Measuring physician adherence with gout quality indicators: a role for natural language processing

Area of Science:

• Rheumatology
• Health Informatics
• Clinical Quality Improvement

Background:

• Gout management requires adherence to quality indicators (QIs) for medication use, monitoring, and behavioral modification (BM).
• Electronic medical records (EMRs) contain valuable clinical narrative data that can be analyzed for quality assessment.

Purpose of the Study:

• To evaluate physician adherence to gout-specific quality indicators (QIs) related to medication management and monitoring.
• To assess the documentation of behavioral modification (BM) counseling for gout patients using natural language processing (NLP).

Main Methods:

• Quality indicators (QIs) assessed included allopurinol dosing in chronic kidney disease (CKD), uric acid monitoring, and monitoring for colchicine initiation.
• Natural language processing (NLP) analyzed EMRs for behavioral modification (BM) counseling on diet, weight loss, and alcohol consumption in gout patients.
• Statistical analyses compared QI compliance and correlated rheumatology visits with QI adherence and BM counseling.

Main Results:

• QI compliance varied widely: 92.1% for initial allopurinol dosage, 44.8% for uric acid monitoring, and 7.7% for colchicine monitoring.
• Behavioral modification (BM) counseling was documented for weight loss (64.0%), diet (24.8%), and alcohol (8.7%).
• Higher rheumatology clinic attendance correlated with improved QI compliance and more frequent BM counseling for weight loss and diet.

Conclusions:

• Rheumatology clinic attendance is associated with better adherence to gout quality indicators.
• Natural language processing (NLP) is an effective tool for extracting and measuring behavioral modification (BM) documentation from clinical narratives in EMRs.