Tailoring "best-of-breed" safety classification for patient fall voluntary reporting.

Voluntary safety event reporting often produces poorly defined data points, which complicate data analyses across health care settings. Such data should be restructured into a standard patient safety language translatable within and outside health care organizations. We designed and implemented a "best-of-breed" patient safety classification for data created by the Duke University Health System Safety Reporting System.We report our approach for patient fall classification. Our strategy was to deploy the International Classification for Patient Safety Framework of the World Health Organization augmented with additional data points of interest, thereby allowing for data translatability while maintaining local practices. System interface redesign using the "best-of-breed" fall classification was mindful of workflows and known reporting barriers. Custom aggregate reports were also developed.We estimated the impact of the redesigned portal on Safety Reporting System usage before and after classification through comparisons of fall report volume and report completion time. When normalized as falls per day, the rate of falls only changed slightly, indicating that the enhancement had little effect on reporting desire. Report completion time increased modestly but not significantly from a practical standpoint. The presence of structured data eliminated substantial hours dedicated to manual data management and enabled evaluation of quality improvement interventions within and outside our organization.Creation and implementation of a "best-of-breed" patient safety classification for voluntary reporting requires multidisciplinary collaboration between clinical experts, frontline clinicians, and functional and technical analysts. Formal usability evaluations of reporting systems are needed to ensure design facilitates effective data collection.

Duke Scholars

Author Jeffrey Michael Ferranti Pediatrics, Neonatology