Using ICD-10-CM codes to detect illicit substance use: A comparison with retrospective self-report

Area of Science:

• Medical Informatics
• Public Health Surveillance
• Clinical Research

Background:

• Accurate detection of substance use is crucial for public health surveillance and research.
• International Classification of Disease, 10th Revision, Clinical Modification (ICD-10-CM) codes are widely used in healthcare data but their utility for substance use detection is not fully understood.

Purpose of the Study:

• To evaluate the accuracy of ICD-10-CM codes in identifying illicit substance use among patients with chronic pain.
• To determine the sensitivity and specificity of ICD-10-CM codes compared to self-reported substance use.

Main Methods:

• Retrospective cohort study using data from 602 safety-net patients prescribed opioids for chronic non-cancer pain (2015-2018).
• Calculated sensitivity and specificity of ICD-10-CM codes for detecting methamphetamine, cocaine, and opioid use.
• Compared code accuracy based on self-reported substance type, frequency, and healthcare encounter type (outpatient vs. emergency department).

Main Results:

• ICD-10-CM codes demonstrated high specificity (85.0-96.4%) but low sensitivity (36.3-49.5%) for detecting self-reported substance use.
• Sensitivity significantly increased for frequent users, particularly for methamphetamine (≥weekly use: 67.2%) and opioids (≥weekly use: 52.6%).
• Code sensitivity was higher for outpatient visits compared to emergency department visits across all substance types.

Conclusions:

• ICD-10-CM codes have high specificity but low sensitivity for detecting general substance use, though they are more effective for identifying frequent use.
• Caution is advised when using ICD-10-CM codes for substance use surveillance, especially from emergency department visits.
• These codes may serve as a valuable lower-bound measure for population-level substance use or for capturing frequent use in specific patient groups.