Log in to save to my catalogue

Evaluation of a community health worker home visit intervention to improve child development in Sout...

Evaluation of a community health worker home visit intervention to improve child development in Sout...

https://devfeature-collection.sl.nsw.gov.au/record/TN_cdi_plos_journals_2814430214

Evaluation of a community health worker home visit intervention to improve child development in South Africa: A cluster-randomized controlled trial

About this item

Full title

Evaluation of a community health worker home visit intervention to improve child development in South Africa: A cluster-randomized controlled trial

Publisher

United States: Public Library of Science

Journal title

PLoS medicine, 2023-04, Vol.20 (4), p.e1004222-e1004222

Language

English

Formats

Publication information

Publisher

United States: Public Library of Science

More information

Scope and Contents

Contents

Effective integration of home visit interventions focused on early childhood development into existing service platforms is important for expanding access in low- and middle-income countries (LMICs). We designed and evaluated a home visit intervention integrated into community health worker (CHW) operations in South Africa.
We conducted a cluster-randomized controlled trial in Limpopo Province, South Africa. CHWs operating in ward-based outreach teams (WBOTs; clusters) and caregiver-child dyads they served were randomized to the intervention or control group. Group assignment was masked from all data collectors. Dyads were eligible if they resided within a participating CHW catchment area, the caregiver was at least 18 years old, and the child was born after December 15, 2017. Intervention CHWs were trained on a job aid that included content on child health, nutrition, developmental milestones, and encouragement to engage in developmentally appropriate play-based activities, for use during regular monthly home visits with caregivers of children under 2 years of age. Control CHWs provided the local standard of care. Household surveys were administered to the full study sample at baseline and endline. Data were collected on household demographics and assets; caregiver engagement; and child diet, anthropometry, and development scores. In a subsample of children, electroencephalography (EEG) and eye-tracking measures of neural function were assessed at a lab concurrent with endline and at 2 interim time points. Primary outcomes were as follows: height-for-age z-scores (HAZs) and stunting; child development scores measured using the Malawi Developmental Assessment Tool (MDAT); EEG absolute gamma and total power; relative EEG gamma power; and saccadic reaction time (SRT)-an eye-tracking measure of visual processing speed. In the main analysis, unadjusted and adjusted impacts were estimated using intention-to-treat analysis. Adjusted models included a set of demographic covariates measured at baseline. On September 1, 2017, we randomly assigned 51 clusters to intervention (26 clusters, 607 caregiver-child dyads) or control (25 clusters, 488 caregiver-child dyads). At endline (last assessment June 11, 2021), 432 dyads (71%) in 26 clusters remained in the intervention group, and 332 dyads (68%) in 25 clusters remained in the control group. In total, 316 dyads attended the first lab visit, 316 dyads the second lab visit, and 284 dyads the third lab visit. In adjusted models, the intervention had no significant impact on HAZ (adjusted mean difference (aMD) 0.11 [95% confidence interval (CI): -0.07, 0.30]; p = 0.220) or stunting (adjusted odds ratio (aOR) 0.63 [0.32, 1.25]; p = 0.184), nor did the intervention significantly impact gross motor skills (aMD 0.04 [-0.15, 0.24]; p = 0.656), fine motor skills (aMD -0.04 [-0.19, 0.11]; p = 0.610), language skills (aMD -0.02 [-0.18, 0.14]; p = 0.820), or social-emotional skills (aMD -0.02 [-0.20, 0.16]; p = 0.816). In the lab subsample, the intervention had a significant impact on SRT (aMD -7.13 [-12.69, -1.58]; p = 0.012), absolute EEG gamma power (aMD -0.14 [-0.24, -0.04]; p = 0.005), and total EEG power (aMD -0.15 [-0.23, -0.08]; p < 0.001), and no significant impact on relative gamma power (aMD 0.02 [-0.78, 0.83]; p = 0.959). While the effect on SRT was observed at the first 2 lab visits, it was no longer present at the third visit, which coincided with the overall endline assessment. At the end of the first year of the intervention period, 43% of CHWs adhered to monthly home visits. Due to the COVID-19 pandemic, we were not able to assess outcomes until 1 year after the end of the intervention period.
While the home visit intervention did not significantly impact linear growth or skills, we found significant improvement in SRT. This study contributes to a growing literature documenting the positive effects of home visit interventions on child development in LMICs. This study also demonstrates the feasibility of collecting markers of neural function like EEG power and SRT in low-resource settings.
PACTR 201710002683810; https://pactr.samrc.ac.za/TrialDisplay.aspx?TrialID=2683; South African Clinical Trials Registry, SANCTR 4407....

Alternative Titles

Full title

Evaluation of a community health worker home visit intervention to improve child development in South Africa: A cluster-randomized controlled trial

Identifiers

Primary Identifiers

Record Identifier

TN_cdi_plos_journals_2814430214

Permalink

https://devfeature-collection.sl.nsw.gov.au/record/TN_cdi_plos_journals_2814430214

Other Identifiers

ISSN

1549-1676,1549-1277

E-ISSN

1549-1676

DOI

10.1371/journal.pmed.1004222

How to access this item