Exploring the “Black Box” of Recommendation Generation in Local Health Care Incident Investigations: A Scoping Review

Abstract

Background Incident investigation remains a cornerstone of patient safety management and improvement, with recommendations meant to drive action and improvement. There is little empirical evidence about how—in real-world hospital settings—recommendations are generated or judged for effectiveness. Objectives Our research questions, concerning internal hospital investigations, were as follows: (1) What approaches to incident investigation are used before the generation of recommendations? (2) What are the processes for generating recommendations after a patient safety incident investigation? (3) What are the number and types of recommendations proposed? (4) What criteria are used, by hospitals or study authors, to assess the quality or strength of recommendations made? Methods Following PRISMA-ScR guidelines, we conducted a scoping review. Studies were included if they reported data from investigations undertaken and recommendations generated within hospitals. Review questions were answered with content analysis, and extracted recommendations were categorized and counted. Results Eleven studies met the inclusion criteria. Root cause analysis was the dominant investigation approach, but methods for recommendation generation were unclear. A total of 4579 recommendations were extracted, largely focusing on individuals’ behavior rather than addressing deficiencies in systems (<7% classified as strong). Included studies reported recommendation effectiveness as judged against predefined “action” hierarchies or by incident recurrence, which was not comprehensively reported. Conclusions Despite the ubiquity of incident investigation, there is a surprising lack of evidence concerning how recommendation generation is or should be undertaken. Little evidence is presented to show that investigations or recommendations result in improved care quality or safety. We contend that, although incident investigations remain foundational to patient safety, more enquiry is needed about how this important work is actually achieved and whether it can contribute to improving quality of care.