When Communities Participate in Primary Health Care: A Randomized Controlled Trial of an NGO Led Community Health Worker Program in the Philippines

In many developing countries public health systems are confronted with a significant shortage in professional health workers. In 2014, more than 4 million health workers were missing worldwide, based on WHO estimations. The shortage is recognized as one of the most important constraints in achieving progress in reaching long-term global health and development goals. Community Health Workers (CHWs) have been proposed as a cost-effective solution to address the enduring shortage. CHWs are lay health workers who, after receiving a short health training, work as support personnel in their communities. With their close relationship to the communities they serve as a bridge between their peers and the public health sector. Among other functions they promote good health behaviors, monitor treatment compliance, and perform small treatments. Besides national programs, NGOs have increasingly gotten engaged in developing and implementing own small-scale CHW programs. So far, little is known about the benefits and implementation barriers of such NGO led initiatives. In our study we evaluate a CHW program in the Philippines using a cluster randomized controlled trial (RCT) design. The program was initiated by a non-profit social development NGO operating in Metro Manila. Extensive data was collected in interviews among 1064 community members from 70 neighborhoods. In contrast to previous studies, we also collected information about the CHWs, e.g. about their competencies, motivation, and relationship to the community. We find that the CHW program has a positive direct effect on the health behavior and knowledge of the CHWs and their families. However, we do not find evidence for positive spillover effects to the communities. Our results suggest that the lack of spillovers is mostly due to missing information and mistrust towards the CHW’s abilities. We discuss potential improvements of the program and derive policy recommendations that can be extended to programs in other contexts.