Using information on clinical conditions to predict high-cost patients

Area of Science:

• Health Services Research
• Medical Informatics
• Epidemiology

Background:

• Accurate prediction of high medical expenditures is crucial for healthcare resource allocation and financial planning.
• Existing models often rely on limited clinical data, potentially underestimating future costs.

Purpose of the Study:

• To compare the predictive performance of various models for identifying individuals likely to incur high medical expenditures prospectively.
• To evaluate the incremental value of clinical data, chronic condition counts, and self-reported health status in expenditure prediction.

Main Methods:

• Utilized nationally representative Medical Expenditure Panel Survey (MEPS) data.
• Developed and validated logistic regression models to predict high expenditure deciles using baseline data.
• Compared Diagnostic Cost Group (DCG) risk scores against chronic condition counts and self-rated health indicators.

Main Results:

• Medical condition information, particularly DCG risk scores, significantly enhanced expenditure prediction beyond demographics.
• The number of chronic conditions, self-rated health, and functional limitations were independently associated with future high costs.
• A comprehensive model incorporating these factors demonstrated good predictive discrimination (c-statistic = 0.836).

Conclusions:

• The number of chronic conditions is a valuable predictor for future high medical expenditures.
• While self-rated health and functional limitations showed associations, their incremental predictive value was modest when controlling for clinical factors.