534 - Impact of the Community-based Health Planning and Services Program (CHPS+) on maternal care seeking behaviors in Ghana

Poster Number: 534
Publication Number: 534.318

Nadir Ijaz, Columbia University Medical Center, New York, NY, United States; Elizabeth Jackson, Columbia University Department of Population and Family Health, New York, NY, United States; James F. Phillips, Columbia University Mailman School of Public Health, New York, NY, United States; Koku Awoonor, Public Health Consultant, Accra, Greater Accra, Ghana; Ayaga A. Bawah, University of Ghana, Accra, Greater Accra, Ghana; S Patrick Kachur, Columbia University, New York, NY, United States

Background: Two decades ago, the Ghana Health Service adopted a policy known as Community-based Health Planning and Services (CHPS) to expand primary health care nationwide. CHPS interventions include community health nurse training and deployment to local health posts, provision of childhood vaccines and integrated management of childhood illness, and systems planning with traditional leaders and community health committees. In sequential plausibility and replication studies in some of Ghana’s most impoverished and rural districts, CHPS implementation was associated with a substantial reduction in under-5 mortality. More recently, in 2015-2020, the CHPS+ program examined the implementation of CHPS interventions in two additional regions.

Objective: This secondary analysis aims to assess the effects of CHPS+ on maternal care seeking behaviors, both for preventative and curative care.

Design/Methods: Within the Northern and Volta Regions, twelve districts were assigned to one of three intervention groups and an additional twelve districts served as comparison areas. Primary data were collected via field interviews both in the pre-implementation baseline period (2017) and post-implementation endline period (2020). We used difference-in-difference (DiD) estimation methodology to compare selected endpoints pre- and post-CHPS+ implementation between intervention and comparison groups. Multivariate logistic regression models were created to determine predictors of maternal care seeking for childhood illness. Models controlled for child, maternal, household, and socioeconomic covariates.

Results: 15,819 mothers of 23,541 children under 5 years from 11,429 unique households were included. Over the study period, 31.1% of children had cough, diarrhea, or fever in the preceding two weeks. 59.0% of mothers in the endline period reported seeking care for an ill child at the health facility compared to 54.7% in the baseline period (p=0.003). In a controlled regression model, mothers in intervention districts were more likely to report seeking care for an ill child than those in comparison districts (DiD coefficient 1.4, 95% CI 1.1-1.8). For each kilometer increase in distance from a health facility, mothers were 6% less likely to report seeking care (95% CI 0.92-0.97).Conclusion(s): CHPS+ program implementation increased maternal care seeking for childhood illness, which may be one pathway through which CHPS can reduce child mortality. Future research should focus on the distillation and implementation of other core CHPS intervention components that improve child health.