Journal of Clinical Toxicology
Open Access

Abstract

Defining Controlled Substances Overdose: Should Deaths from Substance Use Disorders and Pharmaceutical Adverse Events be included?

Nabarun Dasgupta, Scott Proescholdbell, Catherine Sanford, Michele Jonsson Funk, Carri Casteel, Kurt M Ribisl and Steve Marshall

Background: The rate of mortality attributable to drug poisoning has risen consistently since the 1990s. State-based vital statistics registries estimate the incidence of drug overdose deaths using International Classification of Disease 10th revision (ICD-10) codes. Composite ICD-10-based definitions of “overdose” may include deaths that do not involve controlled substances while missing deaths that do. This has implications for incidence estimates, funding priorities, and intervention evaluation. Methods: We evaluated the impact of including substance use disorders and pharmaceutical adverse events codes in definitions of overdose. Seven proposed ICD-10-based definitions, including ones from the Injury Surveillance Workgroup (ISW) and the Centers for Disease Control and Prevention (CDC), were applied to North Carolina mortality data from 2008 through 2011. We examined whether overdose deaths varied among definitions. Results: Approximately 1.5% of all deaths among NC residents were due to overdose. Prescription opioids were involved in 63.4% of drug overdose deaths. The estimated number of overdose deaths ranged from 734 to 1,202 per year depending on which definition was used. Of deaths identified using the CDC definition, between 6.4% and 16.1% were pharmaceutical adverse events that showed no evidence of controlled substance involvement. However, there were 12 additional deaths per year that appeared to be overdoses involving controlled substances that were not identified. We propose a definition that includes 28 deaths from substance use disorders, but removes 88 deaths from adverse events, resulting in 1,149 deaths per year on average from overdoses involving controlled substances. Conclusions: Quantifying mortality from drug overdose depends on close cooperation of diverse health professionals. State health department injury control officials should carefully choose which ICD-10 codes are most relevant for their data. Researchers evaluating interventions to prevent overdoses from controlled substances may obtain a more accurate measure of effect by separating pharmaceutical adverse events from non-controlled medicines.