Creating a Statistical Analysis Plan to Continually Evaluate Intervention Adaptations that Arise in Real-World Implementation.

Evidence-based health interventions are frequently translated into real-world settings where practical needs drive changes to intervention protocols. Due to logistical and resource constraints, these naturally arising adaptations are rarely assessed for comparative effectiveness using a randomized trial. Nevertheless, when observational data are available, it is still possible to identify beneficial adaptations using statistical methods that adjust for differences among intervention groups.