Identifying Patterns of Depression Comorbidities Using Association Rule Learning: Insights from Maryland Medicaid...

Area of Science:

• Health Services Research
• Computational Health
• Psychiatric Epidemiology

Background:

• Depression is a prevalent psychiatric disorder and a significant risk factor for suicide.
• Understanding depression comorbidities is crucial for effective chronic disease management and early diagnosis.

Purpose of the Study:

• To identify association rules in patients with multiple chronic conditions, focusing on depression comorbidities.
• To explore how age and sex influence these comorbidity patterns.

Main Methods:

• Analysis of Maryland Medicaid claims data (2021-2022).
• Application of association rule learning to examine co-occurrence of depression with 62 other chronic conditions.
• Stratification of analyses by sex and age group.

Main Results:

• A total of 582 association rules were identified.
• The number of association rules increased with advancing age, particularly in women.
• Significant variations in comorbidity patterns were observed across demographic subgroups.

Conclusions:

• Association rule learning effectively detects clinically relevant depression comorbidity patterns.
• Identified patterns vary by age and sex, offering insights for clinical practice.
• Findings can improve targeted screening, early diagnosis, and management of patients with multiple chronic conditions.