Abstract
Background and aims:
It is widely believed that the 2018 decline in overdose deaths in the United States was attributable to a range of public health interventions; however, this decline also coincided with the regulation and decline in use of potent fentanyl analogs, especially carfentanil. The aim of this study was to investigate the association between overdose deaths and carfentanil availability in the United States.
Design:
Secondary analysis of drug overdose deaths from the Center for Disease Control and carfentanil exhibit data from drug seizures submitted to drug crime labs and published by the Drug Enforcement Administration (DEA). Trends in overdose deaths were compared in states with high carfentanil exhibits with states with low or no carfentanil exhibits.
Setting:
United States.
Cases:
A total of 1,035,923 drug overdose death records in the United States from 1979 through 2019 were studied.
Measurements:
The outcomes studied were number of overdose deaths and mortality rates by state.
Findings:
Drug overdose deaths have been closely tracking along an exponential curve. The years 2016 and 2017 witnessed a hyper-exponential surge with increases in overdose deaths of 11,228 (+21.4%), and 6,605 (+10.4%), respectively. Subsequently in 2018, drug overdose deaths declined by −2,870 (−4.1%). This rise and then fall coincided with a surge and then decline in carfentanil drug seizure exhibits during these same years: 0 (2015), 1292 (2016), 5,857 (2017) and 804 (2018). The majority of carfentanil exhibits were localized to a few states. The 2018 decline in overdose deaths in the top five states with the greatest spike in carfentanil exhibits in 2017 (Ohio, Florida, Pennsylvania, Kentucky and Michigan) was 2,848, which accounted for nearly all of the total US decline.
Conclusions:
The 2016-2017 acceleration and then 2018 decline in drug overdose deaths in the United States was associated with the sudden rise and then fall of carfentanil availability. Given the regional variation, carfentanil’s decreased availability may have contributed to the reduction in overdose deaths in 2018.
Keywords: opioid epidemic, drug waves, carfentanil, fentanyl, overdose, deaths
Introduction:
Annual mortality rates for drug overdoses in the United States have been inexorably growing along an exponential curve for decades.(1) In 2018, for the first time since 1990, the annual number of drug overdose deaths in the US fell, from 70,237 in 2017 to 67,367 in 2018, a decline of 2,870 deaths, or −4.1%.(2) The New York Times reported “Three decades of ever-escalating deaths from drug overdoses in the United States may have come to an end.”(3) And in his 2020 State of the Union Address, President Trump asserted that “With unyielding commitment, we are curbing the opioid epidemic. Drug overdose deaths declined for the first time in nearly 30 years.” However, in 2019, the number of overdose deaths have increased again along the historical exponential trajectory and above their 2017 level to 71,148, a +5.6% increase, according to provisional data released by the CDC.(4) Thus, it is not clear if the 2018 decline was the start of a long-term trend, or a transient change due to other factors, such as fluctuations in the supply of deadlier new drugs.(5)
The exponential growth of overdose deaths has involved three overlapping opioid waves: prescription opioids, heroin and synthetic opioids, including fentanyl and its analogs.(6,7) Since 2016, the majority of overdose deaths have been linked to synthetic opioids. The surge in availability of fentanyl and its analogs in the US was driven by market forces, including the increased ease of production, increased demand for cheaper opioids, and the ease of online sales of these chemicals.(8)
Carfentanil is a fentanyl analog which is 10,000 times more potent than morphine, and responsible for many overdose outbreaks in the US.(9,10) Carfentanil’s availability increased in 2016, peaked in 2017, and suddenly declined in 2018.(9) It is not clear if the fluctuation in carfentanil’s availability can explain the rise and fall of overdose deaths in the US. This report examines the relationship between trends of carfentanil exhibits submitted to state crime labs and overdose deaths by state.
Methods:
We used drug overdose death data which are publicly available from CDC Wonder.(11) Drug-poisoning (overdose) deaths are identified using underlying cause-of-death International Classification of Diseases 9th revision (ICD-9) codes E850.0-E950.5, E950.0-E950.5, E962.0, E962.1, E962.9 and E980 for the period prior to 1999, and ICD-10 codes X40–X44, X60–X64, X85, and Y10–Y14 for the period starting in 1999. First, we fit an exponential curve to the annual trend of overdose deaths from 1979 through 2015, where log(Deaths) = α + β * year + ε, where α is the intercept, β is regression coefficient of year and represents the annual growth rate of overdose deaths, and ε is the residual term. We examined the 2018 decline and the 2019 increase in the context of the historical exponential trajectory through 2015.
The International Classification of Diseases – 10th revision (ICD-10) which is used to identify the contributory causes of death does not distinguish between fentanyl and its various analogs. Therefore, we used data on drug exhibits submitted to state crime labs to compare to overdose deaths trends by state. The US Drug Enforcement Agency’s (DEA) National Forensic Laboratory Information System (NFLIS) publishes annual reports of the number of times certain chemicals are counted from submitted samples to state drug crime labs.(12) This data collects information from a national network of state and local forensic laboratories and adjusts for variation in reporting by the individual labs.(13)
We examined the change in monthly overdose death data in relation to trends in carfentanil compared to heroin and fentanyl [Supplementary Figure S1]. To reduce the monthly fluctuation in the state overdose death counts and rates, we report a 3-month moving average of these statistics. In addition, we performed a multivariate regression analysis to examine the relationship between the 2018 change in overdose deaths as a function of state-level changes in heroin, fentanyl and carfentanil exhibits. The analysis plan was not pre-registered on a publicly available platform. Thus, the results should be considered exploratory.
Results:
There were 1,035,923 drug overdose deaths in the US from 1979 through 2019. Figure 1 shows the 2018 decline in relation to the overall exponential growth. An exponential fit to the data from 1979 through 2015 reveals an annual growth rate of 0.0649 (p-value < 0.0001) indicating overdose deaths doubling every 10.7 years (R2=0.982). Overdose deaths spiked in 2016 and 2017 well beyond the exponential trend. During this “hyper” exponential spike, the rates of overdose deaths increased by +22% in 2016 and +11% in 2017. This spike resulted in the upward swing of the overdose curve. Even with the −4.1% decline, 2018 deaths were still above the historical exponential trajectory. In 2019, the number of overdose deaths increased again above their 2017 level to 71,148, a +5.6% increase from 2018.(4)
Figure 1.
The hyper-exponential rise in overdose deaths in 2016-2017, and the deceleration towards the exponential trajectory in 2018. The annual number of overdose deaths are shows from 1979 through 2018, with the exponential fit through 2015 which is projected through 2018. The gray shaded area represents the 95% confidence interval of the fit.
State-level analysis of the NFLIS data reveals that the 2017 spike in carfentanil exhibits was concentrated in five states: Ohio, Florida, Pennsylvania, Kentucky and Michigan. Figure 2 compares the relationship between trends of carfentanil exhibits and overdose deaths by state from 2016 through 2018. The five states with the largest carfentanil exhibits are highlighted in red.
Figure 2.
The relationship between rise and fall of carfentanil exhibits and number of overdose deaths by state in 2016, 2017 and 2018. State abbreviations are shown in the boxes. The red arrows represent states with the highest spike in carfentanil exhibits in 2017 (Ohio, Florida, Pennsylvania, Kentucky and Michigan). The rest of the states are shown in blue. The lighter arrows indicate changes from 2016 to 2017, and the darker arrows indicate changes from 2017 to 2018. The insert table represents the changes in overdose deaths from 2016 to 2017, and from 2017 to 2018 by state category. To improve clarity of presentation, zero exhibits were replaced with a small random number between 1 and 2, and the labels for states with low mortality and carfentanil exhibits in the lower left corner of the plot were removed: NM, NH, OR, ME, RI, KS, IA, MS, AR, DE, DC, ID, HI, WY, VT, SD, NE, MT and AK.
The insert table in Figure 2 summarizes the relationship between carfentanil seizure and overdose deaths. The top five carfentanil states witnessed the largest increase in overdose deaths +13.7% in 2017 followed by the largest decline −14.3% in 2018. The rise in overdose deaths in the rest of the states was smaller +9.1% in 2017, and collectively overdose deaths in 2018 remain steady. In addition, the top four states with the largest rise and then fall in overdose deaths in 2018 are those same states with the highest rise and fall in carfentanil. For example, from 2016 to 2017, Ohio had the largest increase in overdose deaths (+782) and carfentanil exhibits (+3540) among all the states. From 2017 to 2018, Ohio again had the largest decline in overdose deaths (−1131) and carfentanil exhibits (−3993). Figure 2 also shows the similar association between carfentanil exhibit trends and overdose death trends for Pennsylvania, Florida, and Kentucky. Michigan ranked tenth in overdose death reduction with a decline of −103 deaths.
The total decline in overdose deaths from these five states with the highest decline in carfentanil exhibits account for nearly 100% of the total national 2017 to 2018 reduction of −2,870 deaths. The states with largest increase in overdose deaths in 2017 to 2018 [California (+480 deaths), Missouri (+243 deaths), New Jersey (+215 deaths), and Arizona (+138 deaths)] were less impacted by the carfentanil surge (<10 carfentanil exhibits each in any year). These states trends appear on the left side in Figure 2.
Figure 3 shows the monthly overdose mortality rates for the five high carfentanil exhibit states. The 3-month moving average for all these states tend to increase through the end of 2016. In early 2017, the trend moves downward while the trend for all the other states remain mostly stable in 2017 and 2018. [In the supplementary figure, we show this trend of overdose deaths by state together with trends of fentanyl and carfentanil exhibits.]
Figure 3.
Trends of overdose deaths in states with high carfentanil exhibits compared to all the other states. The trends represent the moving average of overdose deaths in a three-month window. The moving average is aligned at the center of each window.
Table 1 presents the results of a multivariate regression analysis of the state-level changes in overdose deaths per capita as a function of changes in heroin, fentanyl and carfentanil exhibits per capita. As expected, the coefficients for the changes in exhibits for all the drugs were positive. The change in carfentanil exhibits had the highest association with overdose deaths. On average, a reduction of one carfentanil exhibit per capita was associated with 0.327 reduction in deaths per capita. This association was statistically significant (p-value < 0.0003), while the impact of changes in heroin and fentanyl exhibits were smaller and not statistically significant.
Table 1:
Multivariate regression analysis of the association between changes in carfentanil, fentanyl and heroin exhibit per capita and the change in overdose deaths per capita from 2017 to 2018.
Coefficients | Estimate | Std. Error | t-value | p-value |
---|---|---|---|---|
Intercept | −0.000001 | 0.000005 | −2.049 | 0.0461* |
Change in carfentanil exhibits per capita from 2017 to 2018 | 0.327 | 0.0845 | 3.864 | 0.0003*** |
Change in fentanyl exhibits per capita from 2017 to 2018 | 0.0627 | 0.044 | 1.425 | 0.1609 |
Change in heroin exhibits per capita from 2017 to 2018 | 0.0132 | 0.0367 | 0.36 | 0.7202 |
Significant codes: 0
‘’ 0.001
‘’ 0.01
‘’ 0.05 ‘.’ 0.1 ‘ ’ 1
Residual standard error: 2.686e-05 on 47 degrees of freedom
Multiple R-squared: 0.2419, Adjusted R-squared: 0.1935
F-statistic: 4.998 on 3 and 47 DF, p-value: 0.004329
Discussion:
In this report, we show the association between the hyper-exponential rise and then fall of overdose deaths and the sudden appearance and then disappearance of carfentanil. The decline in overdose deaths in the five states with the highest carfentanil peak accounts for nearly all the national decline. The states that had an increase in overdose deaths in 2018 did not witness a similar prior spike in carfentanil exhibits.
Since 2013, overdose deaths involving synthetic opioids have been increasing rapidly following a heroin wave.(1,6) Although the mortality data from the CDC is limited to a single code for the synthetic opioids, textual analysis of death records indicate that most of these deaths were from fentanyl and its analogs which have doubled every year from 2013 to 2016.(14) Economic factors and advances in chemistry may have contributed to the rapid rise of fentanyl and its analogs.(9) In addition, fentanyl and its analogs are increasingly found laced with other chemicals including heroin, cocaine and methamphetamine further increasing overdose risk from these chemicals.(1,15,16)
Carfentanil has no known medical use in humans, and it was rarely detected in exhibits submitted to the DEA before 2016. The potency of carfentanil can make it an attractive substitute to heroin and even fentanyl, but correctly dosing carfentanil is extremely difficult.(9) According to a DEA document, in 2017 carfentanil was the second most reported synthetic opioid after fentanyl, mostly due to increased reporting from Ohio.(16) Carfentanil exhibits spiked in other states, but to a lesser extent, including Florida, Pennsylvania, Kentucky and Michigan. At the same time, carfentanil also surged in British Columbia, but it is not clear if this is linked to the spike in the US.(9)
Analyzing the monthly overdose deaths data, reveals that the decline in deaths started in early 2017 which coincides with the Chinese regulation of carfentanil and several other fentanyl analogs. Under increased US pressure, China added carfentanil to its list of controlled substances, effective March 1st, 2017.(17) Until that date, carfentanil was manufactured and sold legally by Chinese companies.(17) China’s regulation of carfentanil coincides with the sharp decline in carfentanil exhibits in the Ohio.(5) In addition, a review of overdose death records in Wayne County, Michigan, shows that the percent of deaths involving carfentanil dropped form 33% in Feb 2017 to 3% in March 2017.(18)
Our study has several limitations. First, the NFLIS data only reports on drug exhibits that were submitted to state drug crime laboratories. These drug exhibits measure the number of times carfentanil is counted by each of these labs. As such it does not represent the quantity of carfentanil per exhibit, if the exhibit was from a wholesale or retail, or whether it was mixed with other drugs such as cocaine or heroin. Second, some laboratories may lack the equipment needed to detect carfentanil, thus the number of carfentanil exhibits may be undercounted in some states. Another limitation of our study is that it does not account for other changes that may have contributed to the decline in overdose deaths. Around the same time, and perhaps in response to the surge in opioid related overdose deaths, many states might have expanded naloxone distribution and access. However, the rate of naloxone prescribing, and co-prescribing of naloxone with prescription opioids, in states with the highest decline in overdose deaths does not appear to be different from states that did not witness such decline in 2018.(19,20)
According to the most recent provisional counts of overdose deaths, drug overdose deaths increased in 2019 by 5.6% from 2018, again reaching their highest level yet since the start of the epidemic. This latest increase in overdose deaths is more widespread than the 2018 decline and may indicate a return towards the expected exponential trajectory. Although, there has been reports of resurgence of carfentanil in Ohio and British Columbia, NFLIS exhibit data for 2019 was not available at the time of writing this report, and it is not clear if the 2019 increase correlates with an increase in carfentanil.(21,22)
This is a particularly critical time for the opioid epidemic, given the diversion of healthcare resources and the national attention to the COVID-19 pandemic.(23–25) Individuals with opioid and substance use disorder are at particular risk of overdose due to the economic and social burdens of the COVID-19 pandemic. Many overdose deaths followed job losses in Appalachia in late 2000’s, and the COVID-19 has a much broader economic hardship.(26) Both opioids and the COVID-19 compromise the respiratory system.(23) Those who inject drugs, may do so more often at isolation and decrease their chances of receiving naloxone or timely treatment. Individuals receiving or needing treatment may have been impacted the most by the social distancing because until mid-March 2020, these individual were required a full physical examination before receiving treatment, and could only receive medications in-person in the treatment centers.(27)
The surge in overdose deaths in the US in 2017 led to a declaration of the opioid epidemic as a public health emergency. We are concerned that policy makers may have interpreted the one-year downturn in 2018 as evidence for an especially effective national response or the start of a long-term trend. The analysis we present here suggests an alternative explanation, that the decline in overdose deaths in 2018 was just a return to baseline exponential trajectory after a transient supply side shock of highly potent illicit drugs. Domestic public health efforts should be reinforced to prevent the continued growth of opioid and drug overdose deaths in the US.
Supplementary Material
Acknowledgments.
Funding.
H.J. is supported by NIH/NIDA Grant K01DA048985.
Footnotes
Competing interests: Authors declare no competing interests.
Data and materials availability: All data used in this analysis is publicly available as described in the Methods section.
References:
- 1.Jalal H, Buchanich JM, Roberts MS, Balmert LC, Zhang K, Burke DS. Changing dynamics of the drug overdose epidemic in the United States from 1979 through 2016. Science (80-). 2018;361(6408):6408. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Hedegaard H, Miniño AM, Warner M. Drug Overdose Deaths in the United States, 1999-2018 Key findings Data from the National Vital Statistics System, Mortality. 2020;(356):1–8. Available from: https://www.cdc.gov/nchs/products/index.htm. [Google Scholar]
- 3.Goodnough A, Katz J, Sanger-Katz M. Drug Overdose Deaths Drop in U.S. for First Time Since 1990. The New York Times; [Internet]. 2019. July 17; Available from: https://www.nytimes.com/interactive/2019/07/17/upshot/drug-overdose-deaths-fall.html [Google Scholar]
- 4.Ahmad F, Escobedo L, Rossen L, Spencer M, Warner M, Sutton P. Provisional drug overdose death counts. National Center for Health Statistics. [Internet]. National Center for Health Statistics. 2020. Available from: https://www.cdc.gov/nchs/nvss/vsrr/drug-overdose-data.htm [Google Scholar]
- 5.Rosenblum D, Unick J, Ciccarone D. The Rapidly Changing US Illicit Drug Market and the Potential for an Improved Early Warning System: Evidence from Ohio Drug Crime Labs. Drug Alcohol Depend. 2020;208(January). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Ciccarone D The triple wave epidemic: Supply and demand drivers of the US opioid overdose crisis [Internet]. International Journal of Drug Policy. Elsevier; 2019. p. 0–1. Available from: 10.1016/j.drugpo.2019.01.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Compton WM, Jones CM, Baldwin GT. Relationship between Nonmedical Prescription-Opioid Use and Heroin Use. N Engl J Med. 01/14. 2016;374(2):154–63. [DOI] [PubMed] [Google Scholar]
- 8.McLaughlin K Underground labs in China are devising potent new opiates faster than authorities can respond. Science [Internet]. 2017. March 29; Available from: https://www.sciencemag.org/news/2017/03/underground-labs-china-are-devising-potent-new-opiates-faster-authorities-can-respond
- 9.Pardo B, Taylor J, Caulkins J, Kilmer B, Reuter P, Stein B. The Future of Fentanyl and Other Synthetic Opioids. RAND Corportation. 2019. [Google Scholar]
- 10.Kreek MJ, Reed B, Butelman ER. Current status of opioid addiction treatment and related preclinical research. Sci Adv [Internet]. 2019. [cited 2020 Apr 15];5(10). Available from: https://advances.sciencemag.org/content/5/10/eaax9140?utm_source=TrendMD&utm_medium=cpc&utm_campaign=TrendMD_1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Underlying Cause of Death 1979-2018 on CDC WONDER Online Database, released in 2020. Data are from the Multiple Cause of Death Files, 1979-2018. Centers for Disease Control and Prevention National Center for Health Statistics. 2020. [Google Scholar]
- 12.U.S. Drug Enforcement Administration, Diversion Control Division. (2019). NFLIS-Drug public use data table title. [Internet]. Available from: https://www.nflis.deadiversion.usdoj.gov/Resources/NFLISPublicResourceLibrary.aspx
- 13.Frechtel P, Heller D, Mosquin P. National Forensic Laboratory Information System: Statistical Methodology. Springfield, VA; 2017. [Google Scholar]
- 14.Hedegaard H, Bastian BA, Trinidad JP, Spencer M, Warner M. Drugs most frequently involved in drug overdose deaths: United states, 2011-2016. Natl Vital Stat Reports. 2018;67(9). [PubMed] [Google Scholar]
- 15.Zibbell JE, Aldridge AP, Cauchon D, DeFiore-Hyrmer J, Conway KP. Association of Law Enforcement Seizures of Heroin, Fentanyl, and Carfentanil With Opioid Overdose Deaths in Ohio, 2014-2017. JAMA Netw open. 2019;2(11):e1914666. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.DEA. 2019 National Drug Threat Assessment, Drug Enforcement Administration [Internet]. 2019. Available from: https://www.dea.gov/documents/2020/01/30/2019-national-drug-threat-assessment
- 17.Kinetz E, Butler D. Chemical weapon for sale: China’s unregulated narcotic. Associated Press; [Internet]. 2016. October 7 [cited 2020 Apr 14]; Available from: https://apnews.com/7c85cda5658e46f3a3be95a367f727e6/Chemical-weapon-for-sale:-China’s-unregulated-narcotic [Google Scholar]
- 18.King A, Foley D, Arfken C, Aaron C, Sung L, Hlavaty L. Carfentanil-associated mortality in Wayne County, Michigan, 2015-2017. Am J Public Health. 2019;109(2):300–2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Freeman PR, Hankosky ER, Lofwall MR, Talbert JC. The changing landscape of naloxone availability in the United States, 2011 – 2017. Drug Alcohol Depend. 2018;191(July):361–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Sohn M, Talbert JC, Huang Z, Lofwall MR, Freeman PR. Association of Naloxone Coprescription Laws with Naloxone Prescription Dispensing in the United States. JAMA Netw Open. 2019;2(6):1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Cauchon D SPECIAL REPORT: Ohio overdose deaths on rise again, exceed 4,000 in 2019 [Internet]. 2020. [cited 2020 Aug 10]. Available from: https://www.harmreductionohio.org/special-report-ohio-overdose-deaths-on-rise-again-exceed-4000-in-2019/
- 22.Fentanyl-Detected Illicit Drug Toxicity Deaths, Ministry of Public Safety adn Solicitor General, British Columbia. 2020;1–8. Available from: http://www2.gov.bc.ca/gov/content/safety/public-safety/death-investigation/statistical-
- 23.Volkow ND. Collision of the COVID-19 and Addiction Epidemics. Ann Intern Med [Internet]. 2020;2019:2019–21. Available from: https://annals.org/aim/fullarticle/2764313/collision-covid-19-addiction-epidemics [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Becker WC, Fiellin DA. When Epidemics Collide: Coronavirus Disease 2019 (COVID-19) and the Opioid Crisis. Ann Intern Med [Internet]. 2020. April 2; Available from: https://annals.org/aim/fullarticle/2764312/when-epidemics-collide-coronavirus-disease-2019-covid-19-opioid-crisis [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Alexander GC, Stoller KB, Haffajee RL, Saloner B. An Epidemic in the Midst of a Pandemic: Opioid Use Disorder and COVID-19. Ann Intern Med [Internet]. 2020;19–21. Available from: http://www.ncbi.nlm.nih.gov/pubmed/32240283 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Young AM, Havens JR. Transition from first illicit drug use to first injection drug use among rural Appalachian drug users: A cross-sectional comparison and retrospective survival analysis. Addiction. 2012;107(3):587–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Knopf A DEA, SAMHSA relax OTP/OBOT regulations due to COVID-19. Alcohol Drug Abus Wkly [Internet]. 2020. March 23 [cited 2020 Apr 15];32(12):3–5. Available from: 10.1002/adaw.32664 [DOI] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.