Using diagnoses to describe populations and predict costs.

Area of Science:

• Health economics
• Medical informatics
• Public health

Background:

• Diagnostic Cost Group Hierarchical Condition Category (DCG/HCC) payment models are crucial for summarizing population health and predicting healthcare expenditures.
• These models rely on diagnoses from patient encounters to identify medical conditions and use demographic data to forecast costs.

Purpose of the Study:

• To describe the logic, structure, coefficients, and performance of DCG/HCC models.
• To validate these models on diverse datasets representing different healthcare populations.

Main Methods:

• Development and validation of DCG/HCC models.
• Utilized large-scale datasets from privately insured individuals, Medicaid, and Medicare beneficiaries, each exceeding one million participants.

Main Results:

• Detailed description of the DCG/HCC model's components, including its underlying logic, structural elements, and coefficient values.
• Demonstrated model performance and validity across three distinct large-scale databases, confirming its applicability to varied patient populations.

Conclusions:

• The DCG/HCC models provide a robust framework for understanding population health and predicting healthcare costs.
• Validation across private, Medicaid, and Medicare populations indicates the models' generalizability and utility in healthcare payment systems.