Systematic review: development of a consensus code set to identify cirrhosis in electronic health records

Area of Science:

• Medical Informatics
• Hepatology
• Clinical Research

Background:

• Electronic health records (EHRs) offer valuable longitudinal data for cirrhosis research.
• A lack of standardized coding for cirrhosis in EHRs hinders large-scale studies.
• Existing coding practices for cirrhosis lack consensus and consistent validation.

Purpose of the Study:

• To systematically review and evaluate the validity of diagnostic coding for cirrhosis in EHRs.
• To synthesize a comprehensive set of International Classification of Diseases, 10th Revision (ICD-10) codes for cirrhosis research.
• To develop and validate a consensus code set for improved cirrhosis identification in EHRs.

Main Methods:

• Systematic review of MEDLINE and EMBASE databases for studies using EHRs to identify cirrhosis.
• Extraction and summarization of validated code sets and their performance characteristics.
• Citation analysis to inform the development of a consensus code set, followed by validation in diverse patient cohorts.

Main Results:

• Eighteen studies were identified, with positive predictive value (PPV) reported in 17/18.
• Citation analysis revealed ongoing variation in coding practices.
• A consensus code set of nine frequently used codes was developed and validated, showing high PPV (83%-89%) and improved sensitivity compared to existing methods.

Conclusions:

• Significant variation exists in current EHR coding practices for cirrhosis.
• A validated consensus code set for cirrhosis has been developed, demonstrating superior performance.
• The proposed consensus code set aims to standardize EHR study designs in cirrhosis, fostering international research collaboration.