In the northern arid lands and other remote parts of Kenya, the Capacity Kenya project has been working with the Ministry of Health to design simple packages to attract health workers and encourage them to stay. Starting with a selected list of diverse pilot sites, the project set out to design intervention packages, implement them, and systematically evaluate their impact on health worker retention over time.

A couple of weeks ago I attended the Second Global Forum on Human Resources for Health in Bangkok, which gave me a unique opportunity to learn about the struggles, successes, and constraints of other countries. As I listened to the experiences of others and what they see as best practices, I kept thinking about what new interventions might work best in the Kenyan context. I left Bangkok with more questions than answers, and I believe the community of practice on health worker retention will continue to face challenges as we scale up beyond controlled and localized contexts.

I saw four key challenges in trying to apply the experiences of other countries to the Kenyan context:

• Most studies lacked conceptual and contextual clarity. For instance, I heard examples of retention in “rural towns,” “marginalized areas,” or rural parts of European countries, which are far better resourced than most urban centers in developing countries. In most cases, the presenters did not explain what these terms meant. Even within a country these terms have diverse interpretations and lack standardization when used in health worker retention surveys.
• Most case studies lacked insight on the challenges of mobilizing support for interventions beyond ministries of health (e.g., how ministries of education, public service, finance, as well as professional councils and other agencies could be mobilized to offer long-term support). Without sustained collaboration from other sectors, it will be difficult to sustain health workforce interventions.
• Lacking hard data, some case studies failed to show how interventions were directly attributable to health workers’ decisions to stay or leave and instead were largely descriptive.
• Sadly, a number of case studies only described plans of major intended interventions—some sounded quite ambitious and inspiring but were not at the stage of offering results or lessons that can be replicated in similar contexts. Even then, few of these intervention case studies went beyond merely describing what works to the more elaborate exploration of what makes it work.

I think we can do better and do more. Sifting through these case studies, some clear directions came to mind about what we need to work on and what kind of research is needed in health worker retention.

• We need more rigor in designing, documenting, and evaluating retention interventions.
• We need to define a common standard for a “desirable period” of retention. The opportunity costs of prolonged high-vacancy rates are not consistent across countries and contexts.
• When studies rely heavily on external resources, as is often the case, strategic and deliberate planning is needed on how to scale up and replicate positive lessons from a proliferation of pilot initiatives around retention. Sadly, few examples exist on how to mobilize broad-based funding using sector-wide approaches.
• We need studies that analyze the relative costs compared to the benefits of retention interventions in a way that can be presented to policy-makers and encourage the scale-up of successful projects in diverse contexts.
• We need studies that apply the World Health Organization’s recommendations in ways that account for local “context” and are clear in the methodology and logic behind the “bundles of interventions selected.”
• Knowing that most desirable interventions cost much more than what most crisis countries can afford, we need studies that provide more in-depth explorations of what makes successful interventions work so countries can adapt this to local contexts and local budgets.