Advertisement

Opioid Overdoses Increase at Home During the COVID-19 Stay-At-Home Order Period in Cook County, Illinois

  • Chris Delcher
    Correspondence
    Address correspondence to: Chris Delcher, PhD, Institute for Pharmaceutical Outcomes and Policy, University of Kentucky, 760 Press Avenue, Research Building 2, Suite 260, Lexington KY 40536-0679.
    Affiliations
    Institute for Pharmaceutical Outcomes and Policy, UK College of Pharmacy, University of Kentucky, Lexington, Kentucky

    Department of Pharmacy Practice and Science, UK College of Pharmacy, University of Kentucky, Lexington, Kentucky
    Search for articles by this author
  • Daniel R. Harris
    Affiliations
    Institute for Pharmaceutical Outcomes and Policy, UK College of Pharmacy, University of Kentucky, Lexington, Kentucky

    Department of Pharmacy Practice and Science, UK College of Pharmacy, University of Kentucky, Lexington, Kentucky
    Search for articles by this author
  • Nicholas Anthony
    Affiliations
    Institute for Pharmaceutical Outcomes and Policy, UK College of Pharmacy, University of Kentucky, Lexington, Kentucky

    Department of Pharmacy Practice and Science, UK College of Pharmacy, University of Kentucky, Lexington, Kentucky
    Search for articles by this author
  • Mojde Mir
    Affiliations
    Cook County Medical Examiner's Office, Chicago, Illinois
    Search for articles by this author
Open AccessPublished:June 15, 2022DOI:https://doi.org/10.1016/j.focus.2022.100007

      Highlights

      • The odds of opioid-involved deaths at home were 37% higher during stay-at-home orders.
      • The displacement of heroin by fentanyl was largely uninterrupted during this period.
      • Open data from the medical examiner is a rich source of specific death locations.
      • Repeat locations of fatal overdoses could be suitable for naloxone kit placement.

      Introduction

      Stay-at-home orders during the COVID-19 pandemic decreased population mobility to reduce SARS-CoV-2 infection rates. We empirically tested the hypothesis that this public health measure was associated with a higher likelihood of opioid- and stimulant-involved deaths occurring in homes located in Cook County, Illinois.

      Methods

      The stay-at-home period was from March 21, 2020 to May 30, 2020. We analyzed overdose data from the Cook County Medical Examiner's Office using a death location description from case investigations categorized as home, medical, motel, scene, and other. Two groups of decedents were defined as either having an opioid or stimulant listed in the primary cause of death field. We modeled a weekly time series to detect changes in deaths (number) and trends during segmented time periods. Chi-square or Fisher's exact and adjusted logistic regression was used for testing the differences between the stay-at-home and a 13-week preceding period.

      Results

      There were 4,169 and 2,012 opioid- and stimulant-involved deaths, respectively, from 2018 to 2020. Both groups were demographically similar: 75% male, 52% White, and aged 45 years (mean). In the 13 weeks before stay-at-home orders, 51% of opioid-involved deaths occurred in homes, which increased to 59% (p<0.0001) during the 10 weeks of the order and decreased back to 51% in the 18 weeks after the order expired. For stimulant-involved deaths, 51% were residential immediately before the orders, with a nonsignificant increase to 52% during the stay-at-home period. Before the pandemic, there were 20 deaths/week, increasing to 37 deaths/week (p<0.0001) during stay-at-home enactment. Deaths involving fentanyl among the opioid-involved group increased from 76% to 89%, whereas those involving heroin decreased from 55% to 37%. The adjusted OR for opioid-involved fatal overdoses occurring at home during this period compared with that occurring the 13 weeks before was 1.37 (95% CI=1.05, 1.79).

      Conclusions

      The likelihood of a death occurring at home, especially for people using opioids, increased during the stay-at-home order period. Findings have implications for mitigating overdose risks during social isolation.

      Keywords

      INTRODUCTION

      The stay-at-home orders issued during early 2020 in response to the coronavirus disease 2019 (COVID-19) pandemic decreased population mobility to reduce the spread of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2).
      • Moreland A
      • Herlihy C
      • Tynan MA
      • et al.
      Timing of state and territorial COVID-19 stay-at-home orders and changes in population movement - United States, March 1-May 31, 2020.
      • Murray T.
      Stay-at-home orders, mobility patterns, and spread of COVID-19.
      • Nouvellet P
      • Bhatia S
      • Cori A
      • et al.
      Reduction in mobility and COVID-19 transmission.
      The impact of stay-at-home orders on people with substance use disorders was exacerbated for multiple reasons, including increased mental health stressors, reduced medication access, and social isolation.
      • Czeisler MÉ
      • Lane RI
      • Petrosky E
      • et al.
      Mental health, substance use, and suicidal ideation during the COVID-19 pandemic - United States, June 24-30, 2020.
      ,
      • Mason M
      • Arukumar P
      • Feinglass J.
      The pandemic stay-at-home order and opioid-involved overdose fatalities.
      According to the U.S. Centers for Disease Control and Prevention (CDC), Cook County, Illinois, home to the city of Chicago, was ranked second in the number of fatal overdoses in 2020 with 1,927 deaths, representing 2% of the U.S. total.

      CDC Wide-Ranging Online Data for Epidemiologic Research (WONDER). Centers for Disease Control and Prevention, National Center for Health Statistics. http://wonder.cdc.gov. Updated December 22, 2021. Accessed July 4, 2022.

      Furthermore, the county was ranked in the top 20% of all U.S. counties for drug overdoses per 100,000 people in 2018 and was predicted to transition to the top 10% according to one model.
      • Marks C
      • Abramovitz D
      • Donnelly CA
      • et al.
      Identifying counties at risk of high overdose mortality burden during the emerging fentanyl epidemic in the USA: a predictive statistical modelling study.
      ,
      County-level mobility data clearly show that the stay-at-home order, the 10-week period of March 21, 2020–May 30, 2020, restricted the movement of the general population in Cook County.
      • Unwin H
      • Mishra S
      • Bradley V
      • et al.
      Report 23: State-Level Tracking of COVID-19 in the United States.
      During this time, the number of opioid-related deaths increased by 23% from a baseline of 35.1‒43.4 per week in the county.
      • Mason M
      • Welch SB
      • Arunkumar P
      • Post LA
      • Feinglass JM.
      Notes from the field: opioid overdose deaths before, during, and after an 11-week COVID-19 stay-at-home order – Cook County, Illinois, January 1, 2018-October 6, 2020.
      Weekly overdoses declined after orders were lifted but remained elevated well into late 2020 relative to the prepandemic levels.
      • Mason M
      • Arukumar P
      • Feinglass J.
      The pandemic stay-at-home order and opioid-involved overdose fatalities.
      The well-documented intense increase of drug overdoses during the stay-at-home-order period in this urban/suburban context provides an opportunity to understand additional epidemiologic features at the intersection of COVID-19 and the substance use epidemic such as changes in place-based risk factors
      • Nesoff ED
      • Branas CC
      • Martins SS.
      The geographic distribution of fentanyl-involved overdose deaths in Cook County, Illinois.
      and differences by primary drug type. In this paper, we examine the reported locations of fatal overdoses in Cook County to determine whether the stay-at-home-order period was associated with a higher likelihood of opioid- and stimulant-involved deaths occurring in the home.

      METHODS

      Study Sample

      We accessed the Cook County Medical Examiner's Office's publicly available data on March 9, 2022. This is a deidentified data set that has information about deaths that occurred in Cook County and were under the Medical Examiner's jurisdiction.

      Medical examiner case archive. Cook County Open Data.https://datacatalog.cookcountyil.gov/Public-Safety/Medical-Examiner-Case-Archive/cjeq-bs86. Updated March 6, 2022. Accessed March 6, 2022.

      The cause of death was based on autopsy and toxicology findings.

      Measures

      A descriptive location of death (e.g., apartment, hospital, park) is routinely collected, and we requested it separately. We grouped locations into 5 categories: home, medical, motel, scene, and other (see the Appendix for classification rules). Locations described as scene or missing values were classified as home only when the (1) geocoded street address of the overdose and death location matched exactly (95% of records) and (2) coordinates of the overdose address mapped to either a single-family or multifamily housing land-use category from the Chicago Metropolitan Agency for Planning's 2015 Land Use Inventory file (the most recent year available).

      Land use inventories – CMAP. Data Hub. https://datahub.cmap.illinois.gov/group/land-use-inventories. Updated November 23, 2020. Accessed April 26, 2021.

      We defined 2 groups of decedents by either having an opioid or having a stimulant listed in text as the primary cause of death. Drug counts are not mutually exclusive because some determinations include multiple drug classes co-listed in the primary cause of death. We created a customized and publicly available drug name extraction tool for this purpose.

      University of Kentucky, Institute for Pharmaceutical Outcomes and Policy (UK-IPOP): drug extraction. GitHub. https://github.com/UK-IPOP/drug-extraction. Updated June 13, 2022. Accessed July 4, 2022.

      The prevalence of co-occurring substances in opioid/stimulant classes and specific substances such as fentanyl, heroin, benzodiazepines, cocaine, methamphetamine, amphetamine, and alcohol were included.
      We specified 5 time frames for this study: (1) prepandemic period (January 1, 2018–December 15, 2019), (2) 13-week preorder period (December 16, 2019–March 20, 2020), (3) 10-week stay-at-home-order period (March 21, 2020–May 30, 2020), (4) 18-week postorder period (May 31, 2020–October 6, 2020), and (5) 12-week remainder period (October 7, 2020–December 31, 2020). These time periods were defined around the stay-at-home order to capture stepwise changes in the number of deaths with an additional 12-week period to complete the time series for 2020.
      • Mason M
      • Welch SB
      • Arunkumar P
      • Post LA
      • Feinglass JM.
      Notes from the field: opioid overdose deaths before, during, and after an 11-week COVID-19 stay-at-home order – Cook County, Illinois, January 1, 2018-October 6, 2020.

      Statistical Analysis

      An autoregressive model was used to account for temporal correlation (if necessary) in the weekly intensive time series and test for changes in the number of deaths and linear trends during the segmented time periods. Chi-square or Fisher's exact testing was used for testing the differences between the 10-week stay-at-home period and the 13-week preorder period immediately preceding it. Logistic regression was used to examine the odds of fatal overdose at home using a parsimonious model adjustment for demographics, including age, race, and sex. SAS 9.4 software was used for all analyses.

      SAS software version 9.4 for Windows. SAS Institute Inc., Cary, NC; 2002.

      This data set is public and deidentified.

      Medical examiner case archive. Cook County Open Data.https://datacatalog.cookcountyil.gov/Public-Safety/Medical-Examiner-Case-Archive/cjeq-bs86. Updated March 6, 2022. Accessed March 6, 2022.

      RESULTS

      In Cook County, Illinois, there were 4,169 opioid- and 2,012 stimulant-involved deaths (i.e., listed by the medical examiner as the primary cause of death) from January 1, 2018 to December 31, 2020. The highest number of opioid- and stimulant-involved deaths in Cook County for any week in this 3-year period occurred during the stay-at-home order: 57 and 34 fatal overdoses, respectively. In the 2-year prepandemic period, there were 20 opioid-involved deaths per week on average, which changed in a stepwise fashion: increasing first to 26 deaths/week (p=0.0002), then to 37 deaths/week (p<0.0001) during the stay-at-home order, and returning to prepandemic levels by the end of 2020. Figure 1A shows the number of weekly opioid- and stimulant-involved fatal overdoses from 2018 to 2020.
      Figure 1
      Figure 1(A) Opioid- and stimulant-involved deaths, weekly counts, Cook County and Illinois, 2018–2020. (B) Percentage of opioid-involved deaths with fentanyl and heroin present. Linear and AR trendlines are shown with estimated level changes in periods (only the opioids and fentanyl series had a seasonal autoregressive component). The 10-week stay-at-home order period (March 21, 2020–May 30, 2020) is shown with the darkest shading.
      Apr, April; AR, autoregressive; Aug, August; Dec, December; Jan, January.
      The trend in opioid-involved deaths was increasing significantly by 0.05 deaths/week (p=0.0176) in the prepandemic period. For the opioid-involved weekly time series, we accounted for significant autocorrelation detected by the model. For stimulants, the prepandemic number of weekly deaths was 9.5, abruptly increasing during stay-at-home orders to 16.3 (p=0.0001) and immediately returning to prepandemic levels. The trend in stimulant-involved deaths was increasing significantly by 0.034 deaths/week (p=0.0177) in the prepandemic period.
      For the demographics of opioid-involved deaths, 75% were male, 52% were White, the mean age was 45 years, the vast majority were accidental (97%), and 1.4% were coded as suicide. Trends in the presence of co-occurring drugs from the prepandemic period to the end of 2020 included proportional increases in fentanyl-involved cases (76%–89%), decreases in heroin (55%–37%), and benzodiazepine involvement (21%–13%). Figure 1B shows the weekly proportion of opioid-involved deaths attributable to fentanyl and heroin from 2018 to 2020. The age, race, and sex demographic characteristics of the stimulant-involved decedents were similar to those of the opioid-involved decedents. Trends in the presence of co-occurring drugs in this group from the prepandemic period to the end of 2020 include increases in fentanyl (57%–75%) and methamphetamine (5.8%–8.6%) and decreases in heroin (42%–28%) and benzodiazepines (18%–9%). Table 1 shows the sociodemographic, location of death, and co-occurring drug characteristics of people with an opioid- or stimulant-involved fatal overdose in Cook County for the study period.
      Table 1Opioid- and stimulant-involved deaths by weekly periods, Cook County and Illinois, 2018–2020.
      CharacteristicsOpioid involvedStimulant involved
      Jan 1, 2018–Dec 15, 2019Dec 15, 2019–Mar 20, 2020Mar 21, 2020–May 30, 2020 (Stay at home)May 31, 2020–Oct 6, 2020Oct 7, 2020–Dec 31, 20202018–2020p- valueJan 1, 2018–Dec 15, 2019Dec 15, 2019–Mar 20, 2020Mar 21, 2020–May 30, 2020 (Stay at home)May 31, 2020–Oct 6, 2020Oct 7, 2020–Dec 31, 20202018–2020p-value
      n(%)n(%)n(%)n(%)n(%)n(%)n(%)n(%)n(%)n(%)n(%)n(%)
      Sex
       Male1,728(74.8)331(72.4)329(77.1)473(77.7)282(77.1)3,143(75.4)0.1008846(74.3)157(73.0)165(79.7)230(79.6)119(73.0)1,517(75.4)0.1063
      Race/ethnicity0.53720.7871
       Black1,032(44.7)236(51.6)204(47.8)282(46.3)189(51.6)1,943(46.6)527(46.3)102(47.4)97(46.9)146(50.5)80(49.1)952(47.3)
       White1,257(54.4)214(46.8)216(50.6)319(52.4)176(48.1)2,182(52.3)598(52.6)108(50.2)107(51.7)139(48.1)83(50.9)1,035(51.4)
       Other21(0.9)7(1.5)7(1.6)8(1.3)1(0.3)44(1.1)13(1.1)5(2.3)3(1.5)4(1.4)..25(1.2)
      Age group, years0.11860.2009
       0–24138(6.0)24(5.3)31(7.3)39(6.4)16(4.4)248(6.0)35(3.1)11(5.1)6(2.9)15(5.2)3(1.8)70(3.5)
       25–34421(18.2)68(14.9)87(20.4)96(15.8)62(16.9)734(17.6)220(19.3)36(16.7)48(23.2)42(14.5)28(17.2)374(18.6)
       35–44472(20.4)88(19.3)79(18.5)129(21.2)75(20.5)843(20.2)264(23.2)54(25.1)37(17.9)66(22.8)50(30.7)471(23.4)
       45–54641(27.8)126(27.6)107(25.1)169(27.8)87(23.8)1,130(27.1)324(28.5)47(21.9)53(25.6)80(27.7)34(20.9)538(26.7)
       55–64522(22.6)116(25.4)86(20.1)136(22.3)99(27.1)959(23.0)245(21.5)53(24.7)47(22.7)72(24.9)40(24.5)457(22.7)
       ≥65113(4.9)33(7.2)37(8.7)38(6.2)27(7.4)248(6.0)47(4.1)13(6.1)16(7.7)13(4.5)8(4.9)97(4.8)
      Location of death
       Home1,209(52.3)235(51.4)252(59.0)311(51.1)197(53.8)2,204(52.9)<0.0001563(49.5)109(50.7)108(52.2)134(46.4)76(46.6)990(49.2)0.2617
       Medical663(28.7)143(31.3)75(17.6)144(23.7)94(25.7)1,119(26.8)346(30.4)65(30.2)46(22.2)74(25.6)48(29.5)579(28.8)
       Motel77(3.3)20(4.4)17(4.0)31(5.1)11(3.0)156(3.7)56(4.9)11(5.1)12(5.8)19(6.6)7(4.3)105(5.2)
       Other207(9.0)30(6.6)46(10.8)65(10.7)29(7.9)377(9.0)101(8.9)16(7.4)25(12.1)33(11.4)12(7.4)187(9.3)
       Scene154(6.7)29(6.4)37(8.7)58(9.5)35(9.6)313(7.5)72(6.3)14(6.5)16(7.7)29(10.0)20(12.3)151(7.5)
      Co-occurring drugs
       Fentanyl1,753(75.9)367(80.3)364(85.3)531(87.2)325(88.8)3,340(80.1)0.0543653(57.4)135(62.8)150(72.5)204(70.6)123(75.5)1,265(62.9)0.0339
       Heroin1,277(55.3)243(53.2)202(47.3)219(36.0)134(36.6)2,075(49.8)0.0874484(42.5)85(39.5)75(36.2)71(24.6)46(28.2)761(37.8)0.4845
       Any stimulant/opioid879(38.1)167(36.5)176(41.2)236(38.8)135(36.9)1,593(38.2)0.1460879(77.2)167(77.7)176(85.0)236(81.7)135(82.8)1,593(79.2)0.0530
       Cocaine795(34.4)148(32.4)165(38.6)213(35.0)119(32.5)1,440(34.5)0.04881,023(89.9)193(89.8)191(92.3)259(89.6)144(88.3)1,810(90.0)0.3692
       Methamphetamine49(2.1)15(3.3)6(1.4)21(3.5)12(3.3)103(2.5)0.078766(5.8)17(7.9)10(4.8)23(8.0)14(8.6)130(6.5)0.1968
       Amphetamine49(2.1)8(1.8)5(1.2)9(1.5)9(2.5)80(1.9)0.581357(5.0)9(4.2)7(3.4)13(4.5)10(6.1)96(4.8)0.8004
       Benzodiazepines481(20.8)65(14.2)69(16.2)94(15.4)46(12.6)755(18.1)0.4140207(18.2)30(14.0)26(12.6)41(14.2)15(9.2)319(15.9)0.6732
       Alcohol642(27.8)102(22.3)103(24.1)160(26.3)96(26.2)1,103(26.5)0.5133350(30.8)59(27.4)64(30.9)72(24.9)47(28.8)592(29.4)0.4322
      Total2,3104574276093664,1691,1382152072891632,012
      Note: Boldface indicates statistical significance (p<0.05).
      Stay-at-home order period was March 21, 2020–May 30, 2020. Statistical testing for differences between italicized periods corresponding to the 10-week stay-at-home period and the preceding 13-week period. The following listed drugs were used to identify opioids: opioid/opiate, heroin, methadone, hydromorphone, morphine, oxymorphone, oxycodone, hydrocodone, butorphanol, codeine, buprenorphine, meperidine, nalbuphine, tramadol, tapentadol, pentazocine, dihydrocodeine, levomethorphan, levorphanol, U-47700, U-49900, fentanyl, acetyl fentanyl, ANPP, butryl, carfentanil, cyclopropyl, FIBF, furanyl, methoxyacetyl, norfentanyl, FBF, and valerylfentanyl; the following listed drugs were used to identify stimulants: cocaine, amphetamine, methamphetamine, methylphenidate, cathinone, mitragynine, topiramate, and phentermine. There were 5 missing/unknown sex and 7 missing age group for opioids and 5 missing age group for stimulants.
      ANPP, Anilino-N-phenethyl-piperidine; Dec, December; FBF, fluorobutyryl fentanyl; FIBF, fluoroisobutyryl fentanyl; Jan, January; Mar, March; Oct, October.
      Immediately before stay-at-home orders, 51% of opioid-involved deaths occurred in residential locations, increasing to 59% (p<0.0001) during the order and decreasing back to 51% after the order expired (Figure 2). The adjusted OR (aOR) for opioid-involved fatal overdoses occurring at home during this period compared with that occurring the 13 weeks before was 1.37 (95% CI=1.05, 1.79) (Table 2). In this group, Black people were less likely to die at home than White people (aOR=0.80; CI=0.70, 0.91), as were males (aOR=0.86; 95% CI=0.75, 0.99). Medical facilities accounted for only 17.6% of opioid-involved deaths during the stay-at-home period, decreasing from 31.2% in the preorder period. For stimulant-involved deaths, 50.7% occurred at home locations immediately before the order, with a nonsignificant increase to 52.2% during the stay-at-home period. Deaths involving stimulants at medical facilities declined from 30.2% to 22.2% of the total from the preorder to the stay-at-home-order period, respectively.
      Figure 2
      Figure 2Percentage of opioid- and stimulant-involved fatal overdoses attributed to homes by period, Cook County, Illinois, 2018–2020. The stay-at-home order period was Mar 21, 2020–May 30, 2020.
      Dec, December; Jan, January; Mar, March; Oct, October.
      Table 2ORs Estimates and 95% CIs for Fatal Overdoses Occurring in the Home During the Stay-At-Home-Order Period, Cook County and Illinois, 2018–2020
      Opioid involved
      EffectOR95% Wald confidence limits
      Demographics
       Age1.0101.0051.015
       Black versus White0.7980.6970.913
       Other versus White0.8910.4811.649
       Male versus female0.8600.7460.993
      Time periods
       Prepandemic versus before1.0330.8441.265
       Stay-at-home versus before1.3711.0491.792
       After versus before0.9880.7741.262
       Remainder 2020 versus before1.1010.8351.452
      Stimulant involved
      EffectOR95% Wald confidence limits
      Demographics
       Age1.0111.0031.019
       Black versus White0.8240.6751.006
       Other versus White1.4770.6553.327
       Male versus female0.8970.7311.101
      Time periods
       Prepandemic versus before0.9520.7101.276
       Stay-at-home versus before1.0620.7241.559
       After versus before0.8520.5971.215
       Remainder 2020 versus before0.8540.5671.287
      Notes: Before refers to the 13-week period immediately before stay-at-home orders, December 15, 2019–March 20, 2020.

      DISCUSSION

      There were 427 opioid-involved deaths during the enactment of the stay-at-home order in Cook County, Illinois, with 59% of these people found at home. The odds of dying at home were 37% greater in this period than in the preorder period. The proportional increase to home locations was modest for stimulant-involved deaths and similar to location changes seen in all-cause mortality at the national level (unpublished data from CDC).

      CDC Wide-Ranging Online Data for Epidemiologic Research (WONDER). Centers for Disease Control and Prevention, National Center for Health Statistics. http://wonder.cdc.gov. Updated December 22, 2021. Accessed July 4, 2022.

      These dynamics are consistent with decreasing use of medical facilities during the early phases of the COVID-19 pandemic. Although it may now seem obvious that deaths at home could increase with mobility restrictions, we tested this hypothesis empirically and further found variability by drug type.
      The intersection of social isolation, the potential fear of seeking medical treatment during the COVID-19 pandemic, geographic context, and the intervening influence of emergency medical services during this time is complex. For example, a study of Chicago area emergency department (ED) presentations reported no increase in the number of opioid-related visits immediately after the pandemic started. Likewise, in a sample of Washington, District of Columbia/Baltimore area EDs, opioid overdoses did not increase (although other drug overdoses did).
      • Hall GT
      • Cruz DS
      • Lank PM
      • McCarthy DM
      • Kim HS.
      Opioid-related emergency department visits during COVID-19 in a large health system.
      ,
      • Chandran K
      • Mazer-Amirshahi M
      • Shankar N
      • Desale S
      • Nelson L
      • Mete M.
      Impact of COVID-19 pandemic on emergency department substance use screens and overdose presentations.
      Our findings could be consistent with those of these studies because if more people have fatal opioid overdoses at home, then transport to the ED is unnecessary. However, research from Kentucky showed that EMS transports for opioid overdoses increased by 17%, and national data also showed sharp weekly increases in opioid overdoses after the start of the pandemic.
      • Slavova S
      • Rock P
      • Bush HM
      • Quesinberry D
      • Walsh SL.
      Signal of increased opioid overdose during COVID-19 from emergency medical services data.
      ,
      • Khare A
      • Sidana A
      • Mohemmed A
      • et al.
      Acceleration of opioid-related EMS runs in the spring of 2020: the National Emergency Medical Services Information System data for 2018-2020.
      Findings that opioid-involved deaths were more likely to occur at home than stimulant-involved deaths may reflect the underlying differences in social determinants between these populations associated with their use of opioid and stimulant combinations. Specifically, in terms of housing stability, Barocas et al.
      • Barocas JA
      • Wang J
      • Marshall BDL
      • et al.
      Sociodemographic factors and social determinants associated with toxicology confirmed polysubstance opioid-related deaths.
      found that people experiencing a fatal overdose involving stimulants were more likely to be homeless than those experiencing overdoses involving opioids alone, and a national study found that respondents who used methamphetamine alone were more likely to have unstable housing than those who used opioids alone.
      • Shearer RD
      • Howell BA
      • Bart G
      • Winkelman TNA.
      Substance use patterns and health profiles among US adults who use opioids, methamphetamine, or both, 2015–2018.
      A larger proportion of patients with opioid use disorder were found to have unstable housing if they were also positive for amphetamines.
      • Chawarski MC
      • Hawk K
      • Edelman EJ
      • et al.
      Use of amphetamine-type stimulants among emergency department patients with untreated opioid use disorder.
      Likewise, around 30% of people with opioid use disorder experiencing homelessness reported stimulant abuse compared with 9% among those with stable housing.
      • Ali MM
      • Sutherland H
      • Rosenoff E.
      Comorbid health conditions and treatment utilization among individuals with opioid use disorder experiencing homelessness.
      Thus, our findings may reflect the reality that people who use stimulants were less likely to die from those drugs at home during the pandemic because they were less likely to have homes to die in. We could not effectively examine these underlying causes of housing instability or homelessness with the available open data beyond an occasional note from the medical examiner's case investigation in the location of death field. Future research will involve extracting these social determinants of health from case investigation notes using natural language processing.
      • Harris DR
      • Eisinger C
      • Wang Y
      • Delcher C.
      Challenges and barriers in applying natural language processing to medical examiner notes from fatal opioid poisoning cases.
      Previous studies from Cook County left the question as to whether opioid-involved deaths were accelerating before stay-at-home orders unanswered.
      • Mason M
      • Arukumar P
      • Feinglass J.
      The pandemic stay-at-home order and opioid-involved overdose fatalities.
      ,
      • Mason M
      • Welch SB
      • Arunkumar P
      • Post LA
      • Feinglass JM.
      Notes from the field: opioid overdose deaths before, during, and after an 11-week COVID-19 stay-at-home order – Cook County, Illinois, January 1, 2018-October 6, 2020.
      We found that opioid- and stimulant-involved deaths were indeed increasing weekly from 2018 to 2019.
      • Mason M
      • Welch SB
      • Arunkumar P
      • Post LA
      • Feinglass JM.
      Notes from the field: opioid overdose deaths before, during, and after an 11-week COVID-19 stay-at-home order – Cook County, Illinois, January 1, 2018-October 6, 2020.
      Although the COVID-19 pandemic and isolation have made the substance use epidemic worse by many metrics, a return to the intensifying prepandemic norms is not ideal. Unfortunately, Cook County data for 2021 show that annual deaths from opioids and stimulants have surpassed 2020 levels, which appears to be a national trend according to CDC's provisional data as of November 2021.

      Products - vital statistics rapid release - provisional drug overdose data. Centers for Disease Control and Prevention. https://www.cdc.gov/nchs/nvss/vsrr/drug-overdose-data.htm. Updated June 15, 2022. Accessed January 17, 2022.

      Opposing trends in the proportional contributions to opioid-involved mortality attributable to fentanyl and heroin continued unabated throughout this period. There is some compelling visual evidence (Figure 1) that abrupt changes in the heroin supply occurred immediately after the stay-at-home order. By the end of 2020, fentanyl was involved in 89% of opioid-involved overdoses, a level maintained in Cook County in 2021 (unpublished analysis from this data set). Research on fentanyl-involved deaths in Cook County found that these fatalities were concentrated in more socially deprived neighborhoods.
      • Nesoff ED
      • Branas CC
      • Martins SS.
      The geographic distribution of fentanyl-involved overdose deaths in Cook County, Illinois.
      Recent declines in the apparent prevalence of heroin are consistent with the findings of several studies.
      • Mason M
      • Arukumar P
      • Feinglass J.
      The pandemic stay-at-home order and opioid-involved overdose fatalities.
      ,
      • Twillman RK
      • Dawson E
      • LaRue L
      • Guevara MG
      • Whitley P
      • Huskey A
      Evaluation of trends of near-real-time urine drug test results for methamphetamine, cocaine, heroin, and fentanyl.
      • Daniulaityte R
      • Juhascik MP
      • Strayer KE
      • et al.
      Trends in fentanyl and fentanyl analogue-related overdose deaths - Montgomery County, Ohio, 2015–2017.
      • Palamar JJ
      • Le A
      • Carr TH
      • Cottler LB.
      Shifts in drug seizures in the United States during the COVID-19 pandemic.
      Our study relied on medical examiner investigation notes, which are valuable for rapidly identifying the location of overdose fatalities and showed sensitivity to large-scale shifts in population mobility. This information can help to target specific places for intervention. For example, we found that approximately 156 (3.7%) and 105 (5.2%) of opioid- and stimulant-involved deaths, respectively, occurred at motels or hotels, with 23 of these location addresses detected more than once. These locations could be local hot spots suitable for naloxone kit placement. The Cook County data have both street addresses for incident and death locations; further research comparing these locations is possible. Linking addresses to land-use files in the absence of medical examiner‒described locations may be a viable alternative to monitoring place-based risk factors.

      Limitations

      Our weekly counts may differ slightly from previous reports owing to either increased sensitivity in our newer methods to detect drugs from text written in the cause of death field, the continuously updated open data file, and/or our counts derived from regression models. Overall, the proportion of opioid- and stimulant-involved deaths occurring at home in Cook County was lower than that reported for 25 states from January 2019 to June 2019 (79.9% for any home setting).
      • O'Donnell J
      • Gladden RM
      • Mattson CL
      • Hunter CT
      • Davis NL.
      Vital signs: characteristics of drug overdose deaths involving opioids and stimulants - 24 states and the District of Columbia, January-June 2019.
      Our results remained unchanged when we restricted the data to this time frame. This difference highlights the fact that CDC, states, and other jurisdictions may document these locations differently. Our categorization of the location of death is subject to misclassification owing to judgments made about nonspecific locations (e.g., bathroom). Our categories for opioid- and stimulant-involved deaths were not mutually exclusive.
      Future localized studies could address other changes occurring during this period that would impact drug use behavior in the home, such as illegal offline and online drug market restructuring including the increased use of alcohol delivery services.
      • Barratt MJ
      • Aldridge J.
      No magic pocket: buying and selling on drug cryptomarkets in response to the COVID-19 pandemic and social restrictions.
      ,
      • Castaldelli-Maia JM
      • Segura LE
      • Martins SS.
      The concerning increasing trend of alcohol beverage sales in the U.S. during the COVID-19 pandemic.

      CONCLUSIONS

      The stay-at-home order time period in Cook County increased the likelihood that a fatal overdose would occur at home for people dying from opioids compared with that for those dying from stimulants. This work has implications for detecting, mitigating, and responding to overdose risks during times of social isolation and drug use behaviors occurring in the home.

      ACKNOWLEDGMENTS

      The authors wish to thank the Cook County Government of Illinois and the Office of the Medical Examiner for opening their data to the public and Jungjun Bae for data visualization.
      The contents are those of the author(s) and do not necessarily represent the official views of nor an endorsement by the Centers for Disease Control and Prevention/HHS or the U.S. Government.
      This project is supported by the Centers for Disease Control and Prevention of HHS as part of Grant 1R01CE003360-01-00 .
      No financial disclosures were reported by the authors of this paper.

      CRediT AUTHOR STATEMENT

      Chris Delcher: Conceptualization, Formal analysis, Writing - original draft. Daniel R. Harris: Writing - review and editing. Nicholas Anthony: Data curation, Writing - review and editing. Mojde Mir: Data curation, Funding acquisition, Writing - review and editing.

      Appendix. Supplementary materials

      REFERENCES

        • Moreland A
        • Herlihy C
        • Tynan MA
        • et al.
        Timing of state and territorial COVID-19 stay-at-home orders and changes in population movement - United States, March 1-May 31, 2020.
        MMWR Morb Mortal Wkly Rep. 2020; 69: 1198-1203https://doi.org/10.15585/mmwr.mm6935a2
        • Murray T.
        Stay-at-home orders, mobility patterns, and spread of COVID-19.
        Am J Public Health. 2021; 111: 1149-1156https://doi.org/10.2105/AJPH.2021.306209
        • Nouvellet P
        • Bhatia S
        • Cori A
        • et al.
        Reduction in mobility and COVID-19 transmission.
        Nat Commun. 2021; 12: 1090https://doi.org/10.1038/s41467-021-21358-2
        • Czeisler MÉ
        • Lane RI
        • Petrosky E
        • et al.
        Mental health, substance use, and suicidal ideation during the COVID-19 pandemic - United States, June 24-30, 2020.
        MMWR Morb Mortal Wkly Rep. 2020; 69: 1049-1057https://doi.org/10.15585/mmwr.mm6932a1
        • Mason M
        • Arukumar P
        • Feinglass J.
        The pandemic stay-at-home order and opioid-involved overdose fatalities.
        JAMA. 2021; 325: 2495-2496https://doi.org/10.1001/jama.2021.6700
      1. CDC Wide-Ranging Online Data for Epidemiologic Research (WONDER). Centers for Disease Control and Prevention, National Center for Health Statistics. http://wonder.cdc.gov. Updated December 22, 2021. Accessed July 4, 2022.

        • Marks C
        • Abramovitz D
        • Donnelly CA
        • et al.
        Identifying counties at risk of high overdose mortality burden during the emerging fentanyl epidemic in the USA: a predictive statistical modelling study.
        Lancet Public Health. 2021; 6: e720-e728https://doi.org/10.1016/S2468-2667(21)00080-3
      2. ODPredict explorer. https://charles-marks.shinyapps.io/OD_predict_dash/. Accessed May 1, 2022.

        • Unwin H
        • Mishra S
        • Bradley V
        • et al.
        Report 23: State-Level Tracking of COVID-19 in the United States.
        United Kingdom: Imperial College, London2020https://doi.org/10.25561/79231
        • Mason M
        • Welch SB
        • Arunkumar P
        • Post LA
        • Feinglass JM.
        Notes from the field: opioid overdose deaths before, during, and after an 11-week COVID-19 stay-at-home order – Cook County, Illinois, January 1, 2018-October 6, 2020.
        MMWR Morb Mortal Wkly Rep. 2021; 70: 362-363https://doi.org/10.15585/mmwr.mm7010a3
        • Nesoff ED
        • Branas CC
        • Martins SS.
        The geographic distribution of fentanyl-involved overdose deaths in Cook County, Illinois.
        Am J Public Health. 2020; 110: 98-105https://doi.org/10.2105/AJPH.2019.305368
      3. Medical examiner case archive. Cook County Open Data.https://datacatalog.cookcountyil.gov/Public-Safety/Medical-Examiner-Case-Archive/cjeq-bs86. Updated March 6, 2022. Accessed March 6, 2022.

      4. Land use inventories – CMAP. Data Hub. https://datahub.cmap.illinois.gov/group/land-use-inventories. Updated November 23, 2020. Accessed April 26, 2021.

      5. University of Kentucky, Institute for Pharmaceutical Outcomes and Policy (UK-IPOP): drug extraction. GitHub. https://github.com/UK-IPOP/drug-extraction. Updated June 13, 2022. Accessed July 4, 2022.

      6. SAS software version 9.4 for Windows. SAS Institute Inc., Cary, NC; 2002.

        • Hall GT
        • Cruz DS
        • Lank PM
        • McCarthy DM
        • Kim HS.
        Opioid-related emergency department visits during COVID-19 in a large health system.
        J Addict Med. 2021; 15: 345-348https://doi.org/10.1097/ADM.0000000000000850
        • Chandran K
        • Mazer-Amirshahi M
        • Shankar N
        • Desale S
        • Nelson L
        • Mete M.
        Impact of COVID-19 pandemic on emergency department substance use screens and overdose presentations.
        Am J Emerg Med. 2021; 50: 472-476https://doi.org/10.1016/j.ajem.2021.08.058
        • Slavova S
        • Rock P
        • Bush HM
        • Quesinberry D
        • Walsh SL.
        Signal of increased opioid overdose during COVID-19 from emergency medical services data.
        Drug Alcohol Depend. 2020; 214108176https://doi.org/10.1016/j.drugalcdep.2020.108176
        • Khare A
        • Sidana A
        • Mohemmed A
        • et al.
        Acceleration of opioid-related EMS runs in the spring of 2020: the National Emergency Medical Services Information System data for 2018-2020.
        Drug Alcohol Depend. 2022; 232109271https://doi.org/10.1016/j.drugalcdep.2022.109271
        • Barocas JA
        • Wang J
        • Marshall BDL
        • et al.
        Sociodemographic factors and social determinants associated with toxicology confirmed polysubstance opioid-related deaths.
        Drug Alcohol Depend. 2019; 200: 59-63https://doi.org/10.1016/j.drugalcdep.2019.03.014
        • Shearer RD
        • Howell BA
        • Bart G
        • Winkelman TNA.
        Substance use patterns and health profiles among US adults who use opioids, methamphetamine, or both, 2015–2018.
        Drug Alcohol Depend. 2020; 214108162https://doi.org/10.1016/j.drugalcdep.2020.108162
        • Chawarski MC
        • Hawk K
        • Edelman EJ
        • et al.
        Use of amphetamine-type stimulants among emergency department patients with untreated opioid use disorder.
        Ann Emerg Med. 2020; 76: 782-787https://doi.org/10.1016/j.annemergmed.2020.06.046
        • Ali MM
        • Sutherland H
        • Rosenoff E.
        Comorbid health conditions and treatment utilization among individuals with opioid use disorder experiencing homelessness.
        Subst Use Misuse. 2021; 56: 571-574https://doi.org/10.1080/10826084.2021.1884723
        • Harris DR
        • Eisinger C
        • Wang Y
        • Delcher C.
        Challenges and barriers in applying natural language processing to medical examiner notes from fatal opioid poisoning cases.
        Proc IEEE Int Conf Big Data. 2020; 2020: 3727-3736https://doi.org/10.1109/BigData50022.2020.9378443
      7. Products - vital statistics rapid release - provisional drug overdose data. Centers for Disease Control and Prevention. https://www.cdc.gov/nchs/nvss/vsrr/drug-overdose-data.htm. Updated June 15, 2022. Accessed January 17, 2022.

        • Twillman RK
        • Dawson E
        • LaRue L
        • Guevara MG
        • Whitley P
        • Huskey A
        Evaluation of trends of near-real-time urine drug test results for methamphetamine, cocaine, heroin, and fentanyl.
        JAMA Netw Open. 2020; 3e1918514https://doi.org/10.1001/jamanetworkopen.2019.18514
        • Daniulaityte R
        • Juhascik MP
        • Strayer KE
        • et al.
        Trends in fentanyl and fentanyl analogue-related overdose deaths - Montgomery County, Ohio, 2015–2017.
        Drug Alcohol Depend. 2019; 198: 116-120https://doi.org/10.1016/j.drugalcdep.2019.01.045
        • Palamar JJ
        • Le A
        • Carr TH
        • Cottler LB.
        Shifts in drug seizures in the United States during the COVID-19 pandemic.
        Drug Alcohol Depend. 2021; 221108580https://doi.org/10.1016/j.drugalcdep.2021.108580
        • O'Donnell J
        • Gladden RM
        • Mattson CL
        • Hunter CT
        • Davis NL.
        Vital signs: characteristics of drug overdose deaths involving opioids and stimulants - 24 states and the District of Columbia, January-June 2019.
        MMWR Morb Mortal Wkly Rep. 2020; 69: 1189-1197https://doi.org/10.15585/mmwr.mm6935a1
        • Barratt MJ
        • Aldridge J.
        No magic pocket: buying and selling on drug cryptomarkets in response to the COVID-19 pandemic and social restrictions.
        Int J Drug Policy. 2020; 83102894https://doi.org/10.1016/j.drugpo.2020.102894
        • Castaldelli-Maia JM
        • Segura LE
        • Martins SS.
        The concerning increasing trend of alcohol beverage sales in the U.S. during the COVID-19 pandemic.
        Alcohol. 2021; 96: 37-42https://doi.org/10.1016/j.alcohol.2021.06.004