The Recovery Deficit: A Simulation Model of Physician Performance Under Sleep Deprivation.

BACKGROUND: Medical education policy has focused on duty hours as the primary lever for safety and training quality. This work-centric framing treats sleep as residual time rather than a prerequisite for learning and safe practice. As a result, trainees often operate in chronic sleep debt, accumulating a performance and safety deficit that duty-hour counts alone cannot resolve. OBJECTIVE: This study uses Monte Carlo simulation to quantify how sleep deprivation affects clinician performance, safety, and well-being and establishes evidence-based sleep visualizations to promote recovery-centered residency redesign. DESIGN, SETTING, AND PARTICIPANTS: Computer-based Monte Carlo simulation (Stata 18.0), parameterized from experimental and cohort literature on sleep restriction and clinical performance. Each of 500 iterations generated 11,000 synthetic clinician-days uniformly across medical students, residents, and attending physicians. Nightly sleep was the primary exposure; outcomes included reaction time, clinical and diagnostic accuracy, attentional lapses, medical errors, mood, and burnout. Outputs were baselined to the ≥8 h sleep category to visualize the deficit relative to a rested state. RESULTS: Performance declined steeply with shorter sleep. Composite score fell from 83.3 (≥8 h) to 56.8 (4 h), 38.9 (2 h), and 9.0 (0 h). Reaction time slowed from 280 ms (≥8 h) to 404 ms (4 h) and 602 ms (0 h). Error risk rose markedly: <4 h vs ≥4 h, 16.2% vs 6.8% (OR: 2.62; risk difference +9.3 pp); <6 h vs ≥6 h, 9.7% vs 6.0% (RR: 1.62; OR 1.68; risk difference +3.7 pp). Burnout prevalence (modeled probability) increased from 12.8% (≥8 h) to 51.3% (0 h). Uncertainty was summarized using simulation percentiles across iterations. CONCLUSIONS AND RELEVANCE: Duty-hour counts alone obscure a daily performance and safety deficit driven by inadequate recovery sleep. Quantifying and visualizing this deficit shows severe, dose-dependent impairments and elevated error risk with short sleep, supporting a recovery-first training model that designs duty hours around protected seven to nine hours of sleep to enhance patient safety, learning, and clinician well-being.