Effect of digitizing Community Health Information System on the delivery and use of maternal and child health services: Propensity score matching analysis

Abstract

Introduction In 2018, Ethiopia’s government digitized its Community Health Information System (eCHIS) to enhance Health Extension Program service delivery. However, the impact of eCHIS on health outcomes remains unclear. This study, therefore, examined the effects of eCHIS on the uptake of maternal and child health services. Methods A post-test-only, non-equivalent group household survey design was used, collecting data from July to August 2024 in rural communities with and without eCHIS. A stratified multistage sampling technique was employed to recruit respondents. A sample of 1,728 women of reproductive age (271 in eCHIS and 1,457 in non-eCHIS woredas), 1,118 women with children ages 0–11 months (188 in eCHIS and 930 in non-eCHIS), and 569 women with children ages 12–23 months (301 in eCHIS and 268 in non-eCHIS) were included in this study. Propensity scores were used to match intervention and comparison communities based on women’s age, religion, parity, education, household wealth, distance to health facilities, and autonomy. A modified Poisson regression analysis, adjusting for covariates, was conducted to estimate the adjusted prevalence ratios (PR) for maternal and child health outcomes. Results Modern contraceptive use was 66% in intervention areas versus 56% in comparison areas. Institutional delivery rates were 95% in intervention and 79% in comparison areas. In intervention areas, 94% of children received the first dose of the Pentavalent vaccine compared to 83% in comparison areas, while 85% and 68%, respectively, received the third dose. Adjusted analyses showed that eCHIS intervention areas had statistically significant increases in contraceptive use (13.0 percentage points; PR: 1.25; p-value < 0.01) and institutional deliveries (8.9 percentage points; PR: 1.18; p-value < 0.01) compared to comparison areas. Similarly, the average treatment effect of eCHIS on Pentavalent 1, Pentavalent 3, and full vaccination coverage showed significant increases of 10.3 (PR: 1.13; p-value < 0.01), 13.6 (PR: 1.26; p-value < 0.01), and 14.6 (PR: 1.36; p-value < 0.01) percentage points, respectively. Conclusion This study underscores the transformative potential of eCHIS in improving maternal and child health outcomes in Ethiopia. Sustained investment in digital health systems can help scale effective health care services to rural and underserved communities.