REPORT OF THE JOINT CONSULTATION ON APPROACHES TO

REPORT OF THE JOINT CONSULTATION

ON APPROACHES TO MEASURE

COVERAGE OF NUTRITION

COUNSELING INTERVENTIONS

Washington, DC | September 17, 2018

ii 
 
Contents 
Acknowledgements ...................................................................................................................................... iii 
Abbreviations and Acronyms ....................................................................................................................... iii 
1.  Introduction .......................................................................................................................................... 1 
1.1.  Aim of the consultation................................................................................................................. 1 
1.2.  Objectives of the consultation ...................................................................................................... 1 
1.3.  Expected outcomes of the consultation ....................................................................................... 1 
2.  Setting the scene ................................................................................................................................... 2 
2.1.  The need for indicators of IYCF counseling coverage ................................................................... 2 
2.2.  Learning from the MNCH coverage measurement process ......................................................... 3 
3.  Review of current efforts to measure IYCF counseling coverage in program evaluations and large-
scale surveys ......................................................................................................................................... 4 
3.1.  Development of conceptual framework of 7 elements of coverage measures ........................... 4 
3.2.  Findings from case studies of coverage measurement across program evaluations ................... 6 
3.3.  Experiences from large nationally representative surveys ........................................................... 9 
4.  Group discussions about coverage measurement .............................................................................. 12 
4.1.  Framing the group discussions ................................................................................................... 12 
4.2.  Summary of group discussions ................................................................................................... 13 
Question 1: What survey questions can best capture exposure to counseling interventions? ......... 13 
Question 2: What counseling coverage measurement issues should be addressed in an extended 
guidance? ............................................................................................................................................ 15 
Question 3: What are additional research needs for development and validation of survey-based 
coverage measures? ........................................................................................................................... 15 
5.  Closing remarks and way forward ...................................................................................................... 16 
Annex 1. Agenda ......................................................................................................................................... 18 
Annex 2. List of participants ........................................................................................................................ 20 
Annex 3. List of studies and surveys reviewed ........................................................................................... 22 
Annex 4. Questions about IYCF advice in last 6 months ............................................................................. 23 
Annex 5. Exercise on additional survey questions on counseling coverage ............................................... 25 
 
   

iii

Acknowledgements

The meeting was organized by Alive & Thrive and DataDENT with the WHO-UNICEF Technical Expert

Advisory Group on Nutrition Monitoring (TEAM) and hosted by the International Food Policy Research

Institute (IFPRI). The core planning team comprised Karin Lapping, Silvia Alayon, Rebecca Heidkamp,

Kuntal Saha, Larry Grummer-Strawn, Chika Hayashi, Purnima Menon and Sunny Kim.

We gratefully acknowledge the contributions of the participants during and after the meeting, and

funding support from the Bill & Melinda Gates Foundation.

Jowel Choufani, Sunny Kim and Purnima Menon (IFPRI) prepared this report.

Abbreviations and Acronyms

ANC antenatal care

BF breastfeeding

CF complementary feeding

DHS Demographic and Health Surveys program

GNMF Global Nutrition Monitoring Framework

IYCF infant and young child feeding

MCH maternal and child health

MICS Multiple Indicator Cluster Survey

MIYCN maternal, infant and young child nutrition

MNCH maternal, newborn and child health

PNC postnatal care

TEAM Technical Expert Advisory group for nutrition Monitoring

UNICEF United Nations Children’s Fund

USAID United States Agency for International Development

WHO World Health Organization

1. Introduction

1.1. Aim of the consultation

The consultation aimed to consolidate insights from research and implementation efforts and facilitate

dialogue that can help make progress towards the development of indicators to measure coverage of

infant and young child feeding (IYCF) counseling or behavior change interventions.

1.2. Objectives of the consultation

The three objectives of the consultation were:

1) To take stock of survey-based approaches to measure exposure to IYCF counseling and support

interventions;

2) To move towards consensus on potential approaches to measuring coverage of IYCF counseling

and support interventions; and

3) To identify research needs related to coverage measurement of IYCF counseling interventions.

1.3. Expected outcomes of the consultation

The consultation was expected to lead to the development of core survey questions to capture exposure

to IYCF counseling interventions, guidance for an approach to design counseling coverage measures and

summary of research needs. Outcomes of deliberations at this meeting then fed into a larger global

consultation on Measuring Nutrition in Population-Based Household Surveys and Associated Facility

Assessments which occurred that same week in Washington, DC. The results will also be reported back

to the World Health Organization (WHO), United Nations Children’s Fund (UNICEF) and the Technical

Expert Advisory group for Nutrition Monitoring (TEAM) and will contribute to guidance to WHO Member

States on measuring coverage of maternal, infant and young child nutrition (MIYCN) counseling or

behavior change interventions.

2. Setting the scene

2.1. The need for indicators of IYCF counseling coverage

The indicators for assessing IYCF practices, defined by the WHO and UNICEF in 2008, have advanced the

ability of the nutrition community to collect harmonized IYCF-related data. The availability of these

indicators has allowed countries to identify gaps in nutrition practices, make valid cross-country

comparisons and evaluate the progress of IYCF programs. It has spurred action to invest in programs to

strengthen these practices globally – a majority of countries surveyed in the WHO’s Global Nutrition

Policy Review (WHO, 2018

) report that they implement programs to support breastfeeding and

complementary feeding.

However, little is known about the reach and coverage of such programs because current survey-based

data collection systems do not include indicators to capture the coverage of interventions to support

IYCF. Guidance on tracking progress on global nutrition commitments has also recently highlighted that

an investment in strengthening measurement of coverage for counseling programs or interventions is

central to measuring program performance (WHO, 2017

A result of the lack of availability of coverage indicators for nutrition counseling interventions to support

MIYCN practices is that global monitoring initiatives such as the Countdown to 2030 and the Global

Nutrition Report use IYCF practices, such as infant feeding, as proxies for program coverage or

performance. This raises several challenges because indicators of practices likely do not accurately

reflect intervention coverage.

The Global Nutrition Monitoring

Framework (GNMF), released in 2016,

included a draft indicator on

breastfeeding counseling

. The WHO-

UNICEF Technical Expert Advisory

group for nutrition Monitoring

(TEAM) has been tasked to further

develop and validate this indicator to

support WHO Member States to

operationalize and report to the

World Health Assembly starting 2018.

Since the data to compute an

indicator on breastfeeding counseling

Available at: http://www.who.int/nutrition/topics/global-nutrition-policy-review-2016.pdf

http://apps.who.int/iris/bitstream/10665/259904/1/9789241513609-eng.pdf?ua=1

The Global Nutrition Monitoring Framework includes the following as an indicator of reach of counseling

programs: “Proportion of mothers of children 0-23 months who have received counseling, support or messages on

optimal breastfeeding at least once in the previous 12 months”.

does not yet exist in most national data systems, current guidance suggests an interim indicator on

program availability

In the absence of global guidance and indicators on how MICYN counseling should be delivered, the

interventions can vary tremendously depending on context. For example, the types of providers that are

tasked with delivering counseling, location of services and the timing and frequency of contact vary from

one context to another. This will make standardizing coverage measures challenging. Achieving

consensus on a set of indicators is imperative given the current scale of global programming to support

MIYCN behavior change interventions and the global monitoring needs described above.

This consultation was particularly timely in 2018 because of ongoing work by the WHO to develop global

guidance for breastfeeding counseling and support programs. In addition, there is recognition that an

investment in strengthening measurement of coverage for counseling programs/interventions is central

to assessing program performance.

2.2. Learning from the MNCH coverage measurement process

The consultation included an overview of the lessons learned from research to improve coverage

measurement related to maternal, newborn and child health (MNCH), a continuing body of work under

the Improving Coverage Measurement (ICM) project (2013-2018, continuing under the Improve project).

ICM aimed to increase the availability of evidence for the validity of existing and new MNCH coverage

indicators collected through household surveys. It also aimed to increase the availability of evidence-

based tools and protocols for routine national-level linkage of data on care-seeking from household

surveys with results from service provider assessments.

Findings from several previous studies published in two journal supplements on measuring coverage

were presented. The two collections are Measuring Coverage in MNCH

by the Child Health

Epidemiology Reference Group (CHERG)

and Improving Coverage Measurement

by ICM. As most intervention exposure

is self-reported by respondents, there is

the potential for misclassification and

bias. Insights related to sources of error

in coverage data were discussed, as well

as what has been learned from various

indicator validation studies (Munos et

al., 2018). Key messages that are

applicable to measuring IYCF counseling

coverage include considering whether

the precise cadre of providers matters

(as types of health workers seem to be

poorly reported), whether precise timing of counseling matters (as timing of events may be challenging

The current interim indicator is defined as “availability of a national program that includes provision for delivering

breastfeeding counseling services to mothers of infants 0-23 months of age through health systems or other

community-based platforms”.

https://collections.plos.org/measuring-coverage-in-mnch

http://www.jogh.org/col-coverage-measurement.htm

to recall, particularly during a sensitive time such as delivery), improving recall with memory aids, and

reporting accuracy may be more of a problem than recall (McCarthy et al., 2016). Results from past

coverage indicator validation studies that reporting accuracy did not significantly decrease with longer

recall periods (Chang et al., 2018; McCarthy et al., 2018; Hazir et al., 2013). Moreover, for interventions

targeted at small sub-groups (for example, children with pneumonia or severe acute malnutrition), the

denominator must be measured with high specificity as to avoid large numbers of false positives in the

denominator making the indicator difficult to interpret (Ayede et al., 2018; Hazir et al., 2013).

3. Review of current efforts to measure IYCF counseling coverage in

program evaluations and large-scale surveys

Building on the lessons learned from MNCH coverage measurement and in line with the larger objective

of the consultation meeting, key issues related to IYCF counseling coverage measurement were

discussed in detail, with specific examples from evaluation case studies and large, nationally-

representative, population-based surveys.

3.1. Development of conceptual framework of 7 elements of coverage measures

Background work leading up to the consultation included a review of household survey questions

developed by researchers and program evaluators to measure exposure to or coverage of IYCF

counseling and behavior change interventions in different contexts. A total of 16 studies (Annex 3) of

program evaluations (interpersonal communication/nutrition education interventions related to IYCF) as

well as 3 nationally representative surveys (PMA2020, DHS and MICS) were reviewed.

A framework for IYCF counseling measurement was created to illustrate the potential elements to be

captured in the design of coverage measures (Figure 1).

Figure 1 Seven elements addressed in counseling coverage questions

Target behavior/

content

Timing of

contact

Recall

period

Mode of

intervention

Place of

contact

Type of

service

provider

Frequency

of contact/

duration

The seven elements include:

- Target behavior/content: Age-appropriate target behavior or the content of counseling and

support received, for example early initiation of breastfeeding and support for proper

positioning and attachment.

- Recall period: The length of time respondents are asked to consider in responding to a question.

Recall periods for reporting the counseling received vary in length and detail. For example,

common recall periods included the last six months and the last year, while specific recall

periods included the last 7-8 months of pregnancy, in the first two days after delivery, or during

the last visit with a health care provider.

- Timing of contact: There are six general time-periods compiled across the surveys of when IYCF

counseling occurs: pregnancy or during antenatal care (ANC) visits, at or immediately after

delivery, first six weeks after childbirth or during postnatal care (PNC) visits, 0-5 months of age,

6-23 months of age, and sick child contacts.

- Frequency of contact/duration: Whether any counseling occurred, or the number of sessions,

and the duration of session received.

- Type of service provider: The type or cadre of providers delivering counseling, ranging from

health care professionals (e.g. doctors, nurses and midwives) to community volunteers and

family members.

- Place of contact: The location of counseling such a health care facility and the home.

- Mode of intervention: The mode of counseling includes individual or one-to-one contact, group

sessions and via mobile devices.

The review of survey questions showed substantial variability in recall period, frequency, type of

provider, and place of contact across different timings of counseling contacts. The parameters of survey

design and sampling affected the formulation of survey questions on coverage across all its dimensions.

For example, if a survey is administered to all women of reproductive age, additional questions may be

required to filter out women who do not have children in the particular age range (child age range to be

eligible for breastfeeding and complementary feeding counseling); surveys administered to a specific

target group (e.g. recently delivered women) may include more specified questions, or those

administered more frequently may include questions with shorter recall periods. Furthermore, the

design of interventions assessed in most evaluations varied, thereby also leading to variations across the

dimensions.

There was discussion on the need for core and expanded sets of survey questions that will facilitate

reporting of harmonized indicators. Core questions are the minimum set for inclusion in large nationally-

representative household surveys where there is limited space to include topic-specific questions, while

the expanded questions could be implemented as a stand-alone survey, an add-on module to national

surveys or in smaller scale studies. When considering a set of core questions, it was suggested that

addressing timing, frequency and content of the counseling intervention may be important, with the

recall period as close to the service contact as feasible. For an expanded set of questions, it may be

important to consider additional information on the types of service provider, locations of service,

duration of contact, and mode of intervention.

3.2. Findings from case studies of coverage measurement across program

evaluations

Experiences in developing survey questions on IYCF counseling coverage, consideration of the seven

elements of coverage measurement, and performance of the survey questions were discussed in the

context of specific program evaluations.

Program evaluations case studies

Three case studies of coverage measurement in program monitoring and evaluation were presented:

Alive & Thrive - Bangladesh, Sanitation Hygiene Infant Nutrition Efficacy (SHINE) trial, and Suaahara II.

The program designs were presented as well as key takeaways from analyses looking at some of the

seven elements of coverage. Summaries of each measurement case study from program evaluations and

the key takeaways are below:

Alive & Thrive - Bangladesh

Alive & Thrive took place in

Bangladesh between 2010 and 2014

and was designed to deliver

interventions aimed to improve IYCF

practices through intensified inter-

personal communication, mass media,

and community mobilization delivered

at scale in the context of policy

advocacy. A total of 27 visits by

frontline worker from pregnancy to

the end of the first 24 months after

delivery were intended to be provided

to mothers. Key IYCF counseling content included proper attachment and position, early initiation and

exclusive breastfeeding, and demonstration of age-specific complementary feeding, responsive feeding,

and handwashing.

Takeaways:

- Recall period: Intervention beneficiaries received counseling at various time points, so having

questions with different recall periods helped to triangulate responses and shed light on

program fidelity. However, different recall periods posed various implications for data validity.

Results showed that shorter, more specific recall periods had less spread of data points while

longer, unbound recall periods had a larger spread and more data outliers potentially indicating

lower accuracy.

- Type of service provider: Different types of frontline workers provided counseling, so coverage

questions were developed to capture respective proportions of visits and frequencies of

counseling delivered by each type of service provider. This was important to capture as the

types of information and modes of providing information may vary depending on who provides

counseling.

- Frequency: Frequency of home visits by frontline workers varied at different stages of pregnancy

to the first 2 years of life. Asking about frequency of contact at different periods provided

information on fidelity to program design. When asked about the total amount of time spent in

the last visit, the spread of data points showed that it is harder to remember amount of time

spent on a specific topic compared to the amount of time spent the on general topics.

When assessing coverage, looking beyond any exposure only (did you ever) to better understand when

(timing/recall period), how often (frequency), how long (duration), what (content), from whom (type of

provider), and where (place of contact) mothers received nutrition counseling served to verify responses

and implementation fidelity in the evaluation study context.

SHINE - Zimbabwe

The SHINE trial took place between

2012 and 2016 in Zimbabwe and was

designed to achieve dietary intake that

meets all nutrient requirements by

children between six and 18 months of

age in clusters randomized to receive

the IYCF interventions, and to prevent

all fecal ingestion by the children in

clusters randomized to receive the

WASH interventions. Exclusive

breastfeeding promotion was

delivered to all children from birth

through 6 months of age as a standard

of care. The cluster-randomized trial with a longitudinal design included counseling interventions that

were delivered by community health workers within the Ministry of Health system. A total of 15

structured intervention delivery contacts were scheduled for all participating households. However,

health workers visited each household monthly to encourage proper IYCF behavior adoption. For all

infants from birth to 6 months, counseling content emphasized exclusive breastfeeding promotion. In

IYCF clusters, counseling content emphasized the promoted optimal use of locally available foods for

complementary feeding after 6 months as well as continued breastfeeding and feeding during child

illness.

Takeaways:

- Recall period: Data collection contacts occurred in at least 3-month intervals with a recall period

of 3 months. This was fit for purpose to assess the appropriateness of messages received

depending on the age of children. The intervention beneficiaries reported high exposure to

intervention messages, and the trends were consistent between the community health workers’

contacts and the expected messages recalled (alluding to high program fidelity). For example,

when asked about having received information or heard about how to properly breast feed their

child, 86% responded in the affirmative at 6 months. At 12 months the proportion was similar

at 81%, but at 18 months 62% recalled having recently heard this. This trend was as intended,

as breastfeeding messages were expected to be replaced with messages on complementary

feeding as the child aged.

Suaahara II - Nepal

Suaahara II is an ongoing 5-

year (2016-2021) multi-sector

integrated nutrition program in

Nepal, which builds on

Suaahara I (2011-2016) and is

designed primarily to improve

the nutritional status of

pregnant and lactating women

and children under two years

of age. Other core elements of

Suaahara II include improving

health and family planning

services, nutrition, water,

sanitation and hygiene, homestead food production (gardening and poultry rearing) and strengthening

nutrition governance. The integrated, multi-sectoral program includes the promotion of key MIYCN

practices as part of an intensive social behavior change strategy, including interpersonal communication,

community events, a weekly interactive radio programs, and more recently, the use of mobile

technology. For this, during the 1000-day period, the idea is that a family would receive 4 interpersonal

contacts via home visits, participate in at least 3 community mobilization events, hear 33 radio program

episodes and receive about 35 mobile texts, to total 75 contact points. The Suaahara II monitoring

system includes both monthly monitoring data system (representative at the district level and collected

internally) as well as annual surveys (representative at the program-wide level and collected by an

external survey firm).

Takeaways:

- Content: Because of the multisectoral nature of the program and number of topics covered,

open-ended questions on content discussed in home visits with field supervisors and heard via

mass media were asked. For community events and mobile technology, the content is

standardized such that just asking about exposure to the platform is sufficient. Asking detailed

questions for each topic would have real implications for time.

- Platform specific: Because the content of ANC is clear in government protocol and limited,

specific questions on whether “x” was a part of counselling during ANC are included.

- Type of service provider, frequency, and other details: Consideration was given to whether the

survey should ask about the specific message, the provider, or the platform, as multiple

combinations were possible and have implications for the length of the questionnaire.

- Lessons learned: Phrasing of “counselling” didn’t work during field testing and thus other

wording “what topics did x discuss with you?” were adopted. Also, triangulation is important

and thus Suahara II monitoring datasets include parallel questions to capture counselling

coverage from the perspectives of health workers and female community health volunteers.

Discussion related to the 3 examples

A discussion on the complexity and difficulty in measuring coverage of counseling interventions followed

the three presentations. Some important points that were raised and proposed to help guide the design

of survey questions include 1) what the set of counseling interventions are that need to be measured,

and 2) what the priority is to ask in core questions of a large-scale, nationally-representative survey for

the purpose of tracking over time and making cross-country comparisons as opposed to questions that

will provide more detail about how MIYCN counseling is delivered , which are more appropriate for

inclusion in a nutrition-specific national survey, add-on modules to a core questionnaire or in evaluation

studies. Moreover, defining what is meant by counseling

is needed since the term counseling may not

be well-understood by all populations and is defined differently across contexts, sometimes referring to

exposure to messages via any mode of communication and other times referring to interpersonal

counseling. Most of the surveys reviewed measure coverage of exposure to certain messages via

interpersonal communication, rather than to the process of counseling and support. Finally, there was

some discussion over what is meant by coverage and who should be included in the denominator of

questions. The goal is to know the proportion of a population of mothers who are receiving counseling

among those who need the particular type of counseling. Furthermore, from the perspective of the

WHO global guidance for breastfeeding programs, the focus is on program delivery and whether the

program is reaching all women of reproductive age with the counseling they need. It is not only about

the messages they heard but also about the interaction between the counselors and caregivers.

Participants also emphasized the distinction between what program evaluations or intervention

research studies and large-scale surveys are trying to measure: in evaluations, measurement is

specifically adapted to a known intervention that is delivered, while for large-scale surveys,

measurement is broader to address multiple existing programs and what is routinely being delivered

from various sources in terms of counseling and support.

3.3. Experiences from large nationally representative surveys

The next topic of discussion was related to experiences from three large nationally representative

surveys: PMA2020, DHS, and MICS.

Survey examples: PMA 2020, DHS and MICS

Performance Monitoring and Accountability 2020 (PMA2020)

PMA2020 started in 2013 as a nationally-representative family planning survey program funded by the

Bill & Melinda Gates Foundation and implemented by the Bill & Melinda Gates

Institute for Population and Reproductive Health at Johns Hopkins Bloomberg School of Public health.

Surveys are implemented in 11 countries through a network of local universities and research institutes.

There is a core survey on family planning and WASH that is implemented every 6-12 months across all

countries, and several modules including maternal and child health (MCH) and a new nutrition module

that have been developed and implemented in select countries.

A nutrition module was tested in Burkina Faso and Kenya over two rounds in 2017 and 2018. The survey

includes questions on a number of nutrition interventions that are not measured in other nationally-

The Global Strategy for IYCF (2003) currently defines breastfeeding counseling as “Mothers should have access to

skilled support to help them initiate and sustain appropriate feeding practices, and to prevent difficulties and

overcome them when they occur”.

representative surveys including four time-points for breastfeeding counseling: during ANC, at

delivery/PNC (within 2 days of birth), within the first month of delivery, and during a sick child visit. The

intervention of interest was whether women received information about breastfeeding at all four time

points and whether breastfeeding was observed by the health care provider around delivery/PNC and

within the first month. Both these elements were included as the stem questions in the survey.

Additional questions are asked about place, provider and key messages. The survey also includes

questions on complementary feeding counseling.

Takeaways:

- Recall period: In Burkina Faso and Kenya, there was the expected relationship between the

specific messages reported and the time point of recall. For example, when asked about specific

messages received during pregnancy, more women reported receiving messages on exclusive

breastfeeding and immediate breastfeeding, compared to the 1-month postpartum period when

most women reported hearing messages on proper positioning and attachment. This result was

as expected, as breastfeeding messages were expected to change as the child aged.

- Content: The survey asked about key messages twice – the first pass using an unprompted recall

and the second pass asking specifically about messages that were not identified on the first

pass. The results show that the unprompted responses were much lower than once prompted –

suggesting that method of asking the question is important for comparability.

Demographic and Health Survey (DHS)

The DHS are household surveys that provide nationally representative data on indicators of

demographics, health and nutrition among other topics. Technical assistance for survey design and

implementation is funded by the United States Agency for International Development (USAID) and

implemented by ICF. DHS have been conducted in more than 70 developing countries in Africa, Asia,

Australasia, Europe and Latin America and the Caribbean. The standard household surveys comprise

sampling of 5000–30 000 households and are usually conducted every 5 years. The DHS surveys usually

cover married women and men aged 15–49 years, although they are tailored to the needs of each

country. There is a core questionnaire – currently DHS-7 - that is implemented across all countries with

minor modifications and is updated approximately every 5 years. Countries can add additional

questions to their specific survey.

DHS-7 includes questions on whether a health care provider provided counseling on breastfeeding and

observed breastfeeding during postnatal care (within first 2 days of birth) (Annex 4). There was

discussion on what background work is required before including additional questions to core

questionnaire. The main criteria that need to be met include: 1) relevance to global indicators that lack

data, 2) in line with USAID priorities, 3) has been validated (not specifically defined), 4) translatable

across different countries, and 5) DHS is the right platform to ask this question.

There were discussions on the lessons learned from the Nepal DHS-7 survey which included additional

questions about MIYCN counseling. The surveys asked about whether a woman received MIYCN-related

advice from a health worker in the last 6 months, and the type of service provider, the timing of this

advice, and the content of the advice (Annex 4). There were three main challenges found in the way

these questions were asked: 1) it was difficult for enumerators and respondents to understand what

MIYCN is, 2) the unprompted nature of the questions made it difficult for enumerators to record the

correct response, especially given the very short training allotted to these questions, and 3) the sample

size for these questions were very small due to the specified recall period (if mothers had a child in the

last 1 year, they were then asked whether they received counseling in the last 6 months), which made it

difficult to draw any conclusions. Though there were challenges, it was evident that Nepal and other

countries, have a high demand to ask about IYCF counseling coverage but need guidance on how to do

so effectively. Though the DHS can be used as a platform to ask these questions, it cannot be used as a

form to test questions. The questions need to be tested prior to incorporation into the DHS.

Multiple Indicator Cluster Survey (MICS)

The MICS survey is implemented by UNICEF. These surveys began in 1995 and are one of the largest

sources of internationally comparable data on women and children. MICS was a major source of data on

the World Summit for Children Goals, Millennium Development Goals (MDG) indicators and will

continue to be a major data source during the 2030 Sustainable Development Agenda to measure

Sustainable Development Goals (SDGs) indicators. Internationally agreed indicators drive the MICS

content and all questions included in the surveys are designed to contribute to an indicator – whether to

the numerator, denominator or for data disaggregation.

When existing indicators are changed, MICS subsequently changes the content of the surveys. An

example is the definition for minimum dietary diversity for children (MDD). The change was agreed at a

July 2017 consultation, and in the current round of MICS the new definition in used in the data

tabulation plan. When new indicators are developed, MICS may include the indicators if they align with

MICS priorities and specified criteria.

There were discussions on the timeline of the MICS surveys. MICS is currently on Round 6 which started

in 2016. It includes a household questionnaire, a questionnaire for males and females aged 15–49 years,

a questionnaire for children under the age of 5 (based on maternal or caretaker reports), and a

questionnaire for 5-17-year-olds. Like the DHS, MICS currently includes questions on whether a health

care provider provided counseling on breastfeeding and observed breastfeeding during postnatal care

(Annex 4).

As MICS prepares for Round 7, it is considering a shorter core set of questionnaires with other questions

kept as optional in a separate category. As MICS Round 6 is ongoing, there are opportunities to include

additional questions for field-testing (in early 2019). New questions need to be validated before being

included in a MICS field test, and new content requires varying degrees of validation. The pilot test for

the full set of MICS questions will take place in early 2020. These pilot tests could potentially yield

results demonstrating that new questions are ready to be included in MICS-7, which is planned to be

launched at the end of 2020.

4. Group discussions about coverage measurement

4.1. Framing the group discussions

Two presentations helped to kick off the group discussions: a presentation on the upcoming global

guidance on breastfeeding counseling programs, and the programmatic experiences of Alive & Thrive.

The upcoming WHO global guidance on breastfeeding counseling programs

In the WHO Global Strategy for IYCF, the current definition of breastfeeding counseling states “Mothers

should have access to skilled support to help them initiate and sustain appropriate feeding practices,

and to prevent difficulties and overcome them when they occur”. It mentions aspects related to who

provides the counseling as well as the time points during which counseling occurs and the aims of

counseling. The WHO is currently providing IYCF counseling training packages that address key IYCF

concepts, but the training does not have a specific target audience. Revisions are currently being made

to the global guidance on breastfeeding counseling programs with regards to updating the definition of

counseling as well as providing details on the timing, frequency, mode of delivery and service providers

of counseling. Current draft recommendations include:

- Breastfeeding counseling should be provided during the ANC period, postnatally and up to 24

months or longer.

- Breastfeeding counseling should be provided at least 6 times and additionally as needed, with

the expectation of at least one contact during ANC, one postpartum related to the baby friendly

hospital initiative, and four other contacts.

- Counseling should be provided through face-to-face contact, and other modes may be used

based on context-specific needs.

- Counseling should be provided as a continuum of care by appropriately trained healthcare

professionals and community-based counsellors.

- Counseling should take the form of anticipatory counseling. It should anticipate and address

important challenges and contexts for breastfeeding in addition to establishing skills,

competencies and confidence among mothers regarding breastfeeding.

- Emphasis on the importance of appropriate and timely support of IYCF, especially breastfeeding

counseling during emergencies.

Considerations from the Alive & Thrive experience

In the context of impact evaluations of Alive & Thrive (A&T), which is a global initiative to improve IYCF

practices, there was heavy investment in developing survey-based measures of the exposure to IYCF

counseling interventions deployed in different countries. In designing intervention-specific coverage

measures, they attempted to capture contact with the countries’ unique delivery platform, exposure to

counseling/messaging via that platform, and frequency/intensity of exposure as per the service

provision protocols in the country. Another approach taken was ‘content-tracing’ where questions were

designed to trace exposure by respondents to very specific content-based messages promoted by A&T.

This resulted in a set of intervention- and context-specific questions that were of relevance to individual

country programs. There was an emphasis on working within the country systems and A&T programs

have used different types of information sharing, depending on the context, from timed and age

appropriate messaging in Ethiopia to standardized interpersonal counseling in Vietnam.

These two presentations led to breakout discussions to consolidate ideas around three key dimensions

of measuring counseling coverage that would also inform the larger global consultation on Measuring

Nutrition in Population-Based Household Surveys and Associated Facility Assessments that occurred the

same week. The three group breakout questions that frames the discussions were:

Question 1: What survey questions can best capture exposure to counseling interventions?

Question 2: What counseling coverage measurement issues should be addressed in an extended

guidance?

Question 3: What are additional research needs for development and validation of survey-based

coverage measures?

The section below summarizes the discussions of each of the groups that discussed these questions and

shared them with the plenary.

4.2. Summary of group discussions

Question 1: What survey questions can best capture exposure to counseling interventions?

The task for this group was to attempt to propose/formulate a very brief set of questions that could be

integrated into a large-scale survey to measure exposure to breastfeeding and complementary feeding

counseling. Participants were asked to draw on insights from the discussions throughout the day,

building on the existing questions in the DHS and MICS questionnaires to ground the discussion.

The group discussed possible options with the intention to add the least number of questions possible

considering the length limit of DHS or MICS. The group agreed that asking about the content of

counseling (e.g. what messages were given, were problems resolved) was too detailed for a DHS or

MICS.

The group agreed that respondents likely have different interpretations of the term “counseling”. On

the other hand, “receiving messages” or “being told something about breastfeeding” does not capture

the interaction that is expected to take place in counseling. The group proposed language such as “Did a

health provider or community worker discuss/talk with you about […]?”

The group suggested to ask about

counseling during pregnancy, in the

first few days after delivery, in the

first month after delivery, and in

the previous 6 months. For the

question on counseling in the

previous 6 months, it was proposed

that differentiating between

counseling related to

complementary feeding and

counseling related to breastfeeding

would be important.

Post-consultation email-based exercise on additional questions

A follow-up exercise with all participants examined the issues related to capturing timing, contact,

content, and provider to develop questions for inclusion in the ANC, PNC and the child health and

nutrition modules of a survey like the DHS. Participants were sent an email the day after the

consultation and asked to indicate their preferences for proposed questions to be asked of caregivers of

children <2 years of age (exercise prompts can be found in Annex 5), which included:

1. During this pregnancy, did a health care provider or community worker talk with you about

breastfeeding? (to be asked with other ANC questions)

2. During the first two days after (NAME)’s birth, did any health care provider do the following:

a. Examine the cord?

b. Measure temperature?

c. Counsel you on danger signs for newborns?

d. Counsel you on breastfeeding?

e. Observe breastfeeding?

3. During the first month after (NAME)’s birth (but after first two days), did a health care provider

or community worker talk with you about breastfeeding?

4. In the last six months, did a health care provider or community worker talk with you about how

to feed your child?

4a. (If yes,) What topics did he or she talk with you about? (list would differentiate topics on

breastfeeding and topics on complementary feeding)

For question 3, some participants believed that the question is not needed since question 4 could

capture this time window. Since the time periods covered in questions 3 and 4 will be overlapping for

mothers of children less than 7 months of age, this could result in double-counting. The majority,

however, believed that counseling in the first month following discharge is a critical window for

breastfeeding problems and deserves a separate indicator.

An alternative formulation of question 4 and 4a would be to ask one question about breastfeeding and

one question about feeding foods and liquids, other than breastmilk. The group was divided on this

question, with a majority favoring the original formulation.

The proposed questions would allow the calculation of the following indicators:

• Proportion of women who received BF counseling in their last pregnancy

• Proportion of women who received BF counseling in the first two days

• Proportion of women who BF counseling in the first month (but after first 2 days)

• Proportion of women who received BF or CF counseling in the last 6 months

• Proportion of women who received at least 4 BF counseling contacts at recommended time

periods

Question 2: What counseling coverage measurement issues should be addressed in an extended

guidance?

The task for this group was to discuss the types of issues that should be covered in a guidance document

on issues to consider in developing robust measures of exposure to nutrition counseling and behavior

change interventions. This covers a more exhaustive set of questions that could be implemented as a

stand-alone survey, an add-on module to national surveys or for use in smaller scale studies and

evaluations so that data are harmonized. The group was asked to help prioritize the dimensions in the

measurement guidance, drawing on the seven dimensions of coverage discussed in relation to

breastfeeding and complementary feeding.

The group agreed that all the elements are important to prioritize and emphasized that asking about

type of service provider is important because it provides more details on where mothers receive

information which could help implementers better target interventions. This is also the case for asking

about location. The group also highlighted the importance of measuring misinformation given during

counseling, exposure to competing messages, e.g. marketing of breast milk substitutes and to provide

guidance on how to measure this. There were discussions on the need for guidance on measuring

specific counseling topics, such as early initiation of breastfeeding as well as the maintenance and

management of supply of breast milk. They also suggested it would be helpful to have guidance and

suggestions for designing on open-ended questions to assess the quality and counseling.

Question 3: What are additional research needs for development and validation of survey-based

coverage measures?

This group was asked to reflect on the issues raised earlier in the day to consider in the development

and validation of survey-based coverage questions and indicators. The group was tasked to reflect on

the issues to examine in validation studies similar to those done in MNCH coverage measurement

research and discuss and summarize what some emerging research needs might be around

development of measures of exposure to breastfeeding and complementary feeding counseling.

The group identified three broad areas of future research:

1. How to select the appropriate indicators and the implications these indicators have for shaping

programs and policies. Such needs will vary based on country context and what indicators

countries need that reflect their own strategies.

2. Unpacking the seven dimensions of coverage - each dimension could have an accompanying

research agenda. The group agreed it is crucial to determine the right time points and recall

periods of questions, and to unpack the different terminology related to counseling and how to

operationalize them (how you ask about service providers, how you phrase the question). The

group also discussed the inevitability of social desirability bias and emphasized the importance

of knowing how to address this as part of future research (an example was provided on using

technology to ask sensitive questions).

3. Research on the different indicators required for different types of surveys. Facility based

surveys may address continuity of service delivery and coverage compared to research

evaluations which are interested in quality and different dimensions of coverage as well as

adoption of practices.

The group also discussed how a guidance on coverage measurement may set expectations that certain

functions of counseling will address IYCF needs. But it is important to also consider other structural

factors in place, such as maternity protection laws and employment options, which also affect IYCF. It is

important to frame the coverage measurement guidance in a way that shows counseling interventions

are one piece of a larger system that works together to improve IYCF.

5. Closing remarks and way forward

The consultation aimed to consolidate insights from research and implementation efforts and facilitate

dialogue that can help make progress towards the development of indicators to measure coverage and

reach of IYCF counseling or behavior change interventions. Throughout the consultation, there were

discussions on the complexity of measuring counseling coverage and the need for validation studies. The

seven elements of counseling coverage were streamlined throughout the discussions and survey case

studies. The working group discussions attempted to address what questions could be included in core

and extended modules on breastfeeding coverage, and existing research needs for the development and

validation of survey-based coverage measures.

There was also a discussion on how data from these modules could be used. Some uses include:

- Provide data to support country reporting on Global Nutrition Monitoring Framework indicator

on coverage of breastfeeding counseling programs. Current GNMF indicator: Proportion of

women with a child <24 months of age who received at least one counseling contact in the last

one year. [NOTE – this indicator can be reformulated with TEAM, and based on data

availability].

- Support the availability of data to report on country programming on BF and CF (which is

currently reported as very high in the Global Nutrition Policy Review) but data gaps preclude

assessment of reach of programs. The proposed indicators are consistent with the draft WHO

guidelines on breastfeeding counseling and capture most of the dimensions described in those

guidelines.

- Coverage monitoring in the context of new large-scale nutrition programs being funded both

nationally and through global financing mechanisms, all of which include programs to support

and promote infant and young child feeding programs.

- Country uses of various types to support programs

Some next steps to move this work forward include:

1. Prepare a meeting report to document discussions, particularly to inform continued guidance

development work by WHO and TEAM. This document serves as that meeting report.

2. Document the review of survey questions and the seven elements of counseling coverage

measurement as a public good such as a discussion paper or manuscript.

3. Identify research opportunities to examine issues related to counseling coverage, such as

secondary analyses of existing datasets and a validation study.

Annex 1. Agenda

TIME

SESSION

PRESENTER

8.30 - 9.00

Registration and breakfast

SESSION 1: WELCOME AND INTRODUCTION

Moderator:

Purnima Menon

9.00 - 9.15

Welcome and meeting background

Silvia Alayon

Rebecca Heidkamp

9.15 - 9.45

Participants’ introductions

9.45 - 9:55

Relevance of coverage measurement in the context of

WHO’s global nutrition monitoring framework

Kuntal Saha

9:55 - 10.30

Keynote address:

Improving coverage measurement in maternal, newborn

and child health

Q&A

Melinda Muños

10.30 - 10.50

Coffee/tea break

SESSION 2: REVIEW OF CURRENT EFFORTS TO MEASURE

IYCF COUNSELLING COVERAGE IN PROGRAM

EVALUATIONS

Moderator:

Silvia Alayon

10.50 - 11.20

Review of coverage indicators for IYCF counseling:

Overview and findings

Sunny Kim

Coverage measurement case studies:

11.20 - 11.30

• Scaling up infant and young child feeding counselling,

Alive & Thrive, Bangladesh

Phuong Nguyen

11.30 - 11.40

• Sanitation and Hygiene to Improve Nutrition and

Enteropathy (SHINE), Zimbabwe

Mduduzi Mbuya

11.40 - 11.50

• Insights from the Suaahara annual monitoring survey

on the reach of counselling services in Nepal

Kenda Cunningham

11.50 - 11.55

Discussant remarks

Melinda Muños

11.55 - 12.20

Q&A/discussion

12.20 - 12.50

Break to pick up lunch

SESSION 3 (WORKING LUNCH): REVIEW OF CURRENT

EFFORTS TO MEASURE IYCF COUNSELLING COVERAGE IN

LARGE-SCALE SURVEYS

Moderator:

Sunny Kim

12.50 - 13.10

Current efforts in large-scale surveys (DHS, MICS, SMART)

and the PMA2020 experience

Rebecca Heidkamp

13.10 - 13.15

Insights on integrating counselling coverage questions into

the DHS

Sorrel Namaste

13.15 - 13.20

Insights on integrating counselling coverage questions into

the MICS

Chika Hayashi

TIME

SESSION

PRESENTER

13.20 - 13.45

Q&A/discussion

SESSION 4: BREAKOUT DISCUSSIONS

Moderator:

Purnima Menon

Framing the discussion:

13.45 - 13.55

• Update on the global guidance on breastfeeding

counselling programs

Laurence Grummer-

Strawn

13.55 - 14.05

• Overview of minimum program elements in Alive &

Thrive’s IYCF counselling/behaviour change

interventions

Silvia Alayon

14.05 - 14.15

Q&A

14.15 - 14.20

Breakout assignments

Coffee/tea available at 14.30

14.20 - 15.20

Discussion questions:

- What survey questions can best capture exposure

to counselling interventions?

- What counselling coverage measurement issues

should be addressed in an extended guidance?

- What are additional research needs for

development and validation of survey-based

coverage measures?

SESSION 5: CLOSING AND NEXT STEPS

Moderator:

Rebecca Heidkamp

15.20 -15.50

Report back on table discussions

Rapporteurs

15.50 -16.45

Reflections and discussion

16.45 -17.00

Summary and closing

Purnima Menon

Ellen Piwoz

Annex 2. List of participants

Silvia Alayon

Monitoring, Learning, and Evaluation Advisor

Alive & Thrive

Jeniece Alvey

Nutrition Advisor

USAID

Rasmi Avula

Research Fellow

International Food Policy Research Institute

Audrey Buckland

Research Associate

Johns Hopkins Bloomberg School of Public Health

Jowel Choufani

Research Analyst

International Food Policy Research Institute

Kenda Cunningham

Suaahara Senior Technical Advisor

Helen Keller International

Jessica Escobar-Alegria

Monitoring, Learning, and Evaluation Advisor

Alive & Thrive

Valerie Flax

Senior Public Health Analyst

RTI International

Edward Frongillo

Professor

University of South Carolina

Laurence Grummer-Strawn

Technical Officer

World Health Organization

Chika Hayashi

Senior Advisor, Monitoring and Statistics

UNICEF

Rebecca Heidkamp

Assistant Scientist

Johns Hopkins Bloomberg School of Public Health

Arja Heustis

Monitoring and Evaluation Associate

PATH

Justine Kavle

Nutrition Team Lead, MCSP

PATH

Sunny Kim

Research Fellow

International Food Policy Research Institute

Monica Kothari

M&E Lead/MQSUN+

PATH

Habtamu Lashtew

Nutrition Director

Save the Children

Mduduzi Mbuya

Senior Technical Specialist

Global Alliance for Improved Nutrition

Vrinda Mehra

Data Analyst

UNICEF

Purnima Menon

Senior Research Fellow

International Food Policy Research Institute

Erin Milner

Nutrition Advisor

USAID

Melinda Munos

Assistant Professor

Johns Hopkins Bloomberg School of Public Health

Sorrel Namaste

Senior Nutrition Technical Advisor

ICF

Phuong Nguyen

Research Fellow

International Food Policy Research Institute

Ellen Piwoz

Senior Program Officer

Bill & Melinda Gates Foundation

Alissa Pries

Technical Advisor

Helen Keller International

Rebecca Robert

Assistant Professor

The Catholic University of America

Tina Sanghvi

Director Programs, Africa

Alive & Thrive

Kuntal Saha

Technical Officer

World Health Organization

Andrew Thorne-Lyman

Associate Scientist

Johns Hopkins Bloomberg School of Public Health

Annex 3. List of studies and surveys reviewed

Project

Country

PI institution

Survey round

Alive & Thrive

Bangladesh

IFPRI

Endline 2014

Alive & Thrive

Vietnam

IFPRI

Endline 2014

Alive & Thrive

Ethiopia

IFPRI

Endline 2017

Alive & Thrive

Burkina Faso

LSHTM

Endline 2017

Alive & Thrive (maternal nutrition

study)

Bangladesh

IFPRI

Endline 2016

Alive & Thrive (maternal nutrition

study)

India

IFPRI

Baseline 2017

CAS (Common App Software)

India

IFPRI

Process Evaluation

2017

Convergence

India

IFPRI

2014

Mama SASHA

Kenya

Emory

Endline 2014

PM2A

Guatemala,

Burundi

IFPRI

Endline 2014

PROMIS

Burkina Faso, Mali

IFPRI

Endline 2017

SELEVER

Burkina Faso

IFPRI

Endline 2017

SHINE

Zimbabwe

Zvitambo,

JHU

Baseline - 18mo visit

2016

TMRI

Bangladesh

IFPRI

Endline 2014

TRAIN

Bangladesh

IFPRI

Baseline 2016

WINGS

India

IFPRI

Midline 2017

PMA2020

Burkina Faso,

Kenya

JHU

Round 2 2018

DHS, MICS

Various

ICF, UNICEF

Annex 4. Questions about IYCF advice in last 6 months

DHS questions on breastfeeding counseling and observation as part of PNC question set

Nepal DHS-7 questions on breastfeeding counseling and observation in the last 6 months

MICS questions on breastfeeding counseling and observation as part of PNC question set

PN25. During the first two days after birth, did any

health care provider do any of the following either at

home or at a facility:

[A] Examine (name)’s cord?

[B] Take the temperature of (name)?

[C] Counsel you on breastfeeding?

YES NO DK

EXAMINE THE CORD ....................... 1 2 8

TAKE TEMPERATURE...................... 1 2 8

COUNSEL ON BREASTFEEDING .... 1 2 8

PN26. Check MN36: Was child ever breastfed?

YES, MN36=1 ......................................................... 1

NO, MN36=2 ........................................................... 2



PN28

PN27. Observe (name)’s breastfeeding?

YES NO DK

OBSERVE BREASTFEEDING ............ 1 2 8

Annex 5. Exercise on additional survey questions on counseling

coverage

Note: this exercise was circulated to all participants in a follow-up email after the meeting

*Proposed questions highlighted in yellow.

**Option 2 formulation for IYCF coverage in first two years is in green.

SECTION 4. PREGNANCY AND POSTNATAL CARE

NO.

QUESTIONS

LAST BIRTH, NAME/

CODING CATEGORIES

SKIP

4xx

During this pregnancy, did a health care provider

or community worker talk with you about

breastfeeding?

YES

DON’T KNOW

457

During the first two days after (NAME)’s birth, did

any health care provider do the following:

a) Examine the cord?

b) Measure temperature?

c) Counsel you on danger signs for

newborns?

d) Counsel you on breastfeeding?

e) Observe breastfeeding?

YES

DON’T KNOW

4xx

During the first month after (NAME)’s birth (but

after first two days), did a health care provider or

community worker talk with you about

breastfeeding?

YES

DON’T KNOW

SECTION 6. CHILD HEALTH AND NUTRITION

OPTION 1: TWO TOPIC-SPECIFIC QUESTIONS – ONE ON BF AND ONE ON CF

6xx

In the last six months, did a health care provider

or community worker talk with you about

breastfeeding?

YES

DON’T KNOW

6xx

In the last six months, did a health care provider

or community worker talk with you about how to

feed your child foods and liquids, other than

breastmilk?

YES

DON’T KNOW

OPTION 2: 1 LEAD-IN QUESTION AND THEN 1 FOLLOW-UP QUESTION ON CONTENT

6xx

In the last six months, did a health care provider

or community worker talk with you about how to

feed your child?

YES

DON’T KNOW

NO… 2

(SKIP TO 6xx)

6xx

What topics did he or she talk to you about?

1) BREASTFEEDING

2) NOT GIVING

WATER IN THE

FIRST SIX

MONTHS OF

LIFE

3) FEEDING OTHER

FOODS

STARTING AT 6

MONTHS OF

AGE

4) FEEDING A

VARIETY OF

FOODS

5) FEEDING

ANIMAL

SOURCE FOODS

6) HANDWASHING

BEFORE

FEEDING

7) **TOPIC LIST

CAN BE

REDUCED OR

EXPANDED**

Scenarios for indicator creation using the above questions in a survey instrument:

There are two potential scenarios for indicator creation using these questions in a survey instrument – one with

the 1 month postpartum included and one without that included. We demonstrate these below, highlighting the

options for indicators using the one month postpartum and not using the one month postpartum contact question.

6 contacts:

Pregnancy/ANC

2 days/PNC

1-2 weeks

3-4 months

5/6 months

6-23 months

Questions:

During last

pregnancy…

During the first

2 days…

During first

month of life

(but after first

2 days)…

In the last 6

months…

If child age

3 months

Y/N

(overlaps

with

pregnancy,

2 days, 1-2

wk)

11 months

Y/N

(any time

when child

was 5-11

mo)

Y/N

(anytime

during child

age 5-11 mo)

Indicator

that can be

created:

Proportion of

women who

received BF

counseling in

their last

pregnancy

Proportion of

women who

received BF

counseling in

the first two

days

Proportion of

women who BF

counseling in

the first month

(but after first

2 days)

Proportion of women who received BF or CF

counseling in the last 6 months

Questions:

During last

pregnancy…

During the first

two days…

In the last six

months…

If child age

3 months

Y/N

(overlaps

with

pregnancy,

2 days)

11 months

Y/N

(anytime

during child

age 5-11 mo)

Indicator

that can be

created:

Proportion of

women who

received BF

counseling in

their last

pregnancy

Proportion of

women who

received BF

counseling in

the first two

days

Proportion of women who received BF or CF counseling in the

last 6 months