Evaluation of the Cambodia Nutrition Project

Last registered on August 19, 2021

Pre-Trial

Trial Information

General Information

Title
Evaluation of the Cambodia Nutrition Project
RCT ID
AEARCTR-0008086
Initial registration date
August 16, 2021

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
August 19, 2021, 10:56 AM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Primary Investigator

Affiliation
World Bank

Other Primary Investigator(s)

PI Affiliation
World Bank
PI Affiliation
World Bank

Additional Trial Information

Status
In development
Start date
2021-08-18
End date
2025-02-18
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
The Cambodia Nutrition Project (CNP) is a five-year project to anchor the RGC’s enhanced and coordinated response to accelerate the human capital formation focusing on MCHN in the early years. The CNP addresses Cambodia’s persistent high levels of stunting and aims to close equity gaps in nutrition, routine immunization, and neonatal survival.
The project will enhance supply-side delivery and quality of priority health and nutrition services in health facility and health outreach activities. It will also extend service provision beyond health facilities through enhanced integrated health outreach activities for pregnant and lactating women children and a harmonized approach to engaging local government for community mobilization and service delivery for health and nutrition. In so doing, the project will aim to increase demand for priority services. The project will build capacity from the central to the frontline levels in functional, technical, and financial management aspects.
The CNP’s impact evaluation aims to quantify and attribute the plausible impact of the project interventions in achieving changes in priority health and nutrition outcomes and to document contextual factors that influence the operational effectiveness. To answer these questions, the study follows a matched difference-in-differences design. Administrative districts (ADs) receiving the CNP intervention were randomly allocated into two groups implementing the intervention in two phases. Control ADs were matched based on key outcomes prior to the start of the baseline.
The study will deploy repeated cross-sections of mothers of children and children. Three survey rounds will be conducted to collect both longitudinal and cross-sectional data: a baseline survey, a midline survey one year after intervention rollout, and an endline survey after about 3.5 years of intervention.
External Link(s)

Registration Citation

Citation
Alderman, Harold, Jed Friedman and Anne Provo. 2021. "Evaluation of the Cambodia Nutrition Project." AEA RCT Registry. August 19. https://doi.org/10.1257/rct.8086-1.0
Sponsors & Partners

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information
Experimental Details

Interventions

Intervention(s)
The Cambodia Nutrition Project (CNP) is a five-year project to anchor the RGC’s enhanced and coordinated response to accelerate the human capital formation focusing on MCHN in the early years. The CNP addresses Cambodia’s persistent high levels of stunting and aims to close equity gaps in nutrition, routine immunization, and neonatal survival.
The project will enhance supply-side delivery and quality of priority health and nutrition services in health facility and health outreach activities. It will also extend service provision beyond health facilities through enhanced integrated health outreach activities for pregnant and lactating women children and a harmonized approach to community mobilization and service delivery for health and nutrition. In so doing, the project will aim to increase demand for priority services. The project will also build capacity from the central to the frontline levels in functional, technical, and financial management aspects.
Two main platforms are used to deliver CNP interventions: health facilities and communities (C/S administrations). For simplicity, the key interventions across these components are summarized as health facility and community-based interventions:
Health Facility-Based Interventions - Key interventions include
o Cambodia has a national performance-based financing scheme in which all health centers nationwide now receive quarterly assessments through the NQEM process assessment accompanied by a performance-based grant Service Delivery Grants (SDGs). The project supports quality improvement in priority services through the addition of a MCHN Scorecard to the quarterly National Quality Enhancement Monitoring (NQEM) Tools the addition of systematic coaching, and an accompanying SDG top-up payment. Performance on the scorecard will be accompanied by a grant to health centers, paid on a quarterly basis, which can be allocated to individual staff incentives (up to 80% of total grant amount) and operational costs. The staff incentives are distributed across all health centers staff according to individual staff performance and other criteria determined at the health center level.
o Expansion of the Health Equity Fund (HEF) system to provide transport allowances for poor women and children to use health facilities for antenatal care, deliver, postnatal care, and well-child visits; and expand HEF benefits to cover children of informal workers to receive fee waivers for priority services
o Upgrading the package of social behavior change communication materials, training, and coaching to be provided to health centers

Community-Based Interventions - Key intervention includes:
o Development and delivery of group- and home-based health, nutrition, and HEF promotion activities through a commune program for women and children (CPWC)
o Performance-based grants to C/S for effectively delivering the CPWC
o Update of the package of Social and behavior change communication materials, C/S and village-level coaching, and training to be provided in communities
The project’s primary beneficiaries are women and children: pregnant and lactating women and children under the age of two (e.g. the first 1,000 days of life). The project prioritizes seven of Cambodia’s 25 provinces: Mondul Kiri, Ratanak Kiri, Kratie, Steung Treng, Preah Vihear, Kampong Chhnang, and Koh Kong. The seven priority provinces were purposively selected through multi-stakeholder consultation based upon the following characteristics: lagging Reproductive, Maternal, Newborn, Child and Adolescent Health and Nutrition (RMNCAH-N) outcomes, coverage and equity gaps of key RMNCAH-N services, multidimensional poverty index score, and gaps in supply side health service readiness.

Intervention Start Date
2021-10-29
Intervention End Date
2025-02-18

Primary Outcomes

Primary Outcomes (end points)
• MCHN service quality score in health centers
• Height-for-age Z-score (HAZ) (children 12-36 months)
• Stunting (% HAZ <-2 SD) (children 12-36 months)
• Weight-for-age Z-score (WAZ) (children 12-36 months)
• Underweight (% WAZ <-2 SD) (children 12-36 months)
• Minimum dietary diversity (mean number of food groups and % infants and young children 6-23 months consuming at least 4/5* food groups in the past 24 hours)
• Fully immunized children (% children 12-23 months)
Primary Outcomes (explanation)
The MCHN Service quality score will be based on the checklist tool, developed by the Cambodian Ministry of Health and replicated in the independent evaluation. The quality score will consist of structural, process, and outcome quality as per the MOH scorecard. Sub-scores for each of the three domains above will be generated to measure improvement in the service quality domains in addition to overall quality.

Secondary Outcomes

Secondary Outcomes (end points)
• Age-appropriate child development outcomes (CREDI or similar)
• Early initiation of breastfeeding (children born within past 24 months)
• Minimum meal frequency (% infants and young children 6-23 months)
• Minimum acceptable diet (%infants and young children 6-23 months)
• Introduction of solid, semi-solid or soft foods (% children 6-8.9 months)
• Consumption of non-milk animal-source (egg and/or flesh foods) during the previous day (% children aged 6–23 months who consumed
• Receipt of DPT-HepB-Hib 3 (% children 12-23 months)
• Birth weight (caregiver reported)
• Underweight mothers of children <2 (% mothers aged 15-49 years with BMI <18.5)
• Receipt of 4+ ANC visits (mothers of children age 6-36 months)
• Receipt of 4 PNC visits (mothers of children age 6-36 months)
• Health Center provision of integrated outreach according to guideline
• Participation in GMP according to minimum package
Secondary Outcomes (explanation)
These secondary outcomes relate to a number of key maternal, infant and young child nutrition behaviors and services utilization which would be expected to be influenced by the study but are not primary outcomes.

Experimental Design

Experimental Design
The impact evaluation of the Cambodia Nutrition Project will focus on the following primary research question:
1. What is the combined impact of interventions delivered through health facilities and communities on priority outcomes of interest compared to the standard of care (e.g. those not receiving similar interventions)?
The gradual expansion of community-based interventions will allow for opportunistic exploration of two secondary research questions:
2. What is the net impact of health facility-based interventions on MCHN service quality and utilization when compared to the standard of care?
3. Is there a differential impact of the health facility-based interventions on the priority outcomes of interest for those receiving the community-based interventions compared to those not receiving the community-based interventions?
A plausibility assessment is proposed to answer our first and second research questions, aiming to demonstrate that the impact achieved is likely due to the influence of CNP, above and beyond other external influences. Based on CNP’s scope, design and phased implementation, a quasi-experimental, matched difference-in-differences was selected.
The study will deploy repeated cross-sections of mothers of children age 6-36 months and children ages 12-36 months. In total, three survey rounds will be conducted to collect both longitudinal and cross-sectional data: a baseline survey at the beginning of the project, a midline survey approximately one year after intervention rollout, and an endline survey after about 3.5 years of intervention. The ADs in the study are organized according to whether they received: no intervention (Group A), only health facility interventions at baseline (Group B) or the combined health facility and community-based interventions since baseline (Group C).
Evaluation Phase 1 - Baseline to Midline: Taking advantage of the phased rollout of the community-based interventions, this phase is an opportunity to perform a randomized evaluation of the relative effectiveness of health + community-based interventions against the health facility-based interventions only (a comparison of Group C to Group B). This phase serves as a learning experience for implementation of the community-based interventions during which start-up and operational issues are being resolved. Data from the midline survey will be used to assess the impact of the CNP health facility-based interventions on service quality and utilization (a comparison of Group C to Group A) and to provide key process information related to the programmatic coverage and quality.
Evaluation Phase 2 - Midline to Endline: The beginning of Phase 2 (midline) constitutes the initiation of the evaluation primary research question to compare Group B to Group A. Impact will be assessed based on the progress between midline and endline. Phase 2 aligns with the introduction of revised behaviour change communication training, coaching, and materials that may have the potential to modify the community-based intervention’s effectiveness. Therefore, the main evaluation will assess the impact of the “tried and improved” intervention approach by comparing data from the midline to the endline. Secondarily, the evaluation will analyse progress over the period baseline-endline for each Group B and (those ADs initially allocated to) Group C, though this will be to inform project feedback and cost-effectiveness rather than impact.
Experimental Design Details
Not available
Randomization Method
Within the seven CNP provinces, seven ODs were selected for inclusion in the study using stratified random sampling, ensuring that the following criteria were met:
• At least three large ODs (contain > 3 ADs)
• At least three small ODs (containing 1-3 ADs)
• Each province with >1 OD is represented in the final sample:
After selecting the ODs, 12 ADs were randomized to receive the health facility-based interventions alone in Phase 1 and 12 ADs to receive the health facility and community-based interventions in the same Phase. The ADs were randomly assigned to Group B or Group C according to the following procedure:
a) Stratify by OD size, so that large ODs contribute 2 ADs to Group B and 2 ADs to group C, while smaller ADs will contribute one AD to each arm;
(b) Stratify also by urban/rural status so that an equal proportion of urban ADs are in both Groups B and C;
(c) Pursue iterated randomization until the standardized mean difference in four population health measures from the 2014 DHS – child HAZ, WAZ, dietary diversity among children 6-23 months old, child full immunization—do not exceed a value of 0.10 each.
From within the AD, all health centers were selected and 4 villages from each health center catchment area, thus constituting the primary sampling unit (PSU). This randomization was done using Probability proportional to size, based on the number of households in each village.

Randomization Unit
The study randomized the villages in each health center’s catchment area. Thus, the village constitutes the primary sampling unit. In addition, the study randomized the administrative district for intervention groups B and C, and thus the allocation of the community-based interventions.
Was the treatment clustered?
Yes

Experiment Characteristics

Sample size: planned number of clusters
48 Administrative Districts in 15 provinces
144 Health centers
576 villages
5760 households
Sample size: planned number of observations
5760 households 144 Health centers
Sample size (or number of clusters) by treatment arms
Control: 24 ADs; 72 HCs; 288 villages, 2880 households
Intervention group B (Health facility-based interventions baseline + community interventions midline): 12 ADs; 36 HCs; 144 villages, 1440 households
Intervention group C (Health facility-based interventions baseline + community interventions since baseline): 12 ADs; 36 HCs; 144 villages, 1440 households
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Standardized Minimum Detectible Effect Size Outcome 1: height for age (HAZ) (12-35 months) - 0.13 Outcome 2: weight for age (WAZ) (12-35 months) - 0.13 Outcome 3: Full Immunization rate (12-23 months) - 0.15 Outcome 4: Child diet diversity (6-23 months) - 0.13
IRB

Institutional Review Boards (IRBs)

IRB Name
National Ethics Committee for Health Research(NECHR)
IRB Approval Date
2021-02-17
IRB Approval Number
N/A