Defining Major Depressive Disorder Cohorts Using the EHR: Multiple Phenotypes Based on ICD-9 Codes and Medication...

Area of Science:

• Computational psychiatry
• Health informatics
• Observational studies

Background:

• Major Depressive Disorder (MDD) is a leading cause of global disability.
• Electronic Health Records (EHR) offer potential for large-scale MDD research.
• Current methods lack generalizable approaches for classifying depression severity in EHR data.

Purpose of the Study:

• To develop and validate a pragmatic algorithm for classifying depression phenotypes in EHR.
• To stratify an entire patient population based on depression likelihood and severity.

Main Methods:

• A five-group ordinal electronic phenotype algorithm was tested.
• Data from 278,026 patients over 10 years from an integrated health system were used.
• Convergent validity was assessed using external measures like prescriptions, suicidality, comorbidity, mortality, healthcare utilization, and polygenic risk scores.

Main Results:

• Consistent patterns of increasing morbidity and adverse outcomes were observed across the five depression phenotype groups.
• These findings provide evidence for the algorithm's convergent validity.

Conclusions:

• The developed algorithm offers meaningful face and convergent validity for stratifying patient populations by depression severity.
• This simple, generalizable algorithm can be applied to most EHR datasets for depression research.