Episode-based payment: evaluating the impact on chronic conditions.

Area of Science:

• Health Economics
• Health Services Research
• Chronic Disease Management

Background:

• Policymakers aim to improve healthcare efficiency by shifting from fee-for-service to episode-based bundled payments.
• Current bundled payment initiatives primarily target surgical procedures, with limited understanding of chronic condition applications.
• Developing effective episode-based payment models for chronic conditions like diabetes and coronary artery disease (CAD) is crucial.

Purpose of the Study:

• To compare the performance of two distinct episode-creation algorithms.
• To analyze differences in patient identification and payment estimation for diabetes and CAD using these algorithms.

Main Methods:

• A retrospective analysis of enrollees with continuous coverage from 2003-2006 was conducted.
• Two algorithms, Episode Treatment Groups (ETGs) and the Prometheus model, were used to assess annual episode-based payments.
• HEDIS criteria were used to identify patients with diabetes or CAD.

Main Results:

• For diabetes, ETGs identified 92% of patients with a mean payment of $12,731, while Prometheus identified 96% (1,195 patients) with a mean payment of $23,250.
• For CAD, ETGs identified 60% of patients with a mean payment of $24,362, and Prometheus identified 69% (948 patients) with a mean payment of $26,536.
• Both algorithms identified different patient cohorts and showed significant payment variations for diabetes, but similar estimates for CAD.

Conclusions:

• The ETG and Prometheus models yield different patient populations and payment estimates for chronic conditions.
• Significant discrepancies in episode-based payment estimates exist for diabetes, highlighting implementation challenges.
• Careful consideration and discussion are necessary to establish appropriate episode-based payment structures for chronic diseases.