Evaluating markers for methamphetamine use in electronic health records

Area of Science:

• Public Health Surveillance
• Health Informatics
• Substance Use Research

Background:

• Methamphetamine use is a growing public health concern in the US.
• Electronic health records (EHRs) offer timely data for substance use surveillance.
• Current methods for identifying methamphetamine use in EHRs are limited by non-specific codes and varied clinical documentation.

Purpose of the Study:

• To identify and evaluate markers for detecting methamphetamine use within EHRs.
• To provide guidance on selecting appropriate markers based on accuracy and computational feasibility.
• To improve the accuracy of methamphetamine use data for public health initiatives.

Main Methods:

• Identified eight distinct markers for methamphetamine use in an integrated EHR system.
• Assessed markers based on their predictive value and computational lift for EHR extraction.
• Conducted chart reviews to validate the accuracy of electronic definitions.

Main Results:

• Evaluated 814 charts across eight marker groups, with positive predictive values (PPVs) ranging from 0.21 to 1.00.
• Four marker groups achieved PPVs greater than 0.70 when compared to chart review.
• Marker extraction complexity varied significantly, from simple diagnosis codes to challenging free-text clinical notes.

Conclusions:

• Multiple validated options exist for extracting methamphetamine use data from EHRs.
• Marker selection should consider project needs, technical resources, and EHR system specifics.
• Enhanced data accuracy on methamphetamine use can support targeted public health interventions.