Nonfatal Drug Overdoses Treated in Emergency Departments — United States, 2016–2017

In 2017, drug overdoses caused 70,237 deaths in the United States, a 9.6% rate increase from 2016 (1). Monitoring nonfatal drug overdoses treated in emergency departments (EDs) is also important to inform community prevention and response activities. Analysis of discharge data provides insights into the prevalence and trends of nonfatal drug overdoses, highlighting opportunities for public health action to prevent overdoses. Using discharge data from the Healthcare Cost and Utilization Project’s (HCUP) Nationwide Emergency Department Sample (NEDS), CDC identified nonfatal overdoses for all drugs, all opioids, nonheroin opioids, heroin, benzodiazepines, and cocaine and examined changes from 2016 to 2017, stratified by drug type and by patient, facility, and visit characteristics. In 2017, the most recent year for which population-level estimates of nonfatal overdoses can be generated, a total of 967,615 nonfatal drug overdoses were treated in EDs, an increase of 4.3% from 2016, which included 305,623 opioid-involved overdoses, a 3.1% increase from 2016. From 2016 to 2017, the nonfatal overdose rates for all drug types increased significantly except for those involving benzodiazepines. These findings highlight the importance of continued surveillance of nonfatal drug overdoses treated in EDs to inform public health actions and, working collaboratively with clinical and public safety partners, to link patients to needed recovery and treatment resources (e.g., medication-assisted treatment).

The 2017 HCUP NEDS data set is a nationally representative, stratified sample of ED visits from nonfederal, hospital-based EDs in 36 U.S. states and the District of Columbia.* Hospital discharge data represent the reference standard in nonfatal overdose surveillance and allow generation of population-level estimates to examine rate changes over time. Using 2016 and 2017 NEDS data, six drug overdose indicators were classified using International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) discharge diagnosis codes: 1) all-drugs, 2) all opioids, 3) nonheroin opioids, 4) heroin, 5) benzodiazepines, and 6) cocaine. All diagnosis fields were searched for initial encounter^† visits for intent (i.e., unintentional, intentional self-harm, assault, and undetermined).^§ Polysubstance overdoses could be classified under multiple overdose indicators; thus, groups are not mutually exclusive.

Annual rates for drug overdose per 100,000 population were calculated by sex, age group, U.S. Census region of facility,^¶ county urbanization level of facility,** and intent. All rates, except age group, were age-adjusted.^†† Absolute and relative rate changes^§§ were calculated from 2016 to 2017 by patient, facility, and visit characteristics for each overdose indicator; z-tests were used to compare changes that occurred from 2016 to 2017 and for pairwise comparisons between groups for 2017 rates, with p-values <0.05 considered statistically significant. Only selected comparisons were tested for statistical significance, and all results presented were statistically significant. Analyses were conducted using SAS (version 9.4; SAS Institute) to account for HCUP’s complex survey design and weighting.

In 2017, there were 967,615 nonfatal drug overdose ED visits (300.2 per 100,000 population) (Table 1). From 2016 to 2017, rates for nonfatal overdoses increased for those involving all drugs (4.3%), all opioids (3.1%), nonheroin opioids (3.6%), heroin (3.6%), and cocaine (32.9%), whereas the rate for overdoses involving benzodiazepines decreased 5.2% (Table 1) (Table 2) (Table 3).

In 2017, the highest overdose rates for all drugs were among females (308.2), persons aged 15–34 years (range = 427.1–476.4), persons in the Midwest (378.6), and persons in micropolitan (nonmetro) counties (363.3) (Table 1). From 2016 to 2017, overdose rates for all drugs increased 5.0% among males and 3.7% among females. The highest overdose rates for all opioids were among males (112.6), persons aged 25–34 years (209.3), persons in the Midwest (129.2), and persons in medium metro counties (111.4). Rates for all opioid overdoses decreased 4.7% among persons aged 0–14 years, 10.5% in persons aged 15–19 years, and 9.6% among persons aged 20–24 years. In the Midwest, overdose rates for all drugs increased by 6.1% and for all opioids by 7.9%; in the South rates for all drugs and all opioids increased by 3.2% and 5.2%, respectively; and in the West by 8.1% and 3.9%, respectively. In the Northeast, the overdose rate for all drugs remained stable, and the overdose rate for all opioids decreased 5.8%.

Overdose rates for nonheroin opioids and heroin were highest among males (44.5 and 66.7, respectively), persons aged 25–34 years (63.7 and 144.3, respectively), persons in the Midwest (51.2 and 77.0, respectively), and those in medium metro counties (51.6 and 57.8, respectively) (Table 2). Increases in rates for heroin overdose were observed among males (4.1%) and females (3.0%), whereas rates for nonheroin opioid overdoses increased only among males (7.0%). Heroin overdose rates decreased 50% among persons aged 0–14 years, 21.8% among persons aged 15–19 years, and 14.1% among persons aged 20–24 years. Rates for overdoses involving nonheroin opioids and heroin increased 8.5% and 8.6% in the Midwest, respectively, and 3.4% and 8.2%, respectively, in the South. Heroin overdose rates also increased 11.3% in the West. In the Northeast, the rate for heroin-involved overdoses decreased 8.8%.

In 2017, the highest overdose rates for benzodiazepines were among females (42.3), persons aged 20–44 years (range = 50.5–51.0), persons in the Midwest (41.4), and persons in medium metro counties (43.5) (Table 3). The rates for cocaine overdoses in 2017 were highest among males (15.2), persons aged 25–34 years (19.6) and aged 45–54 years (20.5), as well as persons in the South census region (15.1) and large central metro counties (16.5). From 2016 to 2017, rates for benzodiazepine overdoses decreased 5.1% among males and 5.2% among females. Benzodiazepine overdose rates decreased among most age groups, and cocaine-involved overdoses rates increased across all age groups. All regions of the country experienced decreases in the rates of benzodiazepine overdoses and increases in the rates of cocaine overdoses.

In large central metro counties, overdose rates increased for all drugs (11.7%), all opioids (15.2%), nonheroin opioids (11.9%), heroin (21.4%), benzodiazepines (4.7%), and cocaine (71.9%) (Table 1) (Table 2) (Table 3). Most overdoses were unintentional (75% overall; range = 48% for benzodiazepines to 91% for heroin). A consistent finding across all overdose indicators, except for benzodiazepines, was that unintentional overdoses significantly increased from 2016 to 2017. Intentional self-harm overdoses increased 4.8% for all drugs but decreased 6.7% for all opioids, 7.4% for nonheroin opioids, and 3.4% for benzodiazepines.

Acknowledgment

R. Matthew Gladden, Division of Overdose Prevention, National Center for Injury Prevention and Control, CDC.

Corresponding author: Alana M. Vivolo-Kantor, avivolokantor@cdc.gov, 770-488-1244.

¹Division of Overdose Prevention, National Center for Injury Prevention and Control, CDC; ²Epidemic Intelligence Service, CDC; ³Oak Ridge Institute for Science and Education, Oak Ridge, Tennessee.

References

1. Hedegaard H, Miniño AM, Warner M. Drug overdose deaths in the United States, 1999–2017. NCHS data brief, no. 329. Hyattsville, MD: US Department of Health and Human Services, CDC, National Center for Health Statistics; 2018. https://www.cdc.gov/nchs/data/databriefs/db329-h.pdf CrossRef PubMed
2. Hoots B, Vivolo-Kantor A, Seth P. The rise in non-fatal and fatal overdoses involving stimulants with and without opioids in the United States. Addiction 2020. Epub January 7, 2020. CrossRef PubMed
3. Hempstead K, Phillips J. Divergence in recent trends in deaths from intentional and unintentional poisoning. Health Aff (Millwood) 2019;38:29–35. CrossRef PubMed
4. Drug Enforcement Administration, Diversion Control Division. National Forensic Laboratory Information. System: NFLIS-Drug 2017 annual report. Washington, DC: US Department of Justice, Drug Enforcement Administration; 2018. https://www.nflis.deadiversion.usdoj.gov/DesktopModules/ReportDownloads/Reports/NFLIS-Drug-AR2017.pdf
5. Kariisa M, Scholl L, Wilson N, Seth P, Hoots B. Drug overdose deaths involving cocaine and psychostimulants with abuse potential—United States, 2003–2017. MMWR Morb Mortal Wkly Rep 2019;68:388–95. CrossRef PubMed
6. Vivolo-Kantor AM, Seth P, Gladden RM, et al. Vital signs: trends in emergency department visits for suspected opioid overdoses—United States, July 2016–September 2017. MMWR Morb Mortal Wkly Rep 2018;67:279–85. CrossRef PubMed
7. Scholl L, Seth P, Kariisa M, Wilson N, Baldwin G. Drug and opioid-involved overdose deaths—United States, 2013–2017. MMWR Morb Mortal Wkly Rep 2018;67:1419–27. CrossRef PubMed
8. Morrow JB, Ropero-Miller JD, Catlin ML, et al. The opioid epidemic: moving toward an integrated, holistic analytical response. J Anal Toxicol 2019;43:1–9. CrossRef PubMed
9. Houry DE, Haegerich TM, Vivolo-Kantor A. Opportunities for prevention and intervention of opioid overdose in the emergency department. Ann Emerg Med 2018;71:688–90. CrossRef PubMed
10. Dowell D, Haegerich TM, Chou R. CDC guideline for prescribing opioids for chronic pain—United States, 2016. MMWR Recomm Rep 2016;65(No. RR-1). CrossRef PubMed

Abbreviation: SE = standard error.
* Rates are age-adjusted using the direct method and the 2000 U.S. Census standard population, except for age-specific crude rates. All rates are per 100,000 population. Statistical testing was completed using rates rounded to 1 decimal place and standard errors rounded to 3 decimal places.
^† Categories of nonfatal drug overdose visits are not mutually exclusive because overdose visits might involve more than one drug. Summing of categories will result in greater than the total number of visits in a year.
^§ Nonfatal drug overdose visits are classified using the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD–10-CM). ICD-10-CM diagnosis codes for all drugs included codes with T36-T50 with a sixth character of 1, 2, 3, or 4 (exceptions for T36.9, T37.9, T39.9, T41.4, T42.7, T43.9, T45.9, T47.9, and T49.9, which were included if the code had a fifth character of 1, 2, 3, or 4). Only codes with a seventh character of “A” (initial encounter) were included.
^¶ ICD-10-CM diagnosis codes for all opioids included T40.0X1A–T40.0X4A, T40.1X1A–T40.1X4A, T40.2X1A–T40.2X4A, T40.3X1A–T40.3X4A, T40.4X1A–T40.4X4A, T40.601A–T40.604A, and T40.691A–T40.694A.
** Absolute rate change is the difference in rates from 2016 to 2017. Relative rate change is the absolute rate change divided by the 2016 rate, multiplied by 100. Z-tests were used to determine significance.
^†† Statistically significant (p-value <0.05).
^§§ Facility geographic regions were derived from U.S. Census regions: https://www.hcup-us.ahrq.gov/db/vars/hosp_region/nedsnote.jsp. Northeast: Connecticut, Maine, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania, Rhode Island, and Vermont. Midwest: Illinois, Indiana, Iowa, Kansas, Michigan, Minnesota, Missouri, Nebraska, North Dakota, Ohio, South Dakota, and Wisconsin. South: Alabama, Arkansas, Delaware, District of Columbia, Florida, Georgia, Kentucky, Louisiana, Maryland, Mississippi, North Carolina, Oklahoma, South Carolina, Tennessee, Texas, Virginia, and West Virginia. West: Alaska, Arizona, California, Colorado, Hawaii, Idaho, Montana, Nevada, New Mexico, Oregon, Utah, Washington, and Wyoming.
^¶¶ County urbanization levels for facilities were determined using the 2013 National Center for Health Statistics Urban-Rural Classification Scheme for Counties. https://www.cdc.gov/nchs/data_access/urban_rural.htm.
*** In ICD-10-CM, the fifth or sixth character in the diagnosis code indicates intent. Possible values include accidental (unintentional), intentional self-harm, assault, undetermined intent, adverse effect, and underdosing. Adverse effect and underdosing are not applicable values for all of the different drug poisoning diagnosis codes. In this report, the intent was set to “Missing” for emergency department visits with multiple overdose intents listed.

Abbreviation: SE = standard error.
* Rates are age-adjusted using the direct method and the 2000 U.S. Census standard population, except for age-specific crude rates. All rates are per 100,000 population. Statistical testing was completed using rates rounded to one decimal place and standard errors rounded to three decimal places.
^† Categories of nonfatal drug overdose visits are not mutually exclusive because overdose visits might involve more than one drug. Summing of categories will result in greater than the total number of visits in a year.
^§ Nonfatal drug overdose visits are classified using the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD–10-CM). ICD-10-CM diagnosis codes for nonheroin opioids included T40.0X1A–T40.0X4A, T40.2X1A–T40.2X4A, T40.3X1A–T40.3X4A, T40.4X1A–T40.4X4A, T40.601A–T40.604A, and T40.691A–T40.694A.
^¶ ICD-10-CM diagnosis codes for heroin included T40.1X1A–T40.1X4A.
** Absolute rate change is the difference in rates from 2016 to 2017. Relative rate change is the absolute rate change divided by the 2016 rate, multiplied by 100. Z-tests were used to determine significance.
^†† Statistically significant (p-value <0.05).
^§§ Facility geographic regions were derived from U.S. Census regions: https://www.hcup-us.ahrq.gov/db/vars/hosp_region/nedsnote.jsp. Northeast: Connecticut, Maine, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania, Rhode Island, and Vermont. Midwest: Illinois, Indiana, Iowa, Kansas, Michigan, Minnesota, Missouri, Nebraska, North Dakota, Ohio, South Dakota, and Wisconsin. South: Alabama, Arkansas, Delaware, District of Columbia, Florida, Georgia, Kentucky, Louisiana, Maryland, Mississippi, North Carolina, Oklahoma, South Carolina, Tennessee, Texas, Virginia, and West Virginia. West: Alaska, Arizona, California, Colorado, Hawaii, Idaho, Montana, Nevada, New Mexico, Oregon, Utah, Washington, and Wyoming.
^¶¶ County urbanization levels for facilities were determined using the 2013 National Center for Health Statistics Urban-Rural Classification Scheme for Counties. https://www.cdc.gov/nchs/data_access/urban_rural.htm.
*** In ICD-10-CM, the fifth or sixth character in the diagnosis code indicates intent. Possible values include accidental (unintentional), intentional self-harm, assault, undetermined intent, adverse effect, and underdosing. Adverse effect and underdosing are not applicable values for all of the different drug poisoning diagnosis codes. In this report, the intent was set to “Missing” for emergency department visits with multiple overdose intents listed.

Abbreviation: SE = standard error.
* Rates are age-adjusted using the direct method and the 2000 U.S. Census standard population, except for age-specific crude rates. All rates are per 100,000 population. Statistical testing was completed using rates rounded to one decimal place and standard errors rounded to three decimal places.
^† Categories of nonfatal drug overdose visits are not mutually exclusive because overdose visits might involve more than one drug. Summing of categories will result in greater than the total number of visits in a year.
^§ Nonfatal drug overdose visits are classified using the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD–10-CM). ICD-10-CM diagnosis codes for benzodiazepines included T42.4X1A–T42.4X4A.
^¶ ICD-10-CM diagnosis codes for cocaine included T40.5X1A–T40.5X4A.
** Absolute rate change is the difference in rates from 2016 to 2017. Relative rate change is the absolute rate change divided by the 2016 rate, multiplied by 100. Z-tests were used to determine significance.
^†† Statistically significant (p-value <0.05).
^§§ Facility geographic regions were derived from U.S. Census regions: https://www.hcup-us.ahrq.gov/db/vars/hosp_region/nedsnote.jsp. Northeast: Connecticut, Maine, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania, Rhode Island, and Vermont. Midwest: Illinois, Indiana, Iowa, Kansas, Michigan, Minnesota, Missouri, Nebraska, North Dakota, Ohio, South Dakota, and Wisconsin. South: Alabama, Arkansas, Delaware, District of Columbia, Florida, Georgia, Kentucky, Louisiana, Maryland, Mississippi, North Carolina, Oklahoma, South Carolina, Tennessee, Texas, Virginia, and West Virginia. West: Alaska, Arizona, California, Colorado, Hawaii, Idaho, Montana, Nevada, New Mexico, Oregon, Utah, Washington, and Wyoming.
^¶¶ County urbanization levels for facilities were determined using the 2013 National Center for Health Statistics Urban-Rural Classification Scheme for Counties. https://www.cdc.gov/nchs/data_access/urban_rural.htm.
*** In ICD-10-CM, the fifth or sixth character in the diagnosis code indicates intent. Possible values include accidental (unintentional), intentional self-harm, assault, undetermined intent, adverse effect, and underdosing. Adverse effect and underdosing are not applicable values for all of the different drug poisoning diagnosis codes. In this report, the intent was set to “Missing” for emergency department visits with multiple overdose intents listed.