Testing mHealth solutions at the last mile: insights from a study of technology-assisted community health referrals in rural Kenya

Abstract

Background

mHealth technologies are already disrupting conventional healthcare delivery by making innovative solutions more accessible in terms of reach and price across reach and price across the developing world. However, much less has been documented on the process of mHealth innovation introduction in the context of rural communities of Africa. Pending still is the widespread adoption of standards and the removal of barriers to introduction, testing and scale. This paper documents the innovation process of technology introduction, results and lessons learned through a case study of two mHealth initiatives: closed-loop referrals for maternal and child health; and HIV self-testing. Both initiatives were implemented and evaluated in Kisii County, Kenya by Living Goods.

Methods

Living Goods applied an innovation framework to introduce and evaluate two interventions integrated into the Living Goods Smart Health app, a smartphone-based digital health application designed to carry out household registration, assessment, and diagnosis at community level. Community health workers (CHWs) used digitally assisted, standardized Ministry of Health algorithms to assess and refer clients to the nearest health facility for diagnosis confirmation and treatment as appropriate. Routine data as well as periodic household surveys were captured to incorporate performance data and outcomes into activity management. A quasi-experimental evaluation was carried out using a Propensity Score Matching (PSM) methodology to evaluate intervention arms for each intervention.

Results

Findings suggest that the initiatives increased the frequency of visits to households with participants in the treatment groups being more likely to have been visited more than six times within the last six months. The interventions contributed in part to an increase in the frequency of CHW follow-up visits within the treatment group. Attitudes of trust and confidence in CHWs were high but limited to referral services and not to diagnostic and curative services.

Conclusions

The innovation process effectively positioned and tested at community level the two interventions to address key barriers to service delivery acceptance and uptake. Despite extensive pre-testing and field iterations to adapt the solutions to the local context, behavioral and technology barriers persisted. The study highlights important implications for both innovators and service providers: technology introduction and adaptation at community level requires multiple, rapid iteration loops to ensure product refinement and user-acceptance; behavioral assessments of acceptability require a wholistic approach to ensure effective alignment of senders, receivers and trusted intermediaries of novel services.