A conceptual learning analysis of paired after action and intra action reviews for health emergencies.

Background: Processes of self‐reflection and the learning they allow are crucial before, during, and after acute emergencies, including infectious disease outbreaks. Tools—such as Action Reviews—offer World Health Organization (WHO) member states a platform to enhance learning. We sought to better understand the value of these tools and how they may be further refined and better used. Methods: We searched the publicly available WHO Strategic Partnership for Health Security website for paired reports of Action Reviews, that is, reports with a comparable follow‐up report. We complemented the paired action reviews, with a literature search, including the gray literature. The paired action reviews were analyzed using the "Learning Health Systems" framework. Results: We identified three paired action reviews: Lassa Fever After Action Reviews (AARs) in Nigeria (2017 and 2018), COVID‐19 Intra‐Action Reviews (IARs) in Botswana (2020 and 2021), and COVID‐19 IARs in South Sudan (2020 and 2021). Action Reviews allowed for surfacing relevant knowledge and, by engaging the right (in different contexts) actors, asking "are we doing things right?" (single loop learning) was evident in all the reports. Single loop learning is often embedded within examples of double loop learning ("are we doing the right things?"), providing a more transformative basis for policy change. Triple loop learning ("are we learning right"?) was evident in AARs, and less in IARs. The range of participants involved, the level of concentrated focus on specific issues, the duration available for follow through, and the pressures on the health system to respond influenced the type (i.e., loop) and the effectiveness of learning. Conclusion: Action Reviews, by design, surface knowledge. With favorable contextual conditions, this knowledge can then be applied and lead to corrective and innovative actions to improve health system performance, and in exceptional cases, continuous learning. [ABSTRACT FROM AUTHOR]