Abstract
BACKGROUND: Group antenatal (G-ANC) care has been introduced in Kenya and is associated with increased ANC contacts. Previous studies did not report higher likelihood of facility-based deliveries, where facility delivery rates are high (> 80%). It is unknown whether G-ANC influences women's choice of delivery facility. This study sought to understand if and how exposure to and experience of G-ANC, among other factors, influence women's decisions on which facility to deliver. Study findings can inform health system strategies including the Service Delivery Re-design. METHODS: We conducted a cross-sectional survey from December 2023 to January 2024 in eight health centers in Machakos County, Kenya with women who participated in G-ANC and had a facility delivery to understand women's choice of health facility for childbirth, i.e., whether they returned to the same G-ANC facility for delivery or went to another facility. The survey interviews were administered through telephone, as a nested component of a larger implementation research focused on the adoption of G-ANC. RESULTS: Of the 470 women who participated in the phone survey, 29.8% of women returned to deliver at the facility where they received G-ANC and 70.2% delivered elsewhere. Most women (84.3%) delivered in the public sector. Regression analysis models revealed two significant predictors for delivery in the same G-ANC facility: proximity to the facility (adjusted AOR 4.13, 95% CI: 2.73-6.23, p < 0.001); and positive staff attitudes (adjusted AOR 4.68, 95% CI: 2.06-10.60, p < 0.001). Two significant predictors of delivering in a different facility were being "told/advised to give birth there during pregnancy/ANC" (adjusted AOR 0.23, 95% CI: 0.15-0.35, p < 0.001) and high household wealth status (quintile four AOR 0.29, 95% CI: 0.09-0.91, p = 0.034; quintile five AOR 0.23, 95% CI: 0.07-0.72, p = 0.011). Intervention exposure was not significant. CONCLUSION: Most women participating in G-ANC chose to deliver in another facility. The choice of facility for childbirth was most strongly influenced by factors other than the intervention, such as proximity, positive staff attitudes, health advice, and household wealth status.