Comparing methods for identifying future high-cost mental health cases in Medicaid

Area of Science:

• Health Services Research
• Mental Health Economics
• Healthcare Policy

Background:

• Accurate identification of high-cost individuals is essential for effective resource allocation in healthcare.
• Predicting future high healthcare expenditures allows for proactive interventions and cost containment strategies.
• Existing methods for identifying high-cost cases may not be equally effective across different healthcare domains, such as mental health services.

Purpose of the Study:

• To evaluate and compare different methodologies for identifying individuals likely to incur high costs for mental health services covered by Medicaid.
• To determine the most effective predictive model for future high-cost cases within the mental health service landscape.

Main Methods:

• Analysis of Florida Medicaid claims data.
• Comparison of three predictive models: prior year cost, concurrent diagnosis-based, and prospective diagnosis-based.
• Defining high-cost individuals based on expenditure deciles.

Main Results:

• Individuals in the top decile of prior year costs averaged $13,684 in the following year, with 50% remaining in the top spending decile.
• Diagnosis-based models identified high-cost individuals averaging $10,935-$10,974, with 34% meeting the high-cost criteria in the subsequent year.

Conclusions:

• Prior year healthcare cost is a superior predictor of future high-cost cases for mental health services compared to diagnosis-based models.
• Findings contrast with previous research on physical healthcare, highlighting the need for specialized approaches in mental health cost prediction.
• The study underscores the importance of utilizing historical cost data for targeted interventions in Medicaid mental health care.