Optimizing Community Health Worker Resources to Scale Child Nutrition Programs in Rural Pakistan

How to cite

Afzal, Uzma; Shonchoy, Abu; Akram, Agha Ali; Kurosaki, Takashi; Khan, Akib. 2025. "Optimizing Community Health Worker Resources to Scale Child Nutrition Programs in Rural Pakistan." AEA RCT Registry. AEARCTR-0016595. https://doi.org/10.1257/rct.16595-2.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Uzma Afzal* , Abu Shonchoy , Agha Ali Akram , Takashi Kurosaki , Akib Khan

* Primary investigator

Registry ID
AEARCTR-0016595
DOI
https://doi.org/10.1257/rct.16595-2.0
First registered
Last updated
Status
Completed
Trial dates
2022-12-01 – 2025-12-31
Countries
Pakistan (Sindh)
JEL codes
I15, I18, I38, J13, O15, C93, D91
Registry record
http://www.socialscienceregistry.org/trials/16595

Abstract

This study evaluates home-based growth monitoring (HBGM) as a strategy to reduce child stunting in rural Pakistan by optimizing community health worker —locally known as community resource person (CRP)— resources. Pakistan has one of the highest regional stunting rates, with rural Sindh particularly affected. Building on evidence from an earlier urban trial in Karachi, we implement a cluster-randomized controlled trial in 200 villages, sampling 15 households per village. Villages are assigned to a control group, Treatment 1 (HBGM with monthly Lady Health Worker visits and counselling), or Treatment 2 (HBGM-light with initial counselling and biweekly automated voice messages). The study examines the effects of the treatments on stunting, wasting, caregiver knowledge, dietary and hygiene practices, and gender equity. It also explores scalability and cost-effectiveness of HBGM as an instrument for improving child health outcomes.

Intervention

The two treatments or interventions were as follows:
HBGM: A CRP (Community Resource Person) installed the growth-poster inside the home, counseled the caregiver on how to use it to measure their child's height, and revisited the household once every month for one year to support the caregiver in measurement, and provided continued counseling on optimal nutritional practices, and recorded the anthropometric measurement in a booklet as an administrative record; and
HBGM-light: A CRP installed the growth-poster inside the home, counseled the caregiver on how to use it and on optimal nutritional practices during the first visit, followed-up with the mother the next month in a second visit to remind caregivers to regularly use the growth chart and reinforce their knowledge on how to use it. In addition to this, the first visit by the CRP was followed by automated voice calls every 2 weeks for the rest of the treatment period to remind the caregiver about using the poster, encouraging regular visits to the nearby health facility for height/weight measurement by a healthcare professional, and conveying key messages on best nutritional practices.

Experimental design

We proposed a cluster-RCT with households in rural Sindh testing two versions of home-based-growth-monitoring (HBGM) vis-a-vis a pure-control group (status-quo). The experimental arms and corresponding sample sizes were as follows, with the sample size for each cluster/village being ~15 children who were 6-18 months old at baseline. Total sampling frame consisted of 3000 children across 200 villages, while the actual sample size comprised 2848 children:
Control: 80 villages (1,200 children, 1129 sampled)
Treatment 1 – HBGM: 60 villages (900 children, 861 sampled)
Treatment 2 – HBGM-light: 60 villages (900 children, 858 sampled)

Design details

Randomization method
There were a total of 285 villages in our selected district as per the RSPN (our field partner) definition of a village. We restricted the sample frame to villages that lay within +/- 1 SD of the mean values of the following selection criteria: 1. Average distance from a health facility within a village 2. Average proportion of households eligible for Benazir Income Support Program (BISP) (The eligibility criterion to receive BISP support defined by the government is any household having a PMT score of less than 24) 3. Average number of households in a village This was to ensure that we did not draw any village that was: ● Either too far away from the nearest health facility, OR ● Had extremely poor or extremely rich households, OR ● Had too few households to select from Our sampling frame was thus restricted to 270 out of 285 villages, from which we randomly drew 200 villages for the study.
Randomization unit
Village level
Clusters
200 villages. Control — 80 villages Treatment 1 – HBGM: 60 villages Treatment 2 – HBGM-light: 60 villages
Observations
2848
Arms
Control —1129 individuals Treatment 1 – HBGM: 861 individuals Treatment 2 – HBGM-light: 858 individuals
Primary outcomes
Height-for-age z-score (HAZ): This measures a child’s linear growth. It compares a child's height to the median height of a healthy reference population of the same age and sex, as established by the WHO Child Growth Standards (WHO 2006). Stunting: Stunting is a binary outcome, taking on a value of 1 if the child’s height-for-age z-score is below -2.
Secondary outcomes
Weight-for-height z-score (WHZ): This measures a child's body proportion or "wasting," indicating acute malnutrition. It compares a child's weight to the median weight of a healthy reference population of the same height and sex, as established by the WHO Child Growth Standards (WHO 2006). Weight-for-age z-score (WAZ): This measures a child's overall body mass relative to their age, reflecting both acute and chronic malnutrition. It compares a child's weight to the median weight of a healthy reference population of the same age and sex, as established by the WHO Child Growth Standards (WHO 2006) for children 0-5 years. Child Health History: Captures recent child health interventions and illness episodes in line with WHO (2006) child health indicators, including Vitamin A supplementation, deworming, iron intake, diarrheal prevalence and management (e.g., ORS, zinc), fever, and acute respiratory infection symptoms, as well as care-seeking behavior, treatment sources, timeliness, and medications used. Underweight: Underweight is a binary variable, taking on a value of 1 if the child’s weight-for-age z-score is below -2. Wasting: Wasting is a binary variable, taking on a value of 1 if the child’s weight-for-height z-score is below -2.

Cited by

No citation data available for this registration yet.

Keywords