Defining nonfatal opioid overdoses using administrative health data in the United States and Canada: A scoping review

To explore definitions of nonfatal opioid overdoses with study populations based in the United States and Canada, to improve nonfatal overdose monitoring.

We searched five databases: PubMed, Medline, Web of Science, Embase, and PsycInfo, including studies published between January 2013-October 2024. We combined overdose-related terms (e.g., overdose, poisoning) and hospital-related terms (e.g., diagnosis code, International Classification of Diseases [ICD], electronic health records) and extracted data from 214 studies that defined opioid overdoses.

Most studies reported on opioid overdoses due to any opioid (n = 175). Studies varied in terms of specific ICD codes. Among studies using ICD, Tenth Revision, Clinical Modification (ICD-10-CM) codes (n = 79), key sources of variation included intent specifications and encounter type; nearly one-third did not specify whether codes reflected an initial encounter, a subsequent visit, or sequelae (n = 24, 30.8%). Of studies that specified, most used all three encounter types (n = 27, 34.6%) or initial only (n = 21, 26.9%). Few specified both initial and sequelae (n = 4, 5.1%) or initial and subsequent encounters (n = 2, 2.6%). Similar patterns were observed across all opioid subtype definitions.

We observed substantial variation in definitions. Detailed, standardized reporting methods when applying diagnostic codes may improve consistency and transparency and could improve public health monitoring.