi!cations. Furthermore, tools would need to take into account whether and how acute malnutrition in the infant a"ects assessment of any of the breastfeeding domains.

There is general consensus among the tools for inclusion of ‘attachment’ and ‘e"ective feeding’ domains. What is not clear, however, is the best way to assess these aspects of the breastfeed given the lack of validation against measurable outcomes. The important domain ‘health of the baby’, is only captured by 4/15 tools. It is crucial that this be included as an infant can be e"ectively breastfed but malnourished for another reason (e.g. HIV infected, severe wasting in the mother). Only half of the tools assess mothers’ ownbehaviour towards the baby, which may be very telling about her psychosocial status and could help guide appropriate management. Furthermore, the in%uence of severe malnutrition on the mother’s interaction with her infant is not addressed in any tool reviewed.#

147



Tab

le 3

9: T

oo

l co

vera

ge

of

bre

astf

eed

ing

do

mai

ns

and

val

ue

of

use

in d

iffe

ren

t co

nte

xts

Do

mai

ns

cove

red

aE

xtra

do

mai

ns

that

may

be

imp

ort

ant

Co

nte

xts

in w

hic

h B

F a

sses

smen

t to

ols

may

be

app

lied

b

BFc To

olBa

byBe

havio

urMo

ther

Beha

viour

Posit

ionAt

tach

-m

ent

E!ec

tive

Feed

ing

Healt

hof

Brea

stHe

alth

of Ba

byMo

ther

'sEx

perie

nce

of fe

ed

Num

ber,

timin

g or

dura

tion

of fe

eds

othe

r

BF Asse

ssm

ent

Scor

e (BA

S)

!!

prev

ious

BF

expe

rienc

e

Gen

eral

:Pros

- si

mpl

e, a

ccep

tabl

e, c

heap

& q

uick

toad

min

iste

r. 3/

5 ite

ms

obje

ctiv

e, q

uant

itativ

e. S

ome

pred

ictiv

eva

lidity

. Cons:

non-

com

para

ble

setti

ng; d

oes

not a

sses

snu

triti

onal

sta

tus;

onl

y ai

med

at 1

st 7

-10

days

of l

ife.

Dia

gn

osi

s:D

oes

not d

iscr

imin

ate

unde

rlyin

g ca

uses

of n

ot B

F.

Infa

nt B

FAs

sess

men

tTo

ol (I

BFAT

)

!!

!!

Gen

eral

:Pros

- si

mpl

e (in

clud

es d

efin

ition

s), a

ccep

tabl

e,ch

eap,

& q

uick

to a

dmin

iste

r. S

core

s w

eakl

y as

soci

ated

w/ m

ilkin

take

. Cons

- on

ly 1

/6 it

ems

obje

ctiv

e. S

ome

item

s va

gue.

Lim

ited

pred

ictiv

e va

lue

& v

aria

ble

relia

bilit

y. D

iag

no

sis:

Doe

sno

t dis

crim

inat

e un

derly

ing

caus

es o

f not

BF.

LATC

HAs

sess

men

t!

!!

!Pros

– si

mpl

e, q

uick

, che

ap &

qua

ntifi

able

. Ide

ntifi

es s

peci

fic B

Fpr

oble

ms.

Mot

her

can

use

mod

ified

ver

sion

for

self-

mon

itorin

g.S

ome

pred

ictiv

e va

lidity

. Cons

- R

elia

bilit

y/va

lidity

inco

nsis

tent

.A

udib

le s

wal

low

n/a

to c

olos

trum

. Acc

epta

bilit

y of

nip

ple

stim

ulat

ion

is q

uest

iona

ble.

No

obje

ctiv

e ite

ms.

Mot

her-i

nfan

tBF as

sess

men

tto

ol

!!

!!

!pr

evio

usB

Fex

perie

nce,

supp

ortiv

epa

rtne

r

Gen

eral

:Pros

- ch

eap

& q

uick

to a

dmin

iste

r. Q

uant

itativ

e 'ri

sk'

scor

e ge

nera

ted.

2/8

obj

ectiv

e ite

ms

re m

ilk in

take

, (e.

g. w

etna

ppie

s-bu

t mis

lead

s if

infa

nt g

iven

oth

er li

quid

s). Cons

- so

me

item

s su

bjec

tive/

unc

lear

des

crip

tions

of b

ehav

iour

s.V

alid

ity/r

elia

bilit

y un

know

n. D

iag

no

sis:

Gen

eral

item

s do

not

allo

w s

peci

fic d

iagn

osis

or

trea

tmen

t dec

isio

n.

Neon

atal

Ora

lM

otor

Asse

ssm

ent

Scal

e (NO

MAS

)

!!

Gen

eral

:Pros

- Id

entif

ies

spec

ific

suck

ing

prob

lem

s. Cons:

desi

gned

for

neon

ates

; sol

e fo

cus

on o

ral-m

otor

pat

tern

s, d

oes

not a

sses

s ov

eral

l BF.

Onl

y 5/

28 it

ems

obje

ctiv

e. C

ompl

ex &

time-

cons

umin

g. C

ateg

oris

atio

ns p

oten

tially

una

ccep

tabl

e to

mot

her

(e.g

. dys

func

tiona

l). V

aria

ble

relia

bilit

y, n

o va

lidity

dat

a.C

om

mu

nit

y se

ttin

g:

Not

sim

ple

to u

se, c

ompl

ex te

rmin

olog

y.

Chec

klist

from

pape

r: 'B

F &th

e use

ofpa

ci!er

s'

!!

!!

!!

!G

ener

al:Pros

- ac

cept

able

, che

ap, q

uick

to a

dmin

iste

r; c

over

sdi

ffere

nt s

tage

s of

BF

from

bef

ore

star

ts to

afte

r fin

ish.

Som

epr

edic

tive

valid

ity.Cons

- 1/

16 it

ems

obje

ctiv

e, n

ot q

uant

itativ

e.R

elev

ance

of s

ucks

/min

ute

is u

ncle

ar &

diff

icul

t to

coun

t.C

om

mu

nit

y se

ttin

g:

med

ical

term

inol

ogy

need

s cl

arifi

catio

n,no

t sim

ple

to u

se.

148



Do

mai

ns

cove

red

aE

xtra

do

mai

ns

that

may

be

imp

ort

ant

Co

nte

xts

in w

hic

h B

F a

sses

smen

t to

ols

may

be

app

lied

b

BFc To

olBa

byBe

havio

urMo

ther

Beha

viour

Posit

ionAt

tach

-m

ent

E!ec

tive

Feed

ing

Healt

hof

Brea

stHe

alth

of Ba

byMo

ther

'sEx

perie

nce

of fe

ed

Num

ber,

timin

g or

dura

tion

of fe

eds

othe

r

Chec

klist

from

pape

r: 'B

F &th

e use

ofpa

ci!er

s'

!!

!!

!!

!G

ener

al:Pros

- ac

cept

able

, che

ap, q

uick

to a

dmin

iste

r; c

over

sdi

ffere

nt s

tage

s of

BF

from

bef

ore

star

ts to

afte

r fin

ish.

Som

epr

edic

tive

valid

ity.Cons

- 1/

16 it

ems

obje

ctiv

e, n

ot q

uant

itativ

e.R

elev

ance

of s

ucks

/min

ute

is u

ncle

ar &

diff

icul

t to

coun

t.C

om

mu

nit

y se

ttin

g:

med

ical

term

inol

ogy

need

s cl

arifi

catio

n, n

otsi

mpl

e to

use

.

IFE -

Sim

ple,

rapi

d & fu

ll BF

asse

ssm

ent

!!

!!

!!

!!

Mix

edfe

edin

g,fe

ed e

nd,

paci

fier

use

Pros

- de

sign

ed fo

r de

velo

ping

/em

erge

ncy

setti

ngs.

Cle

ar &

suita

ble

for

use

by C

HW

dw

/ 1 tr

aini

ng d

ay.

Acc

epta

ble.

Iden

tifie

s un

derly

ing

caus

es o

f BF

pro

blem

s &

pro

mot

es m

ater

nal

conf

iden

ce.

Mix

ture

of o

bjec

tive

ques

tions

(13

/32)

& s

ubje

ctiv

eas

sess

men

t. Cons

- Ti

me-

cons

umin

g, tr

aini

ng c

osts

, no

relia

bilit

y/va

lidity

dat

a, n

ot q

uant

itativ

e.

A BF

Eval

uatio

n &Ed

ucat

ion T

ool

(BEE

T)

!!

!!

!!

!!

!si

gns

ofm

ilktr

ansf

erin

mot

her

Gen

eral

:Pros

- 13

/36

obje

ctiv

e ite

ms.

Man

y ite

ms

indi

cate

milk

tran

sfer

, cou

ld r

eass

ure

mot

her

abou

t milk

ade

quac

y; a

ccep

tabl

e;Cons

- N

ot s

impl

e/qu

ick

to a

dmin

iste

r. S

ome

item

s ne

ed a

ccur

ate

self-

repo

rt o

f qua

ntifi

able

item

s. N

o re

liabi

lity/

valid

ity d

ata.

Som

eite

ms

mon

itor

infa

nt o

ver

wee

ks, b

ut in

our

set

ting

only

1op

port

unity

to a

sses

s. H

osp

ital

set

tin

g:

Eas

ier

to m

onito

r in

-pa

tient

on

som

e ite

ms

on h

ospi

tal c

hart

com

pare

d to

CH

W.

Mot

her-B

aby

Asse

ssm

ent

(MBA

)

!!

!!

!!

sign

s of

milk

tran

sfer

in m

othe

r

Pros

– si

mpl

e &

qui

ck to

adm

inis

ter.

Qua

ntita

tive,

mea

ning

ful

item

s (e

.g. o

dd n

umbe

r sc

ore

indi

cate

s pr

ogre

ss o

f one

of t

hedy

ad b

ut n

ot th

e ot

her)

Acc

epta

ble.

Cons

- co

ntai

ns o

nly

1ob

ject

ive

item

(am

ount

of a

reol

a th

at in

fant

sho

uld

take

into

mou

th).

Poo

r to

mod

erat

e re

liabi

lity.

No

valid

ity d

ata.

Syste

mat

icAs

sess

men

t of

Infa

nt at

Brea

st (S

AIB)

!!

!!

Gen

eral

:Pros

- S

impl

e, a

ccep

tabl

e &

qui

ck to

adm

inis

ter.

Iden

tifie

s sp

ecifi

c B

F p

robl

ems.

Pos

ition

ing

diag

ram

s pr

ovid

edco

uld

be a

n ad

junc

t to

asse

ssm

ent &

to m

othe

rs. Cons

- N

ore

liabi

lity/

valid

ity d

ata.

No

quan

tifia

ble/

obje

ctiv

e ite

ms.

Co

mm

un

ity

sett

ing

:M

edic

al te

rmin

olog

y w

ould

nee

dcl

arifi

catio

n (e

.g. i

liac

cres

t, m

andi

ble

etc)

.

UNIC

EF B

-R-E

-A-

S-T

obse

rvat

iona

lch

eckl

ist

!!

!!

!!

!G

ener

al:Pro

- qu

ick,

sim

ple,

che

ap &

acc

epta

ble.

Iden

tifie

ssp

ecifi

c B

F p

robl

ems.

Con

- de

velo

ped

for

inte

nsiv

e tr

aini

ngm

odul

e &

invo

lves

cos

t & ti

me.

No

relia

bilit

y/va

lidity

dat

a. N

otqu

antit

ativ

e, o

nly

1 ob

ject

ive

item

. Co

mm

un

ity

sett

ing

:som

eite

ms

need

cla

rific

atio

n fo

r C

HW

s. H

osp

ital

set

tin

g:i

dent

ifies

spec

ific

BF

pro

blem

s &

trea

tmen

ts th

at c

ould

eas

e e.

g. n

ippl

e pa

in.

Table 39 cont’d

149



Do

mai

ns

cove

red

aE

xtra

do

mai

ns

that

may

be

imp

ort

ant

Co

nte

xts

in w

hic

h B

F a

sses

smen

t to

ols

may

be

app

lied

b

BFc To

olBa

byBe

havio

urMo

ther

Beha

viour

Posit

ionAt

tach

-m

ent

E!ec

tive

Feed

ing

Healt

hof

Brea

stHe

alth

of Ba

byMo

ther

'sEx

perie

nce

of fe

ed

Num

ber,

timin

g or

dura

tion

of fe

eds

othe

r

Mot

her-I

nfan

tBF

Prog

ress

Tool

(MIB

PT)

!!

!!

!!

!G

ener

al:Pros

- Acc

epta

ble,

che

ap &

qui

ck to

adm

inis

ter.

Goo

dre

liabi

lity.

Cons

- Ite

ms

need

def

inin

g. T

ool t

este

d w

/ mot

hers

w/

good

BF

so

may

not

iden

tify

prob

lem

s w

ell.

1/8

item

s ob

ject

ive.

Som

e am

bigu

ous

item

s (e

.g. '

redn

ess'

to id

entif

y ni

pple

trau

ma)

.N

o va

lidity

dat

a or

qua

ntita

tive

scor

ing.

Co

mm

un

ity

sett

ing

:M

ayno

t be

suita

ble

for

CH

Ws.

Ho

spit

al s

etti

ng

:co

uld

iden

tify

spec

ific

prob

lem

s &

trea

tmen

ts (

e.g.

atta

chm

ent a

dvic

e; n

ippl

e ca

re).

Pre-

term

Infa

nt B

FBe

havio

urSc

ale (

PIBB

S)

!!

!!

Gen

eral

: Pro

s -

iden

tifie

s sp

ecifi

c B

F p

robl

ems.

Ass

esse

s B

F fr

omst

art t

o fin

ish.

Acc

epta

ble

& q

uick

to a

dmin

iste

r. R

elia

bilit

y is

goo

dbe

twee

n he

alth

pro

fess

iona

ls. C

ons

- on

ly 3

obj

ectiv

e/qu

antif

iabl

eite

ms.

No

clea

r sc

orin

g. R

elia

bilit

y be

twee

n m

othe

rs &

hea

lthpr

ofes

sion

als

is p

oor.

No

valid

ity d

ata.

Com

mun

ity s

ettin

g:D

esig

ned

for

prem

atur

e in

fant

s &

neo

nate

s in

hos

pita

l & n

/a to

gene

ral p

opul

atio

ns. H

ospi

tal s

ettin

g: M

aybe

use

ful i

n sp

ecia

lised

nurs

ery

setti

ng.

UNIC

EF 20

06:

BF Obse

rvat

ion

Aid

!!

!!

!!

!!

Gen

eral

: Pro

s -

sim

ple,

acc

epta

ble,

qui

ck &

che

ap to

adm

inis

ter.

Iden

tifie

s pr

oble

ms

that

dire

ctly

tran

slat

e in

to in

terv

entio

n (e

.g.

nipp

le c

ream

for

nipp

le tr

aum

a). C

ons

- so

me

item

s va

gue

(e.g

.'m

othe

r lo

oks

heal

thy'

). N

ot q

uant

itativ

e. N

o re

liabi

lity/

valid

ity d

ata,

but t

ool b

eing

use

d in

Sie

rra

Leon

e 20

09 &

may

pro

vide

rel

evan

tda

ta. C

omm

unity

set

ting:

nee

d to

cla

rify

'oxy

toci

n re

flex'

Hand

s o"

tech

niqu

e(H

OT

!!

!!

!P

ros

– si

mpl

e, a

ccep

tabl

e &

qui

ck to

adm

inis

ter.

Incl

udes

leaf

let w

/pi

ctur

es to

aid

BF

ass

essm

ent &

for

mot

hers

to k

eep.

Goo

dpr

edic

tive

valid

ity &

iden

tifie

s sp

ecifi

c B

F p

robl

ems.

Con

s -

cost

of

leaf

lets

(if

used

), tr

aini

ng r

equi

red.

Ass

umes

opt

imal

sur

roun

ding

s&

rel

axat

ion

aad

apte

d fro

m H

all M

oran

et

al (

2000)

bad

apte

d fro

m M

yatt

et

al (

2006)

to incl

ude:

pop

ula

tion

scr

eenin

g (

by

com

munity

volu

nte

ers

of p

rim

ary

leve

l hea

lthca

re w

orke

rs)

case

fin

din

g in c

linic

al c

onte

xt &

dia

gnos

is in c

linic

al c

onte

xts

cBre

astf

eedin

gd

Com

munity

Hea

lth W

orke

r

Table 39 cont’d

150



7.4.3 Tool performance characteristicsWe were unable to apply a formal grading procedure (e.g. GRADE185) to the tools, given the largely observational study designs of which none would have achieved higher than a GRADE 2, or the complete lack of tool testing. Other reviews of breastfeeding assessment tools have excluded studies on the basis of design and validation issues. We, however, deliberately included as many tools as possible, though this meantwe were unable to assign formal grades. Instead, we analyse here a number of performance characteristics:

a) Community-based careCommunity-based management of acute malnutrition (CMAM) is increasingly the approach of choice for the management of SAM in children six to 59 months with no co-existing medical complications. If such a model of care was extended to infants <6m, breastfeeding assessment tools to help CHWs identify, resolveand refer breastfeeding problems would be needed. Such tools should also bene!t children to two years or beyond, to support sustained and continued breastfeeding.

Eight of the tools are either not relevant to a community setting or would require adaptation for use by a CHW. Many items also involve complex medical terminology, so some adaptation or the provision of clear de!nitions would be of bene!t. Training lasting several days would be necessary for some of the tools, which is a time and !nancial cost that may not be feasible in an emergency setting. However one tool, the BAS, could be a useful basis to help CHWs to decide on referral to a facility. A study to test the BAS tool in acommunity based setting for managing acute malnutrition would be valuable.

b) Hospital-based careEight of the tools appear to be useful in a clinical context. One tool, the BAS, was considered to be of limited use in a hospital setting as it would not provide extra information about speci!c elements of the breastfeed to diagnose speci!c problems. Currently most cases of SAM in infants <6m are managed in inpatient settings. We suggest a study to evaluate a tool(s) for breastfeeding assessment and problem identi!cation in a clinical setting where MAMI takes place.

The recent implementation of a checklist to ensure simple guidelines were followed during surgery resulted in signi!cant reduction of patient mortality and morbidity.186 We believe a breastfeeding checklist as part of admission assessment could operate in a similar way by highlighting simple problems that could be easily overlooked.

Overall, few of the tools were su$ciently versatile to be used across community and hospital contexts without substantial modi!cation.

c) Lack of ‘gold standard’ treatment of malnourished infants <6mThe success of a breastfeeding assessment tool in relation to nutritional/mortality outcome is dependent on the interventions available to address the problems identi!ed. As stated in earlier chapters, there is currently no gold standard treatment for MAMI. Intensive, inpatient programmes as suggested in current guidelines might be enhanced by a tool that identi!es breastfeeding problems and appropriate remedial action. Community-based programmes have a larger potential capacity for provision of treatment; the question remains which infants <6m with SAM can be managed #in the community.

d) Continued breastfeeding in older infants and childrenBreastfeeding assessment and support is not just a consideration for infants <6m. Actions to support and sustain breastfeeding as part of complementary feeding should feature in the therapeutic management and rehabilitation of children up to two years of age and beyond. One promising initiative is the development of a training content to integrate training on infant and young child feeding into CTC/CMAMprogramming, piloted in Sierra Leone Feb/March 2009 (ENN/Valid/UNICEF) and in Zimbabwe in June 2009 (ENN/SCUK/UNICEF) (personal communication, Mary Lung’aho).187 Reliable and valid breastfeeding assessment tools developed for infants <6m could have wider application in children up to two years of age and build upon current initiatives to address shortfalls in infant and young child feeding support available to this older age-group.

7.5 LimitationsThis review was systematically conducted, but it was not exhaustive due to time and resource constraints. Formal grading of each of the breastfeeding assessment tools was not possible due to a lack of appropriatelydesigned study, and no gold standard for treatment of SAM in the <6m age group. The review is therefore more subjective than ideal, which limits any strong recommendations about which tools perform best in each context.

151


7.5 Limitations

7.6 ConclusionsThis review highlights the need for simple, accurate and valid assessment tools predictive of breastfeedingand infant nutritional outcomes. Di"erent tools, or di"erent versions of the same tool, are necessary for several related but separate purposes:

• A short, sensitive and quick version to identify breastfeeding problems and actions at community level.• A more speci!c, detailed tool to aid diagnosis and appropriate treatment in an inpatient/clinical context• A tool with intermediate balance of sensitivity and speci!city for use in diagnosis and management for

outpatient primary healthcare services

At present, there is no single tool or set of tools that meets these ideals. Many existing tools are too narrowin scope or have not been robustly validated, especially in nutritionally vulnerable developing country / emergency settings. This is an important gap that warrants future research.

In the interim, UNICEF b-r-e-a-s-t, the UNICEF 2006 BF observation aid and the aids described in IFE Module 2 o"er the most promise for !eld use in programmes managing infants <6m. Operational researchcould be useful in con!rming their utility as ‘best currently available’ tools. They could also form the basis of future tools optimised to meet the needs of malnourished infants <6m. There is a need to identify the most appropriate outcome(s) for validating new tools (e.g. rate of weight gain after stabilisation).

7.7 Summary !ndings and recommendationsSummary !ndings15 breastfeeding assessment tools were identi!ed in the literature search. No single tool or set of tools was identi!ed for use in MAMI. Many tools are either too narrow in scope or have not been robustly validated against nutritional or health outcomes, or applied in nutritionally vulnerable developing country/emergency settings. #

The in%uence of infant SAM on the mother’s interaction with her infant is not addressed in any tool reviewed.#

The Breastfeeding Assessment Tool (BAS) could be tested for its suitability to community settings (e.g. use by community health workers for case !nding and assessment). Half of the tools reviewed could be useful for inpatient assessment.#

There is no tool that is su$ciently sensitive for community use or speci!c for use in inpatient settings. It is likely that di"erent tools will be needed for each setting, and possibly a third that is a balance of both for use in primary healthcare services.#

In the interim, UNICEF b-r-e-a-s-t, the UNICEF 2006 breastfeeding observation aid and the aids described in IFE Module 2 were identi!ed as most useful to assess breastfeeding in programmes managing infants <6m.

Summary recommendationsThere is a need for simple, accurate and valid assessment tools that are predictive of breastfeeding and infant nutritional outcomes. Such tools could also have wider application for children up to two years.

Quality research studies to test the validity of existing breastfeeding assessment tools are needed. For example, the Breastfeeding Assessment Tool (BAS)188 could be tested for its suitability to community settings (e.g. use by community health workers for case !nding and assessment). Half of the tools reviewed could be useful for inpatient assessment.#

New tools should be developed, possibly based on existing tools, to meet the needs of malnourished infants <6m, suitable for use in case !nding in the community, inpatient settings and outpatient primary healthcare programmes.

The success of a breastfeeding assessment tool and how it relates to nutritional/morbidity outcomes is dependent on the interventions available to address problems identi!ed. The lack of a ‘gold standard’ treatment for infants <6m is a limiting factor in this regard.

Severe maternal wasting and maternal and infant HIV status are important considerations in assessing breastfeeding e"ectiveness.

152


7.7 Summary !ndings and recommendations

153 Moran VH, Dinwoodie K, Bramwell R, Dykes F. A critical analysis of the content of the tools that measure breast-feeding interaction. Midwifery. 2000 Dec;16(4):260-8.

154 Myatt M, Khara T, Collins S. A review of methods to detect cases of severely malnourished children in the community for their admission into community-based therapeutic care programs. Food Nutr Bull. 2006 Sep;27(3 Suppl):S7-23.

155 Moran VH, Dinwoodie K, Bramwell R, Dykes F. A critical analysis of the content of the tools that measure breast-feeding interaction. Midwifery. 2000 Dec;16(4):260-8.


157 Brahmbhatt, Heena PhD; Kigozi, Godfrey MD; Wabwire-Mangen, Fred PhD; Serwadda, David MD; Lutalo, Tom MSc; Nalugoda, Fred MD; Sewankambo, Nelson MD; Kiduggavu, Mohamed MD; Wawer, Maria MD; Gray, Ronald MD (2006). Epidemiology and Social Science Mortality in HIV-Infected and Uninfected Children of HIV-Infected and Uninfected Mothers in Rural Uganda. JAIDS: 1 April 2006 - Volume 41 - Issue 4 - pp 504-508

158 Moran, V.H., Dinwoodie, K., Bramwell, R., Dykes, F. (2000) A critical analysis of the content of the tools that measure breast-feeding interaction. Midwifery. 2000 Dec;16(4):260-8.


160 Mizuno K, Aizawa M, Saito S, Kani K, Tanaka S, Kawamura H, et al. (2006)Analysis of feeding behavior with direct linear transformation. Early Hum Dev. 2006 Mar;82(3):199-204.

161 Johnson et al, 1999. A home visit programme for breastfeeding education and support. JOGNN, 28; 480-485, 1999.162 Tobin, D.L. (1996) A breastfeeding evaluation and education tool. J Hum Lact. 1996 Mar;12(1):47-9.163 Mulford, C. (1992) The Mother-Baby Assessment (MBA): an "Apgar score" for breastfeeding. J Hum Lact. 1992 Jun;8(2):79-82.164 Shrago, L. & Bocar, D. (1990) The infant's contribution to breastfeeding. J Obstet Gynecol Neonatal Nurs. 1990 May-Jun;19(3):209-15.165 WHO/UNICEF, 1993. Breastfeeding counselling a training course. Participants Manual. P21. 1993. WHO/CDR/93.5 UNICEF/NUT/93.3166 Johnson, T.S., Mulder, P.J., Strube, K. (2007) Mother-Infant Breastfeeding Progress Tool: a guide for education and support of the

breastfeeding dyad. J Obstet Gynecol Neonatal Nurs. 2007 Jul-Aug;36(4):319-27.167 Nyqvist, K.H., Rubertsson, C., Ewald, U., Sjoden, P.O. (1996) Development of the Preterm Infant Breastfeeding Behavior Scale (PIBBS): a

study of nurse-mother agreement. J Hum Lact. 1996 Sep;12(3):207-19.168 UNICEF, 2006. BABY-FRIENDLY HOSPITAL INITIATIVE. Revised, Updated and Expanded for Integrated Care. Section 3.2, p91169 Hall, R.T., Mercer, A.M., Teasley, S.L., McPherson, D.M., Simon, S.D., Santos, S.R., et al. (2002) A breast-feeding assessment score to

evaluate the risk for cessation of breast-feeding by 7 to 10 days of age. The Journal of Pediatrics. 2002;141(5):659-64.170 Gianni, M.L., Vegni, C., Ferraris, G., Mosca, F. (2006) Usefulness of an assessment score to predict early stopping of exclusive breast-

feeding. J Pediatr Gastroenterol Nutr. 2006 Mar;42(3):329-30.171 Matthews, M.K. (1988) Developing an instrument to assess infant breastfeeding behaviour in the early neonatal period. Midwifery. 1988

Dec;4(4):154-65.172 Furman, L., Minich, N.M. (2006) Evaluation of breastfeeding of very low birth weight infants: Can we use the infant breastfeeding

assessment tool? Journal of Human Lactation. 2006;22(2):175-83.173 Schlomer, J.A., Kemmerer, J., Twiss, J.J. (1999) Evaluating the Association of Two Breastfeeding Assessment Tools with Breastfeeding

Problems and Breastfeeding Satisfaction. Journal of Human Lactation. 1999;1 5(1):35.174 Riordan, J.M., Koehn, M. (1997) Reliability and validity testing of three breastfeeding assessment tools. J Obstet Gynecol Neonatal Nurs.

1997 Mar-Apr;26(2):181-7.175 Jensen, D., Wallace, S., Kelsay, P. (1994) LATCH: a breastfeeding charting system and documentation tool. J Obstet Gynecol Neonatal

Nurs. 1994 Jan;23(1):27-32.176 Riordan, J., Bibb, D., Miller, M., Rawlins, T. (2001) Predicting Breastfeeding Duration Using the LATCH Breastfeeding Assessment Tool.

Journal of Human Lactation. 2001;17(1):20.177 Kumar, S.P., Mooney, R., Wieser, L.J., Havstad, S. (2006) The LATCH scoring system and prediction of breastfeeding duration. Journal of

Human Lactation. 2006;22(4):391-7.178 Adams, D., Hewell, S. (1997) Maternal and professional assessment of breastfeeding. J Hum Lact. 1997 Dec;13(4):279-83.179 Palmer, M.M., Crawley, K., Blanco, I.A. (1993)Neonatal Oral-Motor Assessment scale: a reliability study. J Perinatol. 1993 Jan-

Feb;13(1):28-35.180 Howe, T.H., Sheu, C.F., Hsieh, Y.W., Hsieh, C.L. (2007) Psychometric characteristics of the Neonatal Oral-Motor Assessment Scale in

healthy preterm infants. Dev Med Child Neurol. 2007 Dec;49(12):915-9.181 da Costa, S.P., van der Schans, C.P. (2008) The reliability of the Neonatal Oral-Motor Assessment Scale. Acta Paediatr. 2008

Jan;97(1):21-6.182 Righard, L., Alade, M.O. (1997) Breastfeeding and the use of pacifiers. Birth. 1997 Jun;24(2):116-20.183 Ingram, J., Johnson, D., Greenwood, R. (2002) Breastfeeding in Bristol: teaching good positioning, and support from fathers and families.

Midwifery. 2002 Jun;18(2):87-101.184 Righard, L., Alade, M.O. (1997) Breastfeeding and the use of pacifiers. Birth. 1997 Jun;24(2):116-20.185 Guyatt, G.H., Oxman, A.D., Vist, G.E., Kunz, R., Falck-Ytter, Y., Alonso-Coello, P., et al. (2008) GRADE: an emerging consensus on rating

quality of evidence and strength of recommendations. BMJ. 2008 Apr 26;336(7650):924-6.186 Haynes AB, Weiser TG, Berry WR, Lipsitz SR, Breizat A-HS, Dellinger EP, et al. A Surgical Safety Checklist to Reduce Morbidity and

Mortality in a Global Population. N Engl J Med. 2009 January 29, 2009;360(5):491-9.187 ENN (2009). Integration of IYCF support into CTC/CMAM. Facilitators Guide and handouts. ENN, IFE Core Group & collaborators. October

2009. 188 Hall et al, 2002. A breast-feeding assessment score to evaluate the risk for cessation of breast-feeding by seven to ten days of age. J

Pediatr 2002;141:659-64

Endnotes

153


Documents

Chapter 7 Review of breastfeeding assessment toolstexastenstep.org/.../12/review-of-breastfeeding-assessment-tools.pdf · 7.2 What should a breastfeeding assessment tool include?