Development of a customised programme to standardise comorbidity diagnosis codes in a large-scale database

Area of Science:

• Health Informatics
• Medical Coding Systems

Background:

• The transition from the International Classification of Diseases, Ninth Revision (ICD-9) to the International Classification of Diseases, Tenth Revision (ICD-10) poses a significant data standardization challenge for large-scale longitudinal research.
• Consistent diagnostic terminology is crucial for accurate analysis and reliable research outcomes.

Purpose of the Study:

• To develop an automated program for standardizing diagnostic terminology between ICD-9 and ICD-10 coding systems.
• To address the data standardization challenges inherent in migrating between different versions of medical coding systems.

Main Methods:

• A program was developed to standardize ICD-9 and ICD-10 terminology into a unified system.
• Code optimization techniques, including hash tables and parallelization, were implemented to reduce runtime.
• Two iterations of the program were tested on a large joint ICD-9/ICD-10 database comprising 15.8 million patients.

Main Results:

• Both the original and improved programs successfully standardized diagnostic terminology within the database.
• The improved program demonstrated a substantial reduction in runtime, translating 3.1 million cells in 38 minutes compared to the original program's 100,000 cells in 12.5 hours.

Conclusions:

• The developed program effectively translates ICD-related data into a standardized format, overcoming the limitations of excessive runtimes in earlier versions.
• Code improvements using hash tables and parallelization exponentially reduced processing times, making large-scale data standardization feasible.
• Sharing the program's implementation aims to aid other researchers in standardizing their ICD-9 and ICD-10 databases for improved analytical capabilities.