Optimizing multi-component implementation packages

With HSPH Epidemiology Sc.D. student Dale Barnhart, for whom I serve as primary mentor, we are learning how to improve the design and data collection processes for multi-phase learn-as-you-go designs in order to design an optimal intervention. We have partnered with Ariadne Lab’s BetterBirth program, which used a ‘learn-as-you-go’ design consisting of several pilot studies to develop a multi-component implementation package designed to promote the use of the World Health Organization’s Safe Childbirth Checklist in Uttar Pradesh, India and ultimately reduce child morbidity and mortality. Through a review of published manuscripts, internal documentation, and data collected across development phases we have identified lessons learned that can be employed by future researchers seeking to develop a multi-component implementation package. We have used data from the BetterBirth study to produce illustrative examples showing how our suggestions could be used to improve the design of future learn-as-you-go studies. As study designs featuring multiple rounds of pilot studies are becoming increasingly common in implementation science, we believe the lessons learned from this research will have the potential to inform the design and analysis of many future studies. A first draft manuscript documenting these lessons learned has been completed, and we expect to submit the paper for publication by October 2018.