Case-mix for assessing primary care value (CPCV).

Area of Science:

• Health Services Research
• Health Economics
• Primary Care Medicine

Background:

• Accurate primary care (PC) performance measurement and value-based payment systems are hindered by a lack of effective risk adjustment mechanisms.
• Existing risk scores for total resource needs are insufficient for assessing PC performance or setting PC payment rates, as high-cost patients may not require extensive PC services and vice versa.
• A significant challenge in developing PC-specific risk algorithms is the unobservable nature of key components of PC providers' workload.

Purpose of the Study:

• To develop a novel risk adjustment algorithm specifically for primary care.
• To identify reliable proxies for unobservable primary care workload within a large patient population.
• To enable more equitable performance comparisons and payment structures in primary care.

Main Methods:

• Analysis of 5,172,773 patients within the U.S. Veterans Affairs (VA) healthcare system.
• Development of a primary care workload proxy by combining the number of PC visits and prescription drug classes.
• Creation of a case-mix algorithm using the developed workload proxy and a publicly available patient classification system for comorbidities.

Main Results:

• A novel case-mix algorithm for primary care was developed, demonstrating high explanatory power (R² = 0.702).
• The algorithm effectively incorporates proxies for unobservable primary care workload.
• The model utilizes a publicly available patient classification system to account for comorbidities.

Conclusions:

• The developed primary care risk adjustment algorithm offers a reliable method for assessing PC performance, workload, and staffing.
• This algorithm can facilitate the design of more equitable payment systems that reward value over volume in primary care.
• Its basis on a publicly available classification system allows for broader adoption and comparison across different health systems.