The objectives of this paper are to discuss the role of risk analysis and event taxonomies in patient safety reporting systems (PSRSs) and present a conceptual model that supports the use of reporting and analysis to help guide patient safety improvement efforts.
This research involves an analysis of the methodologies being used to use medical incident reports to improve patient safety. Areas discussed are risk analysis, incident-reporting contributions to risk measures, and event taxonomies for health care procedures.
Incidents reported in PSRSs are subject to selection bias, have unknown denominators, and require standardized taxonomies for numerators. PSRSs provide a mechanism to identify and learn from mistakes. A conceptual model for using a PSRS to improve safety is proposed. This model includes 4 major elements: (1) recognition and reporting of events, (2) event analysis, (3) analysis of results produced, and (4) process changes developed and implemented. The central themes of this model are education and learning to engage staff and organizations and to affect behavioral change.
PSRS is a widely recommended as a strategy to address the important problem of patient safety. Most efforts have focused on developing reporting systems and collecting incident data. We are now faced with deciding how best to analyze and report information back to stakeholders and what process changes will best decrease harm. We outline a comprehensive conceptual model to help realize the full potential of reporting systems in patient safety improvement efforts.
From the Departments of *Anesthesiology and Critical Care Medicine and †Surgery, ‡The John Hopkins University, Baltimore, Maryland, and §Department of Health Policy & Management, Bloomberg School of Public Health, The Johns Hopkins University, Baltimore, Maryland.
Grant support provided by the Agency for Healthcare Research and Quality (grant no. 1 U18HS11902).
Correspondence: Peter J. Pronovost, MD, PhD, 1909 Thames Street, 2nd Floor, Baltimore, MD 21231 (e-mail: email@example.com).