THE STUDY

The CHV-NEO study is a non-blinded cluster randomized controlled trial. Our objective is to integrate the NEO interactive SMS intervention into routine CHV workflow through the digital Community Health Toolkit and pragmatically evaluate its clinical, service delivery, and implementation outcomes.

AIM 1: Design and develop CHV-NEO, employing a human-centered design approach

Based on iterative input from CHVs and other stakeholders, NEO functionality will be adapted for CHV use and incorporated as a module of the dCHT, named CHV-NEO. Anticipated adaptations include standardized interactive SMS response guidance and integration of interactive NEO SMS into dCHT task management.

AIM 2: Determine the effect of CHV-NEO on neonatal mortality, essential newborn care (ENC) provision, and care-seeking

We will conduct a 1:1 cluster-randomized trial in 20 facilities in counties in Western Kenya where neonatal mortality is among the highest in Kenya and CHVs currently use the dCHT. Facilities will be randomized to use standard of care (no CHV-client SMS) or the Community Health Toolkit with CHV-NEO, from late pregnancy to 6 weeks postpartum.

  • a) Neonatal mortality will be compared between arms using abstracted routine data from 7100 mothers.
  • b) Clinic visit attendance and provision of ENC (cord care, thermal care and initiation of breastfeeding) will be compared between arms using self-report data in a sub-cohort of 1900 mothers enrolled as study participants. Maternal knowledge of neonatal danger signs and satisfaction with care will also be compared.

AIM 3: Determine levels and determinants of CHV-NEO service delivery and implementation outcomes

The 1:1 cluster-randomized trial will also assess the following:

  • a) CHV workflow indicators (number of home visits performed, timeliness of task completion, and number of clinic referrals made) will be compared between arms using dCHT metadata and self-report questionnaires from 600 CHVs. Time-and-motion studies will also be conducted to validate self-reported estimates among 200 CHVs.
  • b) CHV-NEO acceptability, adoption and fidelity of use by CHVs and their supervisors will be determined using questionnaires and dCHT metadata. Qualitative and quantitative data from CHVs and supervisors will be used to define CHV, supervisor and facility drivers of acceptability, adoption and fidelity.