Prevalence and Diagnosis of PCOS Using Electronic Health Records: A Scoping Review and a Database Analysis

Area of Science:

• Endocrinology
• Reproductive Health
• Health Informatics

Background:

• Polycystic ovary syndrome (PCOS) is a common endocrine disorder in reproductive-aged women.
• PCOS increases risks for type 2 diabetes, cardiovascular disease, infertility, and other serious health issues.
• Accurate identification of PCOS is vital for managing associated health risks.

Purpose of the Study:

• To evaluate the prevalence and reliability of diagnosing PCOS using electronic health records (EHRs).
• To assess the feasibility of obtaining PCOS diagnostic variables from patient databases.
• To understand the implications of EHR-based PCOS diagnosis for research.

Main Methods:

• A scoping review of literature on PCOS diagnosis using EHRs was conducted.
• Analysis of a PCOS patient database to determine the availability of diagnostic variables.
• Comparison of EHR-based diagnoses with established PCOS diagnostic criteria (Rotterdam, NIH, AEPCOS).

Main Results:

• PCOS prevalence in EHRs ranged from 0.27% to 5.8%, often lower than expected.
• Reliability of EHRs for PCOS diagnosis varied, with one study showing 50% sensitivity and 29% specificity.
• Few women in the database met full criteria for PCOS using Rotterdam (13.7%), NIH (8%), or AEPCOS (7.5%) criteria.

Conclusions:

• EHRs may lead to underestimation of PCOS prevalence.
• Many women identified via EHRs may not meet complete diagnostic criteria.
• Findings highlight limitations of EHRs for PCOS research and patient identification.