A hospital-specific template for benchmarking its cost and quality

Area of Science:

• Health Services Research
• Hospital Quality Improvement
• Healthcare Economics

Background:

• Traditional hospital audits often lack specificity to individual patient populations.
• Standardization methods may not fully capture institutional nuances in care delivery and outcomes.

Purpose of the Study:

• To develop and validate an improved hospital auditing methodology.
• To create a patient-specific template for direct standardization of hospital performance.

Main Methods:

• Utilized Medicare claims data (2004-2006) for general, gynecologic/urologic surgery, and orthopedic procedures.
• Constructed a 300-patient template from an index hospital and matched 300 patients from 43 other hospitals using multivariate matching.
• Ensured matched hospitals had at least 500 patients over the 3-year study period.

Main Results:

• Achieved close matches on patient characteristics and procedures, exceeding expectations for randomized trials.
• Identified significant differences in mortality, failure-to-rescue, and cost between the index hospital and the 43 matched hospitals.
• Demonstrated minimal differences in patient characteristics despite significant outcome variations.

Conclusions:

• Patient-specific template matching offers a fair and directly standardized approach to hospital auditing.
• This method combines the fairness of direct standardization with the institutional relevance of indirect standardization.
• Empowers hospitals to conduct more accurate performance evaluations and identify drivers of observed results.