Digital communication between mothers and community health workers to support neonatal health (CHV-NEO): study protocol for a randomized controlled trial.

BACKGROUND: Provision of essential newborn care at home, rapid identification of illness, and care-seeking by caregivers can prevent neonatal mortality. Mobile technology can connect caregivers with information and healthcare worker advice more rapidly and frequently than healthcare visits. Community health workers (CHWs) are well-suited to deliver such interventions. We developed an interactive short message service (SMS) intervention for neonatal health in Kenya, named CHV-NEO. CHV-NEO sends automated, theory-based, actionable, messages throughout the peripartum period that guide mothers to evaluate maternal and neonatal danger signs and facilitate real-time dialogue with a CHW via SMS. We integrated this intervention into Kenya's national electronic community health information system (eCHIS), which is currently used at scale to support CHW workflow. METHODS: The effect of CHV-NEO on clinical and implementation outcomes will be evaluated through a non-blinded cluster randomized controlled trial. Twenty sites across Kisumu County in Western Kenya were randomized 1:1 to provide either the national eCHIS with integrated CHV-NEO messaging (intervention) or standard of care using eCHIS without CHV-NEO (control). We will compare neonatal mortality between arms based on abstracted eCHIS data from 7200 pregnant women. Secondary outcomes include self-reported provision of essential newborn care (appropriate cord care, thermal care, and timely initiation of breastfeeding), knowledge of neonatal danger signs, and care-seeking for neonatal illness, compared between arms based on questionnaires with a subgroup of 2000 women attending study visits at enrollment in pregnancy and 6 weeks postpartum. We will also determine CHV-NEO's effect on CHW workflows and evaluate determinants of intervention acceptability, adoption, and fidelity of use through questionnaires, individual interviews, and messaging data. DISCUSSION: We hypothesize that the CHV-NEO direct-to-client communication strategy can be successfully integrated within existing CHW workflows and infrastructure, improve the provision of at-home essential newborn care, increase timely referral of neonatal illness to facilities, and reduce neonatal mortality. The intervention's integration into the national eCHIS tool will facilitate rapid scale-up if it is clinically effective and successfully implemented. TRIAL REGISTRATION: ClinicalTrials.gov, NCT05187897 . The CHV-NEO study was registered on January 12, 2022.

Task shifting and community engagement as the keys to increasing access to modern contraception: Findings from mixed methods operational research in coastal Kenya.

RATIONALE: Kilifi County, Kenya, has a modern contraceptive prevalence of 44%, compared to the national prevalence of 61%. In 2018, the Government of Kenya and Population Services Kenya implemented a pilot project in Kilifi to improve contraceptive awareness and access. AIMS AND OBJECTIVES: The Riziki Demonstration Project task shifted contraceptive injectable and implant provision to community health extension workers (CHEWs) and supported community engagement led by community health volunteers (CHVs). This mixed methods evaluation examined the effect of Riziki on contraceptive service provision and identified drivers of the intervention's success in increasing public sector contraceptive uptake. METHOD: We obtained monthly contraceptive service provision data from DHIS2 for intervention and comparison facilities from June 2018 to July 2020. Controlled interrupted time series analyses were used to assess difference-in-differences effects on the levels and trends in service provision. Qualitative data included key informant interviews with health workers, community-based focus group discussions and contraceptive journey mapping. Topics included contraception-related shifts in the health system and social environment and changes to contraceptive accessibility and demand. RESULTS: We found positive difference-in-differences effects for levels of provision for implant (adj. ß = 7.4 per month, per facility, 95% confidence interval [CI]: 2.8-12.0) and all methods combined (adj. ß = 12.2, 95% CI: 3.0-21.3). The health system provided a legal framework and management support for the pilot. Community engagement by CHVs and CHEWs, including with men and religious leaders, bolstered supportive social norms by emphasizing birth spacing and family wellbeing. Awareness and supportive social norms were strengthened among women and their husbands through counselling that leveraged CHEW and CHV community embeddedness. CONCLUSION: Findings demonstrate the potential for task sharing combined with community engagement efforts to improve contraceptive awareness, knowledge and supply. Community engagement should include men, traditional leaders and other influencers to strengthen norms that position family planning in alignment with community values.