Can hospitals "game the system" by avoiding high-risk patients?

Area of Science:

• Healthcare quality improvement
• Hospital performance metrics
• Surgical outcomes research

Background:

• Quality-improvement benchmarking programs may incentivize physicians and hospitals to avoid high-risk patients to improve rankings.
• This practice could potentially lead to the denial of necessary treatments for vulnerable patient populations.

Purpose of the Study:

• To test the hypothesis that avoiding high-risk patients does not change risk-adjusted hospital rankings.
• To evaluate the impact of risk-avoidance strategies on hospital performance metrics.

Main Methods:

• A simulation analysis was conducted on 6 complex surgical procedures using the Nationwide Inpatient Sample.
• Hospitals were ranked by observed-to-expected (O/E) mortality ratios.
• Half of the hospitals simulated avoiding 25% of high-risk patients (based on Charlson, Elixhauser, age, minority, or uninsured status).

Main Results:

• Across simulations, the median change in O/E ratios was zero, with minimal changes observed in most hospitals.
• Hospital rankings remained largely unchanged, with 70.5% to 98.0% of rankings unaffected by risk avoidance.
• A small percentage of rankings improved (1.3%–13.1%) or worsened (0.7%–14.3%).

Conclusions:

• Risk-adjusted hospital rankings are unlikely to be significantly altered by selectively avoiding high-risk patients.
• In scenarios where rankings did change, improvements were as likely as worsening, indicating no consistent benefit from this strategy.