Enhancing Depression Identification and Stratification with a Claims-Based Analytical Framework.

Area of Science:

• Health Informatics
• Mental Health Research
• Data Analytics in Healthcare

Background:

• Depression significantly impacts socioeconomic status and well-being.
• Current depression severity assessments often use single indicators, limiting accuracy.
• Integrating diverse data sources is crucial for comprehensive health assessments.

Purpose of the Study:

• To evaluate a multi-parameter analytics-enabled Identification and Stratification (IDS) framework for depression.
• To improve the accuracy of depression identification and severity stratification.
• To leverage health insurance claims and electronic health record data for a more robust analysis.

Main Methods:

• Retrospective study using the Highmark Health dataset (members aged 18+).
• Evaluation of an IDS framework integrating claims and electronic health records.
• Comparison of IDS framework results against single-indicator diagnostic approaches.

Main Results:

• The IDS framework identified 720,882 members with depression (16.6% of the population).
• An additional 258,206 members (5.9%) were identified compared to diagnosis codes alone.
• Severity stratification revealed 5.0% mild, 8.5% moderate, and 2.2% severe depression cases.
• The IDS framework adjusted severity for 46% of mild and 19% of moderate cases.

Conclusions:

• The IDS framework effectively identifies and stratifies depression severity by integrating fragmented data.
• This comprehensive approach allows for targeted interventions, potentially reducing costs and improving outcomes.
• The study highlights the utility of integrative platforms for managing depression and addressing care gaps.