Dying from Drugs: A New Look at Overdose Deaths in the U.S.

All graphics presented here were derived by NIHCM using publicly-available data from the multiple cause of death files of the Wide-ranging Online Data for Epidemiologic Research (WONDER) data query system and the Provisional Drug Overdose Death Counts, both maintained by the Centers for Disease Control and Prevention (CDC).

In the WONDER query system, deaths were identified as due to a drug overdose if the ICD-10 code for the underlying cause of death was X40-X44 (accidental poisoning), X60-X64 (intentional self-poisoning) or Y10-Y14 (poisoning of undetermined intent). A small number of overdose deaths that were the result of homicide (underlying cause of death X85) were excluded from the analyses.

The substances involved in the overdose deaths were identified using the following multiple cause of death codes: all drugs, any type (T36.0-T50.9); heroin (T40.1); prescription opioids (T40.2 [natural and semi-synthetic opioids, e.g., oxycodone and hydrocodone] and T40.3 [methadone]); synthetic opioids other than methadone, e.g., fentanyl, tramadol (T40.4); cocaine (T40.5); psychostimulants with abuse potential, e.g., methamphetamine (T43.6); non-opioid pain, fever and antirheumatic medications (T39); benzodiazepines (T42.4); other antiepileptic and sedative-hypnotic drugs (T42.6); antidepressants (T43.0-T43.2); antipsychotics and neuroleptics (T43.3-T43.5); antiallergic and antiemetic drugs (T45.0); muscle relaxants (T48.1); and antitussives (T48.3). Multiple substances may be involved in a single death.

The share of deaths deemed accidental or intentional was derived by type of substance based on whether the underlying cause of death was accidental poisoning (X40-X44) or intentional self-poisoning (X60-X64). Deaths where the poisoning was of undetermined intent (Y10-Y14) were excluded from this analysis. Because deaths may often involve multiple substances, the numbers for a given substance may reflect the involvement of other substances as well. In particular, significant portions of the deaths involving substances in the right-hand panel also involve opioids and/or stimulants (ranging from a low of 46.1 percent for non-opioid pain and fever medications to a high of 82.2 percent for other antiepileptics). Despite this overlap, the analysis shows that when drugs other than opioids and stimulants are involved (either alone or with opioids and stimulants) the death is more likely to be ruled intentional.

Provisional death data are based on death records submitted by state vital registration offices and vary in their timeliness for both reporting and validation. The counts given for a specific month capture all death reports for the prior twelve-month period ending in that month, making the counts insensitive to variation in reporting across seasons. In addition to the raw provisional counts, the CDC also provides predicted provisional counts that have been adjusted to account for reporting and validation delays. The predicted provisional counts are used in this infographic. Provisional data are subject to change and are not strictly comparable with final annual death counts reported elsewhere. The same T codes as specified above were used to identify provisional deaths associated with specific substances.

Information on the COVID-19 risks associated with substance use disorders was drawn from Wang QQ, Kaelber DC, Xu R and Volkow ND. “COVID-19 Risk and Outcomes in Patients with Substance Use Disorders: Analyses from Electronic Health Records in the United States.” Molecular Psychiatry. 2021;26(1):30-39. doi:10.1038/s41380-020-00880-7

Mortality rates for urban (metropolitan) vs. rural (non-metropolitan) areas were based on the decedent’s county of residence as classified by the 2013 NCHS Urban–Rural Classification Scheme for Counties.

Speakers on NIHCM’s May 24, 2021 webinar, Addressing the Growing Overdose and Addiction Epidemic, offered a range of strategies to respond to the crisis of drug overdose deaths. These strategies are summarized broadly here using a framework centered around evidence-based prevention, harm reduction, treatment and recovery strategies and emphasizing wide-ranging partnerships, systemic reforms and improved data and research. Much richer detail on specific initiatives within this framework can be found in the speaker presentations.