Systematic method for classifying multiple congenital anomaly cases in electronic health records

Area of Science:

• Medical informatics
• Genetics
• Pediatrics

Background:

• Congenital anomalies (CAs) impact 3% of births, causing significant infant morbidity and mortality.
• Multiple congenital anomalies (MCA) occur in some individuals, but systematic identification in electronic health records (EHRs) is lacking.
• Understanding MCA's genetic and epidemiologic factors is crucial but hindered by identification challenges.

Purpose of the Study:

• To develop a scalable and accurate method for identifying individuals with multiple congenital anomalies (MCA) within electronic health records (EHRs).
• To enable improved research into the genetic and epidemiologic underpinnings of MCA.
• To establish a standardized approach for MCA characterization across diverse EHR databases.

Main Methods:

• Evaluation of three distinct MCA classification approaches using an anonymized EHR database.
• Implementation of a novel method that bypasses minor vs. major CA differentiation.
• Utilizing phenome-wide association studies to analyze the phenome associated with previously classified minor CAs.

Main Results:

• The developed universal MCA identification method in EHRs demonstrates high accuracy (97.1% PPV).
• Identified MCA cases show increased hospital utilization, with 41% receiving inpatient care.
• The method effectively captures detailed patterns of congenital anomalies and was validated in two additional cohorts.

Conclusions:

• A comprehensive method for identifying individuals with MCA in EHRs has been successfully developed.
• This method facilitates deeper investigation into the genetic causes of MCA.
• The approach is transferable and applicable to EHR systems utilizing billing codes.