A novel approach to attribute responsible physicians using inpatient claims

Area of Science:

• Health Services Research
• Medical Informatics
• Health Policy

Background:

• Physician attribution for inpatient care is crucial for quality assessment.
• Existing rule-based methods using administrative claims data have limitations in accurately attributing physicians.
• There is a need for more robust methods to understand physician-level variation in care quality.

Purpose of the Study:

• To develop and validate a novel physician attribution method for inpatient stays using administrative claims data.
• To compare the proposed attribution method against traditional rule-based benchmarks.
• To assess the impact of the new method on physician attribution across various clinical cohorts.

Main Methods:

• A novel approach was developed to calculate physician-specific attribution ratios based on patient characteristics and comorbidities.
• The proposed method was compared to a rule-based benchmark.
• Attribution results were analyzed across seven Centers for Medicare & Medicaid Services (CMS)-defined clinical cohorts, including COVID-19.

Main Results:

• The novel approach demonstrated a more accurate attribution of physicians to primary clinical conditions compared to the benchmark.
• Significant shifts in attribution were observed across cohorts: cardiovascular diseases (+17.0% AMI, +20.2% HF, +4.0% CABG), pulmonary diseases (+24.0% COPD, +16.2% pneumonia), infectious diseases (+17.4% COVID-19), and neurology (+42.2% stroke).
• The method was validated on 6,835,460 unique patient encounters.

Conclusions:

• A robust and consistent method for physician attribution in inpatient settings was developed.
• This method is essential for understanding physician-level quality of care variations.
• The approach eliminates unattributed patients and provides consistency across healthcare facilities.