Modeling Clinician Burnout and Attrition to Optimize Staffing Ratios and Improve Patient Safety Outcomes in Underfunded Safety-Net Hospitals

Abstract

Background: Underfunded safety-net hospitals face chronic clinician shortages, elevated burnout rates, and higher

patient safety incidents. Problem statement: Existing staffing models ignore the dynamic feedback between clinician

psychological distress, attrition, and patient outcomes, perpetuating unsafe ratios. Purpose: This study develops a

predictive model integrating burnout and attrition as core variables to optimize staffing ratios and improve patient

safety. Methodology: A mixed-methods sequential explanatory design was used. Phase 1: Quantitative longitudinal

data (24 months) from three safety-net hospitals in the Midwestern US, including clinician surveys (Maslach Burnout

Inventory), electronic health record-derived patient safety indicators (PSIs), and daily staffing logs. Phase 2: Semi-

structured interviews with 25 nurse managers and hospital administrators. Data were analyzed using multilevel

mixed-effects regression and agent-based modeling. Key findings: Burnout mediates 68% of the effect of

understaffing on attrition (β = 0.72, p < .001). A 0.5 increase in the patient-to-clinician ratio increases PSIs by 34%,

with full mediation by emotional exhaustion. Conclusion/Implications: Staffing ratios alone are insufficient; dynamic

models that predict burnout-driven attrition reduce adverse events by 22% in simulations. Policy implications include

mandatory burnout surveillance and ratio adjustments for safety-net funding.