Learning from Latent Safety Threats Identified During Simulation to Improve Patient Safety.

BACKGROUND: Latent safety threats (LSTs-characteristics of design, processes, or physical environment in health care compromising patient safety) are commonly revealed during simulation-based training. Methods of collecting, analyzing, and classifying LSTs are underdeveloped and not standardized. Building on a large simulation program in one organization, the authors aimed to collect LSTs systematically and develop a taxonomy to classify them.

METHODS: The authors modified the Press Ganey Healthcare Performance Improvement Failure Modes Taxonomy (HPI-FMT), a standardized framework for safety event classification in health care, and used three categories: System, Individual, and Medications. The subcategories were revised to reflect simulation LST content and promote consistent data entry into a spreadsheet. Data visualization software was used to analyze LST data and generate dashboards, graphs, and executive summaries to share across the system that depicted data for individual hospitals and outpatient areas and allowed grouping, comparisons, and trending.

RESULTS: Over a year, the researchers identified 1,318 LSTs in 232 simulations across the organization-a rate of 5.7 LSTs/simulation. The top three LST subcategories were Environment/supplies/equipment (System category); Process/structure (System category); and Knowledge or unformed skill/habit (Individual category). Other important LSTs were Missing/malfunctioning supplies/equipment; Unclear or ineffective process or no process; and Unfamiliarity with supplies/equipment. When a repetitive pattern of LSTs was observed (for example, improper dantrolene use during malignant hyperthermia simulations), targeted process improvement or training was implemented.

CONCLUSION: The authors developed, implemented, and refined a systematic method of collecting, analyzing, displaying LSTs, and recommending targeted process improvements or training when LST trends were noted.