Clarifying radiology's role in safety events: a 5-year retrospective common cause analysis of safety events at a pediatric hospital

Background Common cause analysis of hospital safety events that involve radiology can identify opportunities to improve quality of care and patient safety. Objective To study the most frequent system failures as well as key activities and processes identified in safety events in an academic children's hospital that underwent root cause analysis and in which radiology was determined to play a contributing role. Materials and methods All safety events involving diagnostic or interventional radiology from April 2013 to November 2018, for which the hospital patient safety department conducted root cause analysis, were retrospectively analyzed. Pareto charts were constructed to identify the most frequent modalities, system failure modes, key processes and key activities. Results In 19 safety events, 64 sequential interactions were attributed to the radiology department by the patient safety department. Five of these safety events were secondary to diagnostic errors. Interventional radiology, radiography and diagnostic fluoroscopy accounted for 89.5% of the modalities in these safety events. Culture and process accounted for 55% of the system failure modes. The three most common key processes involved in these sequential interactions were diagnostic (39.1%) and procedural services (25%), followed by coordinating care and services (18.8%). The two most common key activities were interpreting/analyzing (21.9%) and coordinating activities (15.6%). Conclusion Proposing and implementing solutions based on the analysis of a single safety event may not be a robust strategy for process improvement. Common cause analyses of safety events allow for a more robust understanding of system failures and have the potential to generate more specific process improvement strategies to prevent the reoccurrence of similar errors. Our analysis demonstrated that the most common system failure modes in safety events attributed to radiology were culture and process. However, the generalizability of these findings is limited given our small sample size. Aligning with other children's hospitals to use standard safety event terminology and shared databases will likely lead to greater clarity on radiology's direct and indirect contributions to patient harm.