Hospital-specific Template Matching for Benchmarking Performance in a Diverse Multihospital System

Area of Science:

• Health Services Research
• Quality Improvement
• Comparative Effectiveness

Background:

• Hospital performance measurement is evolving, with newer methods like hospital-specific template matching proposed as fairer alternatives to regression-based benchmarking.
• Existing research on hospital-specific template matching is limited, necessitating further investigation into its applicability and effectiveness.

Purpose of the Study:

• To evaluate the feasibility of implementing hospital-specific template matching for performance assessments within the Veterans Affairs (VA) health care system.
• To determine the statistical power of this method to detect significant variations in 30-day mortality rates.

Main Methods:

• An observational cohort study utilizing hospital-specific template matching was conducted.
• 30-day mortality for hospitalizations in each VA hospital was compared against pooled mortality from matched hospitalizations at similar VA hospitals.
• Simulations were employed to assess the power to detect greater-than-expected mortality.

Main Results:

• The study included 556,266 hospitalizations across 122 VA hospitals in 2017.
• Most hospitals (89.3%) showed no significant difference in 30-day mortality compared to their matched benchmarks.
• A small proportion of hospitals had significantly lower (6.6%) or higher (4.1%) mortality rates.

Conclusions:

• Hospital-specific template matching appears feasible for performance assessment in the diverse VA system, but requires further refinement for operational deployment.
• The findings suggest this methodology could be applicable to other large, diverse multihospital systems.