We published an academic paper tackling challenges associated with designing health intervention pilots. In our implementation work we often hear significant frustration with how successful pilots are not always taken to scale. However, we also often find that pilots are not designed nor executed with scale as the goal.

In this paper we propose four guidelines to guide decisions around when and how to pilot:

1. Only pilot what you can scale: design pilots with implementation at scale as the end point, considering common barriers to implementation such as financial resource limitations, human resource shortages and regulatory restrictions.

2. Ensure that the pilot addresses the fundamental challenges of scale: interventions at scale may face a unique set of challenges that are not evident in smaller scale models. Consider the challenges that an intervention may face at scale and whether your pilot is helping you understand and manage these.

3. Carefully consider the true cost of the pilot: direct costs of an intervention at scale may make a program unviable, including factors such as additional monitoring and management costs. In addition, it is worth considering the opportunity costs of piloting – if the impact is high and the risk is low, then delaying implementation by running a pilot may have significant costs associated with it.

4. Consider alternative approaches to mitigate uncertainty: you may find that alternative approaches beyond piloting may be more appropriate for understanding uncertainty associated with a proposed intervention. These include a phased scale-up or implementing multiple strategies in parallel.

Pilots can be a powerful tool in health systems reform. They can help increase innovation, manage risk, and support rapid scale. However they can also be misused risking frontline fatigue, wasted resources and redirected political attention.

To read the full article:

E.J. Hannay, F. Whelan, Health intervention pilots: thinking big before thinking small, Healthcare (2015)
http://www.sciencedirect.com/science/article/pii/S2213076415000573