Evaluating an MFI Community Health Worker Program: How microfinance group networks influence intervention outreach and impact

Background: Community Health Workers (CHWs) are considered to be a cost-effective and inclusive solution to address the persistent health workforce shortage in many low and middle-income countries. In recent years, microfinance institutions (MFIs) got increasingly engaged in providing health services delivered by CHWs. Despite their growing importance, little is known about the impacts and implementation barriers of these mostly small-scale initiatives. This paper evaluates an MFI-led CHW program in the Philippines and studies the role of microfinance group networks in influencing program outreach and impact. The intervention aims at disseminating information in poor communities, improving health monitoring through increased check-ups and raising social support. Methods: Clustered randomized controlled trial in 70 communities in the greater area of Metro Manila, the Philippines. The main data was collected in a baseline and follow-up survey and is complemented with extensive sociometric network and geographical data. The main outcome variable is a composite health index based on 10 indicators. The role of the health worker's embeddedness and connectedness in the community for program success is tested using tools of social network analysis. Results: The intervention led to a 3.8% (95% confidence interval (CI) = 1.3, 6.4) improvement in the composite health outcome. Effects across indicators are mixed and mainly driven by changes in immediate health monitoring behavior: The probability for routine examinations increased in the treatment group by 10.6% (95% CI = 3.2, 18.1), for regular blood pressure checks by 9.6% (95% CI = 3.3, 15.9), and for having access to a health care provider by 7.2% (95% CI = 0.93, 13.5). No statistical effects on general knowledge and social support are observable. Social networks are a key driver of program outreach and impact. Close friends and acquaintances of health workers used and benefited substantially more from the program than more distant ties. Conclusions: Despite the promising immediate behavioral impacts, it remains questionable to what extent such small-scale MFI initiatives can bring transformative and sustainable changes without external support. Microfinance group networks played an important role for the success of the health intervention and further research is needed to better understand how these affect the health care utilization decisions of the clients.