Coding Free-Text Chief Complaints from a Health Information Exchange: A Preliminary Study

Area of Science:

• Medical Informatics
• Computational Linguistics
• Public Health

Background:

• Chief complaints are valuable textual data within electronic health records (EHRs).
• Extracting structured information from chief complaints can enrich diagnostic and symptom data.
• Existing natural language processing (NLP) techniques may be enhanced for better utilization of this data.

Purpose of the Study:

• To present and evaluate a method for preprocessing chief complaints.
• To assign corresponding ICD-10-CM codes using NLP and medical terminologies.
• To assess the potential of this method for improving EHR data utilization.

Main Methods:

• Utilized the MetaMap NLP system and UMLS Metathesaurus for concept extraction.
• Conducted exploratory analysis on 7,942 unique chief complaints from Rhode Island's health information exchange.
• Evaluated the method on 123,086 chief complaints with associated ICD-10-CM encounter diagnoses.

Main Results:

• The proposed method achieved an 87.82% accuracy in assigning MetaMap-extracted concepts.
• Preliminary findings indicate successful preprocessing and coding of chief complaints.
• The system demonstrated effectiveness in linking chief complaints to standardized diagnostic codes.

Conclusions:

• The explored method shows potential for improving NLP techniques in healthcare.
• This approach can enhance the use of chief complaint data for clinical care, research, and public health surveillance.
• Accurate ICD-10-CM code assignment from chief complaints is feasible and beneficial.