Abstract
Introduction:
Following a nonfatal opioid overdose, patients are at high risk for repeat overdose. The objective of this study was to examine the association of MOUD after nonfatal opioid overdose with risk of repeat overdose in the following year.
Methods:
This retrospective cohort study analyzed Missouri Medicaid claims from July 2012 to December 2021. The study identified opioid overdoses occurring between 2013 and 2020 using diagnosis codes for opioid poisoning in an inpatient or emergency department setting. The study implemented Cox models with a time-varying covariate for post-overdose receipt of MOUD.
Results:
During the study period, MOUD receipt after overdose more than tripled, from 5.1% to 17%. Overall, only 10.7% of patients received MOUD in the year after index. MOUD during follow-up was associated with significantly lower risk of repeat overdose (HR = 0.24, 95% CI = 0.08–0.75). Out of 2,962 individuals meeting inclusion criteria, 12.1% had a repeat opioid overdose within 1 year. Repeat overdose risk was more than doubled for those whose index overdose involved heroin or synthetic opioids (HR = 2.1, 95% CI = 1.62–2.72), but MOUD was associated with significantly reduced risk in this group (HR = 0.21, 95% CI = 0.05–0.84).
Conclusions:
MOUD receipt was associated with reduced risk of repeat overdose. Those whose index overdoses involved heroin or synthetic opioids were at greater risk of repeat overdose, but MOUD substantially mitigated risk in this group.
Keywords: Opioid overdose, Opioid use disorder, Medication for opioid use disorder, Medicaid
1. Introduction
Opioid overdoses are a growing cause of morbidity and mortality in the US, and individuals with a history of nonfatal overdose are vulnerable to a range of adverse outcomes, including repeated overdose and death (Hedegaard et al., 2021; Karmali et al., 2020; Olfson et al., 2018; Weiner et al., 2020). Repeat overdose is of particular concern following nonfatal overdose, with estimates showing as many as 1 in 5 nonfatal overdoses leading to repeat overdose within the following year (Crystal et al., 2022). While evidence also shows that medication for opioid use disorder (MOUD) is highly effective for reducing risk of opioid-related morbidity and mortality, MOUD prescribing rates remain low including for patients with recent overdose events (Fudala et al., 2003; Johnson et al., 1995; Karmali et al., 2020; Larochelle et al., 2018; Ling et al., 1998, 2010; Santo et al., 2021). In this context, greater understanding is needed about populations with elevated risk of adverse opioid-related outcomes following a nonfatal overdose, as well as which populations may be under-treated.
Medicaid is the largest payer for mental health treatment in the United States, including substance use disorder (SUD) treatment, and covers a disproportionate share of adults with opioid use disorder (OUD). Although past research has investigated the risk of adverse events and the protective effect of MOUD after nonfatal overdose (Karmali et al., 2020; Larochelle et al., 2018), limited data exist from especially vulnerable populations such as Medicaid beneficiaries. A previous analysis of nonfatal medically treated overdoses in New Jersey Medicaid found that MOUD was protective against repeat overdose, and that those who had a nonfatal overdose involving heroin or synthetic opioids were at greater risk of repeat overdose than those whose nonfatal overdose involved prescription/other opioids only (Crystal et al., 2022). These results contribute to understanding about the risk of overdose in vulnerable populations and how that risk can be mitigated, however, it is unclear whether previous findings in Medicaid generalize across regions of the US.
Evidence across geographic areas is particularly important given variation in fentanyl penetration and policy environments between states (Shover et al., 2020). The Missouri Medicaid population used to investigate repeat overdose risk in this study differs from prior study samples in two key ways. First, Missouri’s illicit opioid context is unlike states addressed in previous research, with penetration of synthetic opioids lagging behind many other states, especially in the Eastern US (Mattson, 2021). Second, Missouri’s policy environment differs from previous research; for example, Missouri did not expand Medicaid eligibility until 2021. The Missouri data analyzed in this paper thus reflect a consistently lower-income population (Department of Health and Human Services Press Office, 2021). These differences in the opioid supply and policy environment provide an opportunity to assess the generalizability of prior findings: namely, whether MOUD is associated with reduced risk of repeat overdose, including for those whose index overdoses involve heroin or synthetic opioids who may have greater risk of subsequent opioid overdose.
The objective of this study was to examine the association of MOUD after nonfatal opioid overdose with risk of repeat overdose in the following year among Missouri Medicaid beneficiaries overall and stratified by the type of opioid involved in the index overdose. We assessed time-varying receipt of MOUD after nonfatal opioid overdose events because patients may start and stop MOUD, resulting in multiple MOUD treatment episodes (Krawcyk et al., 2020). We also investigated trends in the prescribing of MOUD following nonfatal overdose. To verify the robustness of results across provider types, variation in follow-up, and time periods, we performed multiple sensitivity analyses. To facilitate evaluation of these results in the context of previous research, the study performed direct statistical comparisons with related published findings and we present them at the end of the Results. We hypothesized that MOUD receipt after a nonfatal overdose would be associated with lower risk of repeat overdose, including for those whose index overdose involved heroin or synthetic opioids who were expected to have greater risk of repeat overdose than those whose index overdose involved prescription/other opioids only.
2. Methods
2.1. Data sources
This analysis used Missouri Medicaid claims from July 2012-December 2021 to identify enrollees with index opioid-involved nonfatal overdose events occurring in 2013–2020. This dataset included demographics, outpatient prescription drug claims and inpatient or outpatient medical claims other than nursing home, home health, and dental services. Reporting follows the STROBE guidelines for observational results in epidemiology (Vandenbroucke et al., 2007).
2.2. Sample selection
The sample included Medicaid beneficiaries aged 12–64 years with a nonfatal opioid overdose treated in an inpatient or emergency department setting. We required at least 6 months of continuous enrollment before the index overdose to identify baseline comorbid conditions and healthcare exposures. To allow sufficient time to evaluate the effects of post-overdose MOUD treatment on risk of repeat overdose, we excluded individuals without at least 12 months of enrollment after the index overdose. However, to understand whether results were dependent on enrollment stability, we conducted a sensitivity analysis without this requirement, described in subsection 2.6 below. For all analyses, the study required index events to occur in the years 2013–2020 to ensure that all participants could be observed for 6 months before index and one year after index. The study excluded dual Medicare-Medicaid enrollees to ensure that the study captured all follow-up treatment, as Medicare is the first payer for services. The study excluded subjects with MOUD in the 180 days before the index overdose to ensure that those receiving MOUD in the post-index period were treatment-naive. Sample size and selection criteria are shown in Figure 1.
Figure 1:
Cohort flow diagram
2.3. Exposure and outcome measures
Study staff identified overdose events using ICD-9 and ICD-10 diagnosis codes for opioid poisoning on emergency department and inpatient claims. Opioid poisoning diagnosis codes have been shown to have a high sensitivity, specificity, and positive predictive value for identifying opioid overdose (Green et al., 2017, 2019; Slavova et al., 2020; Vivolo-Kantor et al., 2021). For each person, the study defined the first qualifying opioid overdose in the study period as the index overdose. Using diagnosis codes on the index overdose claim, we identified overdoses involving heroin or synthetic opioids other than methadone (e.g., fentanyl). We classified overdoses without these codes as involving prescription/other opioids only. Due to small numbers of heroin and synthetic opioid overdoses, we grouped these two types of overdoses together to maximize the precision of estimates for this group. We also identified whether the index opioid overdose involved benzodiazepines or stimulants. The study examined co-involvement of opioids and stimulants because past data have indicated that stimulants are commonly present in opioid-involved overdose toxicology in Missouri (Missouri overdose information, 2022). Other variables identified at the time of the index overdose included treatment in an emergency department, inpatient hospitalization, and filled prescriptions for benzodiazepines and opioids. After exclusions, the study observed no missing values for variables of interest. All diagnosis and procedure codes are provided in the Appendix. The primary study outcome was the number of days to the first nonfatal repeat overdose, and the study censored patients at 365 days if it did not observe a repeat overdose by that time. The outcome for the primary analysis included only non-fatal repeat overdoses to allow for direct comparability with prior research. The competing risks sensitivity analysis included both fatal and non-fatal overdoses as outcomes. In addition to the primary cohort analysis, we calculated some supplementary values for reference. The study determined overall rates of MOUD prescribing in Missouri Medicaid over time using counts of total enrollees and total MOUD prescriptions in claims data. To estimate the mix of illicit opioid types in Missouri, we calculated percentages based on drug seizure data (see Appendix) (National Forensic Laboratory Information System, 2022).
2.4. Explanatory variables
To estimate the effect of MOUD on risk of repeat overdose, we defined a time-varying covariate for MOUD in the follow-up period. This definition accounts for commonly observed treatment gaps (Saloner et al., 2017) to estimate the effect of MOUD during treated days only. The study identified MOUD using fill dates and days of supply from prescription claims for buprenorphine excluding formulations used for pain (Butrans, Belbuca, and Buprenex) or naltrexone, as well as procedure codes for methadone, injectable buprenorphine, or naltrexone administration in a healthcare setting (see Appendix). When the study observed procedure codes for injectable buprenorphine or naltrexone administration, we defined time-varying receipt using dates of service. We defined methadone to include a 2-day exposure after the last service date to allow for possible take-home doses.
2.5. Baseline covariates
Mental health comorbidities identified in the 180-day baseline period included schizophrenia, bipolar disorder, major depressive disorder, anxiety disorders, and personality disorders, as well as substance use disorder diagnoses for alcohol, benzodiazepines, stimulants, cannabis, and opioids. Medical comorbidities identified in the baseline period included asthma, cerebrovascular disease, chronic pain, chronic obstructive pulmonary disease (COPD), diabetes, heart failure, hepatitis C, HIV, hypertension, pneumonia, and sleep apnea.
Other variables identified in the baseline period included prescriptions for benzodiazepines and opioids, emergency department visits, and psychosocial service visits. For benzodiazepine and opioid prescriptions, we also used the date filled and days of supply on the claim to determine drug availability at the index overdose date. Models included a variable indicating the quarter in which the index overdose took place (Q1 2013 = 1, Q2 2013 = 2, …, Q4 2020 = 32) as well as a binary variable for whether the index overdose took place during the COVID-19 pandemic, defined as an index overdose during or after March 2020.
2.6. Statistical analysis
The study used chi-square tests to determine bivariate associations of baseline and index overdose characteristics with repeat overdose (Table 1). We also used chi-square tests to compare sample proportions, including the proportion of patients who had a repeat overdose, to previously reported data from a New Jersey sample (Crystal et al., 2022). These chi-square tests were based on comparisons to the published results and did not involve any use of the original data sets used by Crystal, et al. Cochran-Armitage trend tests were used to evaluate the significance of time trends. The study used extended Cox models to evaluate the association of time-varying MOUD receipt during the follow-up period with risk of repeat overdose, adjusting for all baseline and index overdose characteristics (Table 2). For time-invariant covariates, the study evaluated proportional hazards assumption by visual inspection of Schoenfeld residual plots from the primary model. Patients with no repeat overdose in the main analysis were censored at the end of the follow-up period (one year after the index overdose).
Table 1:
Beneficiary characteristics and chi-square tests of association with repeat nonfatal overdose in the 12 months after index
| Characteristics | Beneficiaries with Index Overdose† N (column %) | Repeat Overdose N (column %) | No Repeat Overdose N (column %) | Chi-square p-value†† |
|---|---|---|---|---|
|
| ||||
| Overall | 2,962 (100) | 358 (100) | 2,604 (100) | |
| Index Overdose Characteristics | ||||
| Involved Heroin/Synthetic Opioids | 891 (30.1) | 172 (48) | 719 (27.6) | <0.001 |
| Involved Benzodiazepines | 257 (8.7) | 18 (5) | 239 (9.2) | 0.009 |
| Involved Stimulants | 80 (2.7) | 6 (1.7) | 74 (2.8) | 0.202 |
| Resulted in Inpatient Hospitalization | 1013 (34.2) | 100 (27.9) | 913 (35.1) | 0.008 |
| Year of Index Overdose | 0.001 | |||
| 2013 | 295 (10) | 33 (9.2) | 262 (10.1) | |
| 2014 | 324 (10.9) | 17 (4.7) | 307 (11.8) | |
| 2015 | 387 (13.1) | 37 (10.3) | 350 (13.4) | |
| 2016 | 549 (18.5) | 71 (19.8) | 478 (18.4) | |
| 2017 | 463 (15.6) | 69 (19.3) | 394 (15.1) | |
| 2018 | 335 (11.3) | 50 (14) | 285 (10.9) | |
| 2019 | 261 (8.8) | 33 (9.2) | 228 (8.8) | |
| 2020 | 348 (11.7) | 48 (13.4) | 300 (11.5) | |
| Age | 0.129 | |||
| 12–24 | 490 (16.5) | 45 (12.6) | 445 (17.1) | |
| 25–39 | 1021 (34.5) | 134 (37.4) | 887 (34.1) | |
| 40–55 | 939 (31.7) | 111 (31) | 828 (31.8) | |
| 56–64 | 512 (17.3) | 68 (19) | 444 (17.1) | |
| Racial/Ethnic Group | 0.217 | |||
| White | 2107 (71.1) | 242 (67.6) | 1865 (71.6) | |
| Black | 765 (25.8) | 106 (29.6) | 659 (25.3) | |
| Other/Unknown | 90 (3) | 10 (2.8) | 80 (3.1) | |
| Gender | 0.001 | |||
| Female | 1799 (60.7) | 189 (52.8) | 1610 (61.8) | |
| Male | 1163 (39.3) | 169 (47.2) | 994 (38.2) | |
| SUD Comorbidities | ||||
| Alcohol Use Disorder | 470 (15.9) | 58 (16.2) | 412 (15.8) | 0.854 |
| Benzodiazepine Use Disorder | 105 (3.5) | 15 (4.2) | 90 (3.5) | 0.482 |
| Cannabis Use Disorder | 326 (11) | 51 (14.2) | 275 (10.6) | 0.037 |
| Opioid Use Disorder | 900 (30.4) | 159 (44.4) | 741 (28.5) | <0.001 |
| Stimulant Use Disorder | 427 (14.4) | 62 (17.3) | 365 (14) | 0.095 |
| Psychiatric Comorbidities | ||||
| Schizophrenia | 386 (13) | 53 (14.8) | 333 (12.8) | 0.288 |
| Bipolar Disorder | 690 (23.3) | 81 (22.6) | 609 (23.4) | 0.749 |
| Major Depressive Disorder | 1145 (38.7) | 155 (43.3) | 990 (38) | 0.055 |
| Anxiety Disorder | 1384 (46.7) | 153 (42.7) | 1231 (47.3) | 0.107 |
| Personality Disorder | 219 (7.4) | 30 (8.4) | 189 (7.3) | 0.447 |
| Medical Comorbidities | ||||
| Asthma | 477 (16.1) | 57 (15.9) | 420 (16.1) | 0.92 |
| Cerebrovascular Disease | 218 (7.4) | 32 (8.9) | 186 (7.1) | 0.223 |
| Chronic Pain | 2099 (70.9) | 245 (68.4) | 1854 (71.2) | 0.281 |
| Chronic obstructive pulmonary disease (COPD) | 446 (15.1) | 67 (18.7) | 379 (14.6) | 0.039 |
| Diabetes | 610 (20.6) | 74 (20.7) | 536 (20.6) | 0.97 |
| Heart Failure | 212 (7.2) | 31 (8.7) | 181 (7) | 0.24 |
| Hepatitis C | 334 (11.3) | 58 (16.2) | 276 (10.6) | 0.002 |
| HIV | 40 (1.4) | 8 (2.2) | 32 (1.2) | 0.122 |
| Hypertension | 1206 (40.7) | 159 (44.4) | 1047 (40.2) | 0.129 |
| Pneumonia | 292 (9.9) | 41 (11.5) | 251 (9.6) | 0.281 |
| Sleep Apnea | 191 (6.4) | 18 (5) | 173 (6.6) | 0.243 |
| Prescriptions Before Overdose | ||||
| Benzodiazepines (BZ) | 0.131 | |||
| No BZ Before Index | 1648 (55.6) | 205 (57.3) | 1443 (55.4) | |
| BZ in 6 Months Before Index | 474 (16) | 66 (18.4) | 408 (15.7) | |
| BZ at Time of Index | 840 (28.4) | 87 (24.3) | 753 (28.9) | |
| Prescription Opioids | 0.074 | |||
| No Opioids Before Index | 1080 (36.5) | 148 (41.3) | 932 (35.8) | |
| Opioids in 6 Months Before Index | 779 (26.3) | 94 (26.3) | 685 (26.3) | |
| Opioids at Time of Index | 1103 (37.2) | 116 (32.4) | 987 (37.9) | |
| Health Service Utilization | ||||
| ED Visit in Prior 6 Months | 2017 (68.1) | 244 (68.2) | 1773 (68.1) | 0.979 |
| Any Psychosocial Service in Prior 6 Months | 976 (33) | 124 (34.6) | 852 (32.7) | 0.469 |
Note: Unless otherwise noted, all covariates measured in the 6 months before the index overdose. BZ = Benzodiazepines. SUD = Substance Use Disorder.
Includes MO Medicaid enrollees meeting the study inclusion criteria.
Chi-square tests were used in bivariate tests of association with repeat overdose.
Table 2:
Multivariable Cox model results - Hazard of repeat nonfatal overdose, stratified by drugs involved in index overdose
| Characteristics | Full Sample, no MOUD Before Index Overdose (n = 2,962) | Index Overdose Involved Heroin and/or Synthetic Opioids (n = 891) | Index Overdose Involved Prescription/Other Opioids Only (n = 2,071) |
|---|---|---|---|
|
| |||
| HR (95% CI)* | HR (95% CI) | HR (95% CI) | |
| MOUD after overdose | 0.24 (0.08, 0.75)* | 0.21 (0.05, 0.84)* | 0.3 (0.04, 2.13) |
| Index Overdose Characteristics | |||
| Involved Heroin/Synthetic Opioids† | 2.1 (1.62, 2.72)* | - | - |
| Involved Benzodiazepines (BZ) | 0.77 (0.47, 1.26) | 0.23 (0.03, 1.65) | 0.89 (0.53, 1.5) |
| Involved stimulants | 0.59 (0.26, 1.35) | 0.49 (0.15, 1.57) | 0.9 (0.28, 2.9) |
| Resulted in Inpatient Hospitalization | 0.86 (0.66, 1.11) | 0.77 (0.48, 1.23) | 0.97 (0.71, 1.34) |
| Quarter of Index Overdose (1 = Q1 2013) | 1.02 (1, 1.04)* | 1.02 (1, 1.05) | 1.01 (0.99, 1.04) |
| Age | |||
| 12–24 | 0.82 (0.53, 1.26) | 0.9 (0.45, 1.8) | 0.78 (0.43, 1.42) |
| 25–39 | 1 (0.71, 1.4) | 1.08 (0.62, 1.86) | 0.93 (0.59, 1.47) |
| 40–55 | 0.91 (0.66, 1.24) | 0.78 (0.45, 1.33) | 1.01 (0.68, 1.5) |
| 56–64 | REF | REF | REF |
| Racial/Ethnic Group | |||
| White | REF | REF | REF |
| Black | 0.9 (0.7, 1.16) | 1.01 (0.7, 1.46) | 0.85 (0.57, 1.26) |
| Other/Unknown | 0.97 (0.51, 1.85) | 0.47 (0.11, 2.04) | 1.3 (0.63, 2.68) |
| Gender | |||
| Female | REF | REF | REF |
| Male | 1.24 (0.99, 1.55) | 1.43 (1.01, 2.01)* | 1.11 (0.82, 1.52) |
| SUD Comorbidities †† | |||
| Alcohol Use Disorder | 0.82 (0.6, 1.12) | 0.74 (0.47, 1.17) | 0.92 (0.6, 1.41) |
| Benzodiazepine Use Disorder | 0.98 (0.57, 1.68) | 1 (0.47, 2.12) | 0.97 (0.43, 2.19) |
| Cannabis Use Disorder | 1.23 (0.89, 1.7) | 1.15 (0.72, 1.83) | 1.18 (0.73, 1.9) |
| Opioid Use Disorder | 1.5 (1.18, 1.89)* | 1.1 (0.79, 1.53) | 1.94 (1.4, 2.68)* |
| Stimulant Use Disorder | 0.8 (0.59, 1.1) | 0.71 (0.47, 1.07) | 1.13 (0.69, 1.84) |
| Psychiatric Comorbidities | |||
| Bipolar Disorder | 0.98 (0.73, 1.3) | 1.47 (0.98, 2.2) | 0.66 (0.44, 1)* |
| Major Depressive Disorder | 1.42 (1.11, 1.82)* | 1.41 (0.95, 2.08) | 1.5 (1.08, 2.09)* |
| Anxiety Disorder | 0.79 (0.61, 1.03) | 1.1 (0.74, 1.65) | 0.65 (0.46, 0.92)* |
| Personality Disorder | 1.28 (0.85, 1.91) | 1.38 (0.69, 2.76) | 1.33 (0.8, 2.22) |
| Schizophrenia | 1.01 (0.72, 1.4) | 0.77 (0.46, 1.29) | 1.23 (0.78, 1.93) |
| Medical Comorbidities | |||
| Chronic Pain | 0.99 (0.75, 1.31) | 1.18 (0.83, 1.68) | 0.72 (0.46, 1.13) |
| Heart Failure | 1.13 (0.76, 1.7) | 0.89 (0.43, 1.85) | 1.32 (0.8, 2.16) |
| Hepatitis C | 1.23 (0.92, 1.66) | 1.17 (0.78, 1.77) | 1.24 (0.8, 1.91) |
| HIV | 1.12 (0.55, 2.32) | 1.26 (0.53, 2.97) | 1.03 (0.25, 4.27) |
| Hypertension | 1.24 (0.96, 1.61) | 1.53 (1.04, 2.24)* | 1 (0.71, 1.41) |
| Pneumonia | 1.29 (0.91, 1.83) | 1.89 (1, 3.56)* | 1.18 (0.77, 1.82) |
| Prescriptions Before Overdose | |||
| Benzodiazepines (BZ) | |||
| BZ in 6 Months Before Index | 1.11 (0.82, 1.5) | 0.6 (0.35, 1.01) | 1.56 (1.05, 2.33)* |
| BZ at Time of Index | 0.95 (0.71, 1.28) | 0.8 (0.49, 1.29) | 1.12 (0.75, 1.66) |
| Prescription Opioids | |||
| Opioids in 6 Months Before Index | 1 (0.74, 1.34) | 1.19 (0.82, 1.73) | 0.85 (0.52, 1.38) |
| Opioids at Time of Index | 1.04 (0.75, 1.44) | 0.4 (0.2, 0.82)* | 1.39 (0.89, 2.17) |
| Health Service Utilization | |||
| ED Visit in Prior 6 Months | 0.96 (0.74, 1.23) | 0.94 (0.66, 1.35) | 1 (0.7, 1.42) |
| Any psychosocial service in 6-month baseline | 1 (0.77, 1.29) | 0.91 (0.62, 1.34) | 1.08 (0.76, 1.55) |
| Index During Pandemic (Index >= March 2020) | 0.88 (0.59, 1.31) | 0.77 (0.42, 1.4) | 1 (0.58, 1.75) |
Note: CI = confidence interval; SUD = substance use disorder; BZ = Benzodiazepines; SUD = Substance Use Disorder.
Statistically significant at p < 0.05.
Includes opioid poisonings with any diagnosis code for heroin or synthetic opioid poisoning, even if codes for methadone or other natural/semi-synthetic opioid poisoning were also present.
Reference for each comorbid condition is the absence of the condition.
To verify the robustness of results, the study performed multiple sensitivity analyses. First, we estimated Cox models separately on subgroups with and without heroin or synthetic opioid-involved index overdoses. Second, Cox models included additional fixed effects for the healthcare organization that treated the index overdose to control for variation between hospitals in treatment or referral behavior that could affect the risk of repeated overdose. Third, as described above, we included individuals with less than a year of follow-up enrollment in analyses, with competing risks for death as identified from the claims demographics file and disenrollment before a year of follow-up, to evaluate the generalizability of results to those with less than a year of stable enrollment after the index nonfatal overdose. Last, the study restricted models to those with index overdoses observed prior to the COVID-19 pandemic (before March 2020) to assess whether pandemic-related changes were primary drivers of the main findings. The study used SAS Enterprise Guide 8.3 for all analyses.
3. Results
3.1. Sample characteristics
The sample with index nonfatal opioid overdose events (n=2,962) was 25.8% Black race, 71.1% White race, and 3% other or unknown race (Table 1). Data on Latinx ethnicity was not available in the Missouri claims. The sample was mostly female (60.7%) and the majority of patients (66.2%) were between ages 25 and 55. Nearly one-third of index overdoses (n=891, 30.1%) involved heroin or synthetic opioids. 358 people (12.1%) had a repeat overdose, including 172 (19.3%) of those whose index overdoses involved heroin or synthetic opioids, and 186 (9%) of those whose index overdoses involved prescription/other opioids only. Since a year of follow-up was required for all participants, the study observed a total of 2,962 person-years of follow-up. For those with a repeat overdose, the median time to repeat overdose was 83 days.
3.2. MOUD receipt during follow-up
About 1 in 10 patients (10.7%) received MOUD at any point in the year after nonfatal overdose, including 21.3% of those whose index overdoses involved heroin or synthetic opioids, and 6.1% of those whose index overdoses involved prescription or other opioids only. MOUD receipt was significantly more common among those with a prior OUD diagnosis compared to those with no prior OUD diagnosis (20.3% vs. 6.5%, p < 0.001). Among those receiving MOUD, MOUD coverage included an average of 96 days in the post-period. The median time to MOUD initiation was 102 days (IQR = 24–216). The most common form of MOUD was buprenorphine (7.9%), followed by naltrexone (1.8%) and methadone (1.5%). Rates of MOUD increased over time, from 5.1% for those with index events in 2013 to 17% for those with index events in 2020 (p < 0.001) (Figure 2). This increase in MOUD prescribing paralleled the trend in the raw rate of MOUD prescribing in Missouri Medicaid overall during this time frame, which increased steadily from 17 recipients per 1,000 enrollees in Q1 2013 to 55 recipients per 1,000 enrollees in Q4 2021 (p < 0.001). The rate of overdose during periods in which people received MOUD was 36 per 1,000 person years of observation, compared to 135 overdoses per 1,000 person years of observation without MOUD. In fully adjusted Cox models, MOUD in the post-index period was associated with significantly reduced risk (HR = 0.24, 95% CI = 0.08–0.75).
Figure 2:
Percent of study population receiving any MOUD during the 365 days after index overdose
3.3. Risk factors for repeat overdose
Risk of repeat overdose was significantly higher among those whose index overdose involved heroin or synthetic opioids (HR = 2.1, 95% CI = 1.62–2.72) (Table 2). Kaplan-Meier curves for repeat overdose by opioid-involvement in the index overdose are shown in Figure 3. Baseline diagnosis of OUD was also associated with increased risk of repeat overdose (HR = 1.5, 95% CI = 1.18–1.89), as was a diagnosis of major depressive disorder (HR = 1.42, 95% CI = 1.11, 1.82). Risk of repeat overdose increased with time; the quarter of index overdose was weakly associated with increased risk (HR = 1.02, 95% CI = 1.001, 1.04). Other baseline patient characteristics, service use, and comorbid conditions were not significantly associated with the hazards of repeat overdose.
Figure 3:
Survival probability of repeat opioid overdose during the 365 days after index overdose.
In analyses restricted to those whose index overdose involved heroin or synthetic opioids, the protective association of MOUD remained significant (HR = 0.21, 95% CI = 0.05–0.84), but MOUD was not statistically significant in the model restricted to those whose index overdoses did not involve heroin or synthetic opioids (Table 2). Results of all sensitivity models were consistent with the main findings (Appendix tables 5–7).
3.4. Cross-state comparisons
The overall rate of repeat overdose was lower in Missouri compared to previously published results from New Jersey (12.1% vs. 19.6%, p < 0.001) (Crystal et al., 2022). A significantly lower proportion of index nonfatal overdoses in MO involved heroin or synthetic opioids compared to NJ (30.1% vs. 72.3%, p < 0.001) (Crystal et al., 2022). For overdoses involving heroin or synthetic opioids, the rate of repeat overdose was not significantly different between MO and NJ. For overdoses involving prescription/other opioids only, the rate in MO was 4.6 points lower than in NJ (9% vs. 13.6%, p < 0.001). MOUD receipt after the index overdose was significantly lower in Missouri compared to NJ (10.7% vs. 21.7%, p < 0.001) (Crystal et al., 2022).
4. Discussion
Consistent with expectations, time-varying MOUD receipt was significantly protective against repeat overdose, with a 76% reduction in hazard of subsequent nonfatal overdose despite delays in MOUD initiation. This finding was driven by individuals whose index overdose involved heroin or synthetic opioids. Second, also consistent with expectations, those whose index overdoses involved heroin or synthetic opioids were at greater risk of repeat overdose than those whose index overdose involved prescription/other opioids only. Finally, among those whose overdoses involving heroin or synthetic opioids, major depressive disorder was associated with increased hazard of repeat overdose.
In our population, those whose index overdoses involved heroin and synthetic opioids were at greatest risk of repeat overdose, but also exhibited the greatest opportunity for mitigation of risk. The rate of repeat overdose among those whose index overdoses involved heroin and synthetic opioids was twice as great, and repeat overdose is of special concern for fentanyl users. Even low doses of fentanyl can lead to fentanyl-induced muscle rigidity (FIMR), also known as wooden chest syndrome (WCS), which involves rigidity in the diaphragm, upper airway, and chest wall (Roan et al., 2018; Torralva & Janowsky, 2019). This rigidity can interfere with resuscitation and patients who have overdosed on fentanyl may require higher doses of naloxone than patients who overdose on other opioids (Fairbairn et al., 2017; Schumann et al., 2008). Some prior research has indicated that the ratio of fatal to nonfatal fentanyl overdoses can be greater than 1:1, compared to 1:10 for heroin overdose without the involvement of fentanyl (Slavova et al., 2017). Special attention to mitigation of risk is warranted for those whose overdoses involve an opioid such as fentanyl that carries elevated risk.
While index overdoses involving heroin and synthetic opioids were at greater risk of repeat overdose, this group was also the driver of the protective effect found for MOUD. MOUD receipt was associated with a 79% reduction in risk in the heroin/synthetic overdose population. Heroin and synthetic opioid overdoses thus present a greater risk for patients but also a substantial opportunity for providers to mitigate risk of repeat overdose with prompt MOUD treatment.
Among those whose index overdoses did not involve heroin or synthetic opioids, major depressive disorder was associated with 50% increased risk of repeat overdose. This result aligns with prior findings of positive association between depressive symptoms and increased opioid risk behavior (Cleland et al., 2020).
Patterns of overdose and treatment in Missouri Medicaid differed from those previously reported in the Eastern US. Index overdoses in MO were less than half as likely as index overdoses in NJ to involve heroin or synthetic opioids. This is consistent with findings that heroin and fentanyl are less dominant in Missouri’s illicit opioid supply than in New Jersey’s though the rate of heroin and fentanyl in Missouri opioid seizures is steadily increasing (Appendix Figure 1) (National Forensic Laboratory Information System, 2022). This difference in the heroin and synthetic opioid supply may have contributed to the 7.5 point lower overall rate of repeat overdose in Missouri.
Over time, Missouri has implemented policy changes to make MOUD more widely available. Many barriers may prevent patients from accessing MOUD (Sharma et al., 2017). To address limited access to MOUD in an increasingly risky illicit drug environment, Missouri began using State Targeted Response funding to implement the statewide Medication First program in 2017. The Medication First program required publicly funded addiction treatment programs to include MOUD as part of services following a low-threshold delivery model. During implementation of Medication First, increased use of buprenorphine drove increased rates of MOUD overall, greater retention in MOUD treatment, and decreased costs for care (Winograd et al., 2020). In 2019, the Missouri legislature took an additional step to facilitate MOUD by passing Senate Bill 514, which prohibited prior authorizations as well as annual and lifetime limits for MOUD medications (Missouri Senate, 2019).
In this study, consistent with policy changes and findings of increased prescribing from the evaluation of Medication First (Winograd et al., 2020), rates of MOUD receipt tripled over the study time frame, from 5.1% for index overdoses in 2013 to 17% in 2020, with MOUD receipt peaking in 2020 and repeat overdoses peaking in 2016. While these low rates may reflect in part that many in the sample did not have an OUD diagnosis, since OUD, rather than opioid overdose, is the indication for MOUD treatment (Substance Abuse and Mental Health Services Administration, 2023), the rate of MOUD receipt among those with a prior OUD diagnosis only was also low, peaking at 32.3% in 2020. Until September 2018, however, STR funds were only used to support those without any insurance (Missouri Department of Mental Health, 2018), and very few Opioid Treatment Programs in Missouri accept Medicaid (Missouri Department of Mental Health, 2023), which may explain the low rates of methadone treatment relative to other forms of MOUD in spite of increases in MOUD overall over the study period. The relatively low proportion of patients receiving MOUD as late as 2020 illustrates that, despite progress over time, MOUD access remains exceedingly low in MO and other states since even the higher estimates in the literature suggest only about 1 in 5 patients receive MOUD in the year after nonfatal overdose. This result has now been identified in Medicaid and all-payer data, and in both expansion and non-expansion states (Crystal et al., 2022; Larochelle et al., 2018). Evidence from this analysis suggests a substantial opportunity exists to reduce repeat overdose in the disproportionately minoritized and low-income populations served by Medicaid.
4.1. Limitations
Effect estimates for MOUD in the population whose index overdoses involved prescription or other opioids only, as opposed to heroin or synthetic opioids, may have been affected by limited statistical power. Fewer than 1 in 20 patients in the prescription/other overdose group received any MOUD in the year after index. It should also be noted that a large randomized trial of buprenorphine/naloxone has found substantial improvements related to buprenorphine receipt in opioid dependence and abstinence for those with dependence on prescription opioids (Weiss & Rao, 2017). Further research is needed to fully understand the effectiveness of MOUD in the prescription/other opioid overdose population. Treatment need for MOUD may also have been overestimated to the extent that some in the population did not meet criteria for OUD.
The Missouri Medicaid claims contained only a single self-report value each for gender and race, and did not include data about Hispanic or Latinx ethnicity. As a result, this research is unable to identify differences in overdose risk among those of non-binary gender or Latinx ethnicity. Future research using data sets with additional demographic detail could further illuminate gender and ethnic differences in risk.
Another implication of the use of Medicaid claims is that treatment under self-pay or insurance types other than Medicaid is not reflected in these results. While the study used a washout period to ensure that no Medicaid claims for MOUD were present in the 6 months prior to the index overdose, this does not guarantee that patients had not received MOUD before the washout period or outside of Medicaid. Since the study identified index events in the pre-expansion period, these findings cannot be generalized directly to Missouri Medicaid expansion patients; however, similar results have been found in other states following Medicaid expansion (Crystal et al., 2022).
This work focused only on Medicaid beneficiaries in a single state. While Medicaid covers a substantial proportion of those at risk of opioid overdose, these results may not generalize to overdose survivors who are uninsured or covered by other insurance types. Within Medicaid, further research is needed to understand the impact of expansion on access to MOUD among vulnerable groups (Orgera & Tolbert, 2019).
5. Conclusions
This analysis of overdose risk following a nonfatal overdose found that MOUD was highly protective against repeat overdose. Survivors of an overdose involving heroin or synthetic opioids were at greater risk of repeat overdose and were also the drivers of findings of a protective effect for MOUD. Few patients received MOUD and some of the observed overdoses might have been prevented if MOUD were more widely used. These findings align with prior research in Medicaid samples as well as other insurance types, illustrating that the benefits of MOUD, as well as risk of heroin and synthetic opioids, can be observed across variation in geography, policy environment, and fentanyl penetration. These results point to the importance of proactive treatment for patients who have had an opioid overdose. Prompt treatment with MOUD, especially for survivors of heroin and fentanyl overdose, is a critical tool in efforts to prevent subsequent overdoses.
Highlights.
Only 10.7% of patients received MOUD after non-fatal opioid overdose.
MOUD was associated with a 76% reduced hazard of repeat opioid overdose.
Heroin/synthetic opioid overdose was associated with doubled risk of repeat overdose.
Acknowledgements
The authors wish to thank Samantha Lesinski for assistance in preparing the manuscript for publication.
Funding
This work was supported by the National Institute on Drug Abuse grant R01 DA047347. The National Institute on Drug Abuse had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the paper; and decision to submit the paper.
Appendix Tables and Figures
Appendix Figure 1:
Ratio of seizures of fentanyl, fentanyl analogs, and heroin to seizures of other opioid types in Missouri and New Jersey, 2013–2020
Notes: Data on drug seizures taken from the National Forensic Laboratory Information System (NFLIS) public data tables 2 and 3 for the years 2013 to 2020 (National Forensic Laboratory Information System, 2022). The denominator is based on the top 60 most commonly seized drug types nationally in each year. Opioids that are not in the top 60 most commonly seized drugs are not included in the denominator.
Appendix Table 1:
ICD-9-CM and ICD-10-CM Codes Used to Identify Opioid Poisoning
| Opioid Type | ICD-9-CM Codes | ICD-10-CM Codes |
|---|---|---|
| Heroin poisoning | 965.01, E85.00 | T40.1 |
| Synthetic opioid poisoning | No ICD-9-CM Codes | T40.4 |
| Methadone poisoning | 965.02, E85.01 | T40.3 |
| Other opioid poisoning | 965.00, 965.09, E85.02 | T40.0, T40.2, T40.6 |
Appendix Table 2:
ICD-9-CM and ICD-10-CM Codes Used to Identify Comorbidities
| Disease or Disease Category | ICD-9-CM Codes | ICD-10-CM Codes |
|---|---|---|
| Alcohol use disorder | 291.0, 291.1, 291.2, 291.3, 291.4, 291.5, 291.81, 291.82, 291.89, 291.9, 303.xx, 305.00, 305.01, 305.02, 305.03 | F10.10, F10.120, F10.121, F10.129, F10.14, F10.150, F10.151, F10.159, F10.180, F10.181, F10.182, F10.188, F10.19, F10.20, F10.21, F10.220, F10.221, F10.229, F10.230, F10.231, F10.232, F10.239, F10.24, F10.250, F10.251, F10.259, F10.26, F10.27, F10.280, F10.281, F10.282, F10.288, F10.29, F10.920, F10.921, F10.929, F10.94, F10.950, F10.951, F10.959, F10.96, F10.97, F10.980, F10.981, F10.982, F10.988, F10.99 |
| Benzodiazepine use disorder | 304.10, 304.11, 304.12, 304.13, 305.40, 305.41, 305.42, 305.43 | F13.10, F13.120, F13.20, F13.21, F13.220, F13.221, F13.229, F13.230, F13.231, F13.232, F13.239, F13.24, F13.250, F13.251, F13.259, F13.26, F13.27, F13.280, F13.281, F13.282, F13.288, F13.29, F13.90 |
| Cannabis use disorder | 304.30, 304.31, 304.32, 304.33, 305.20, 305.21, 305.22, 305.23, | F12.10, F12.20, F12.21, F12.220, F12.221, F12.222, F12.229, F12.250, F12.251, F12.259, F12.280, F12.288, F12.29, F12.90 |
| Opioid use disorder | 304.00, 304.01, 304.02, 304.03, 305.50, 305.51, 305.52, 305.53 | F11.10, F11.120, F11.121, F11.122, F11.129, F11.13, F11.14, F11.150, F11.151, F11.159, F11.181, F11.182, F11.188, F11.19, F11.20, F11.220, F11.221, F11.222, F11.229, F11.23, F11.24, F11.250, F11.251, F11.259, F11.281, F11.282, F11.288, F11.29 |
| Bipolar disorder | 296.00, 296.01, 296.02, 296.03, 296.04, 296.05, 296.06, 296.40, 296.41, 296.42, 296.43, 296.44, 296.45, 296.46, 296.50, 296.51, 296.52, 296.53, 296.54, 296.55, 296.56, 296.60, 296.61, 296.62, 296.63, 296.64, 296.65, 296.66, 296.7, 296.80, 296.81, 296.82, 296.89 | F31, F31.0, F31.1, F31.10, F31.11, F31.12, F31.13, F31.2, F31.3, F31.30, F31.31, F31.32, F31.4, F31.5, F31.6, F31.60, F31.61, F31.62, F31.63, F31.64, F31.7, F31.70, F31.71, F31.72, F31.73, F31.74, F31.75, F31.76, F31.77, F31.78, F31.8, F31.81, F31.89, F31.9 |
| Major depressive disorder | 296.20, 296.21, 296.22, 296.23, 296.24, 296.25, 296.26, 296.30, 296.31, 296.32, 296.33, 296.34, 296.35, 296.36 | F32, F32.0, F32.1, F32.2, F32.3, F32.4, F32.5, F32.8, F32.81, F32.89, F32.9, F32.A, F33, F33.0, F33.1, F33.2, F33.3, F33.4, F33.40, F33.41, F33.42, F33.8, F33.9 |
| Anxiety Disorder | 293.84, 300.00–300.02, 300.09, 300.20–300.23, 300.29, 300.3, 300.7, 308.0–308.4, 308.9, 309.21, 309.81, 312.39, 313.0, 313.21, 313.23 | F064, F4001, F4002, F4010, F40219, F40240, F40241, F408-F411, F413, F418, F419, F422, F423, F428, F429, F4521, F4522, F430, F4310, F4312, F633, F6389, F930, F938, F940, R457 |
| Schizophrenia | 295.00, 295.01, 295.02, 295.03, 295.04, 295.05, 295.10, 295.11, 295.12, 295.13, 295.14, 295.15, 295.20, 295.21, 295.22, 295.23, 295.24, 295.25, 295.30, 295.31, 295.32, 295.33, 295.34, 295.35, 295.40, 295.41, 295.42, 295.43, 295.44, 295.45, 295.50, 295.51, 295.52, 295.53, 295.54, 295.55, 295.60, 295.61, 295.62, 295.63, 295.64, 295.65, 295.70, 295.71, 295.72, 295.73, 295.74, 295.75, 295.80, 295.81, 295.82, 295.83, 295.84, 295.85, 295.90, 295.91, 295.92, 295.93, 295.94, 295.95 | F20, F20.0, F20.1, F20.2, F20.3, F20.5, F20.8, F20.81, F20.89, F20.9, F21, F22, F23, F24, F25, F25.0, F25.1, F25.8, F25.9, F28, F29 |
| Personality Disorder | 301.0–301.9 | F60.0-F60.7, F60.81, F60.89, F21, F34.0, F34.1 |
| Asthma | 493., 493.0, 493.00, 493.01, 493.02, 493.1, 493.10, 493.11, 493.12, 493.2, 493.20, 493.21, 493.22, 493.8, 493.81, 493.82, 493.9, 493.90, 493.91, 493.92 | J45., J45.2, J45.20, J45.21, J45.22, J45.3, J45.30, J45.31, J45.32, J45.4, J45.40, J45.41, J45.42, J45.5, J45.50, J45.51, J45.52, J45.9, J45.90, J45.901, J45.902, J45.903, J45.99, J45.990, J45.991, J45.998 |
| Cerebrovascular Disease | 430, 431, 431.1, 431.9, 432.0, 432.1, 432.9, 433, 433.00, 433.01, 433.10, 433.11, 433.20, 433.21, 433.31, 433.80, 433.81, 433.90, 433.91, 434.00, 434.01, 434.1, 434.10, 434.11, 434.2, 434.9, 434.90, 434.91, 435, 435.0, 435.1, 435.2, 435.3, 435.8, 435.9, 436, 437.0, 437.1, 437.2, 437.3, 437.4, 437.5, 437.6, 437.7, 437.8, 437.9, 438.11, 438.12, 438.13, 438.14, 438.19, 438.2, 438.20, 438.21, 438.22, 438.30, 438.31, 438.32, 438.40, 438.41, 438.42, 438.50, 438.51, 438.52, 438.53, 438.6, 438.7, 438.81, 438.82, 438.83, 438.84, 438.89, 438.9 | G45.0, G45.1, G45.2, G45.4, G45.8, G45.9, G46.0, G46.1, G46.2, G46.3, G46.4, G46.5, G46.6, G46.7, G46.8, I60.00, I60.01, I60.02, I60.10, I60.11, I60.12, I60.20, I60.21, I60.22, I60.30, I60.31, I60.32, I60.4, I60.50, I60.51, I60.52, I60.6, I60.7, I60.8, I60.9, I61.0, I61.1, I61.2, I61.3, I61.4, I61.5, I61.6, I61.8, I61.9, I62.00, I62.01, I62.02, I62.03, I62.1, I62.9, I63.00, I63.011, I63.012, I63.019, I63.02, I63.031, I63.032, I63.039, I63.09, I63.10, I63.111, I63.112, I63.119, I63.12, I63.131, I63.132, I63.139, I63.19, I63.20, I63.211, I63.212, I63.219, I63.22, I63.231, I63.232, I63.239, I63.29, I63.30, I63.311,I63.312, I63.319, I63.321, I63.322, I63.329, I63.331, I63.332, I63.339, I63.341, I63.342, I63.349, I63.39, I63.40, I63.411, I63.412, I63.419, I63.421, I63.422, I63.429, I63.431, I63.432, I63.439, I63.441, I63.442, I63.449, I63.49, I63.50, I63.511, I63.512, I63.519, I63.521, I63.522, I63.529, I63.531, I63.532, I63.539, I63.541, I63.542, I63.549, I63.59, I63.6, I63.8, I63.9, I65.01, I65.02, I65.03, I65.09, I65.1, I65.21, I65.22, I65.23, I65.29, I65.8, I65.9, I66.01, I66.02, I66.03, I66.09, I66.11, I66.12, I66.13, I66.19, I66.21, I66.22, I66.23, I66.29, I66.3, I66.8, I66.9, I67.1, I67.2, I67.4, I67.5, I67.6, I67.7, I67.81, I67.82, I67.841, I67.848, I67.89, I67.9, I68.0, I68.2, I68.8, I69.00, I69.020, I69.021, I69.022, I69.023, I69.028, I69.031, I69.032, I69.033, I69.034, I69.039, I69.041, I69.042, I69.043, I69.044, I69.049, I69.051, I69.052, I69.053, I69.054, I69.059, I69.061, I69.062, I69.063, I69.064, I69.065, I69.069, I69.090, I69.091, I69.092, I69.093, I69.098, I69.10, I69.120, I69.121, I69.122, I69.123, I69.128, I69.131, I69.132, I69.133, I69.134, I69.139, I69.141, I69.142, I69.143, I69.144, I69.149, I69.151, I69.152, I69.153, I69.154, I69.159, I69.161, I69.162, I69.163, I69.164, I69.165, I69.169, I69.190, I69.191, I69.192, I69.193, I69.198, I69.20, I69.220, I69.221, I69.222, I69.223, I69.228, I69.231, I69.232, I69.233, I69.234, I69.239, I69.241, I69.242, I69.243, I69.244, I69.249, I69.251, I69.252, I69.253, I69.254, I69.259, I69.261, I69.262, I69.263, I69.264, I69.265, I69.269, I69.290, I69.291, I69.292, I69.293, I69.298, I69.30, I69.320, I69.321, I69.322, I69.323, I69.328, I69.331, I69.332, I69.333, I69.334, I69.339, I69.341, I69.342, I69.343, I69.344, I69.349, I69.351, I69.352, I69.353, I69.354, I69.359, I69.361, I69.362, I69.363, I69.364, I69.365, I69.369, I69.390, I69.391, I69.392, I69.393, I69.398, I69.80, I69.820, I69.821, I69.822, I69.823, I69.828, I69.831, I69.832, I69.833, I69.834, I69.839, I69.841, I69.842, I69.843, I69.844, I69.849, I69.851, I69.852, I69.853, I69.854, I69.859, I69.861, I69.862, I69.863, I69.864, I69.865, I69.869, I69.890, I69.891, I69.892, I69.893, I69.898, I69.90, I69.920, I69.921, I69.922, I69.923, I69.928, I69.931, I69.932, I69.933, I69.934, I69.939, I69.941, I69.942, I69.943, I69.944, I69.949, I69.951, I69.952, I69.953, I69.954, I69.959, I69.961, I69.962, I69.963, I69.964, I69.965, I69.969, I69.990, I69.991, I69.992, I69.993, I69.998 |
| Chronic Pain | 307.81, 337.0, 337.1, 338.0, 338.2, 338.4, 339, 346, 350.2, 354.0, 354.4, 355–357, 377, 710–739, 784.0 | E08.42, E09.42, E10.42, E11.42, E13.42, G43–G44, G50.1, G56.0, G56.4, G57, G58.9, G60–G65, G89.0, G89.2, G89.4, G90.0, G99.0, H46–H47, M00–M02, M05–M08, M11–M25, M30–M99, R26.2, R29.4, R29.898, R51 |
| Chronic Obstructive Pulmonary Disease (COPD) | 491.2, 491.20, 491.21, 491.22 | J44., J44.0, J44.1, J44.9 |
| Diabetes | 250.00, 250.01, 250.1, 250.10, 250.11, 250.2, 250.20, 250.30, 250.31, 250.40, 250.41, 250.50, 250.51, 250.60, 250.61, 250.70, 250.71, 250.80, 250.81, 250.9, 250.90, 250.91, 357.2, 362, 362.01, 362.02, 362.04, 362.05, 362.06, 362.07, 362.10, 362.11, 362.12, 362.13, 362.14, 362.15, 362.16, 362.17, 362.18, 362.20, 362.21, 362.22, 362.23, 362.24, 362.25, 362.26, 362.27, 362.29, 362.30, 362.31, 362.32, 362.33, 362.34, 362.35, 362.36, 362.37, 362.40, 362.41, 362.42, 362.43, 362.50, 362.51, 362.52, 362.53, 362.54, 362.55, 362.56, 362.57, 362.60, 362.61, 362.62, 362.63, 362.64, 362.65, 362.66, 362.70, 362.71, 362.72, 362.73, 362.74, 362.75, 362.76, 362.77, 362.81, 362.82, 362.83, 362.84, 362.85, 362.89, 362.9, 366.41, 648.04, 790.21, 790.22, 790.29 | E08.311, E08.319, E08.321, E08.329, E08.331, E08.339, E08.341, E08.349, E08.351, E08.359, E08.36, E08.40, E08.42, E09.311, E09.319, E09.321, E09.329, E09.331, E09.339, E09.341, E09.349, E09.351, E09.359, E09.36, E09.40, E09.42, E10.10, E10.11, E10.21, E10.22, E10.29, E10.311, E10.319, E10.321, E10.329, E10.331, E10.339, E10.341, E10.349, E10.351, E10.359, E10.36, E10.39, E10.40, E10.41, E10.42, E10.43, E10.44, E10.49, E10.51, E10.52, E10.59, E10.610, E10.618, E10.620, E10.621, E10.622, E10.628, E10.630, E10.638, E10.641, E10.649, E10.65, E10.69, E10.8, E10.9, E11.00, E11.01, E11.21, E11.22, E11.29, E11.311, E11.319, E11.321, E11.329, E11.331, E11.339, E11.341, E11.349, E11.351, E11.359, E11.36, E11.39, E11.40, E11.41, E11.42, E11.43, E11.44, E11.49, E11.51, E11.52, E11.59, E11.610, E11.618, E11.620, E11.621, E11.622, E11.628, E11.630, E11.638, E11.641, E11.649, E11.65, E11.69, E11.8, E11.9, E13.00, E13.01, E13.10, E13.11, E13.21, E13.22, E13.29, E13.311, E13.319, E13.321, E13.329, E13.331, E13.339, E13.341, E13.349, E13.351, E13.359, E13.36, E13.39, E13.40, E13.41, E13.42, E13.43, E13.44, E13.49, E13.51, E13.52, E13.59, E13.610, E13.618, E13.620, E13.621, E13.622, E13.628, E13.630, E13.638, E13.641, E13.649, E13.65, E13.69, E13.8, E13.9, G45.3, H31.101, H31.102, H31.103, H31.109, H31.111, H31.112, H31.113, H31.119, H31.121, H31.122, H31.123, H31.129, H34.00, H34.01, H34.02, H34.03, H34.10, H34.11, H34.12, H34.13, H34.211, H34.212, H34.213, H34.219, H34.231, H34.232, H34.233, H34.239, H34.811, H34.812, H34.813, H34.819, H34.821, H34.822, H34.823, H34.829, H34.831, H34.832, H34.833, H34.839, H34.9, H35.00, H35.011, H35.012, H35.013, H35.019, H35.021, H35.022, H35.023, H35.029, H35.031, H35.032, H35.033, H35.039, H35.041, H35.042, H35.043, H35.049, H35.051, H35.052, H35.053, H35.059, H35.061, H35.062, H35.063, H35.069, H35.071, H35.072, H35.073, H35.079, H35.09, H35.101, H35.102, H35.103, H35.109, H35.111, H35.112, H35.113, H35.119, H35.121, H35.122, H35.123, H35.129, H35.131, H35.132, H35.133, H35.139, H35.141, H35.142, H35.143, H35.149, H35.151, H35.152, H35.153, H35.159, H35.161, H35.162, H35.163, H35.169, H35.171, H35.172, H35.173, H35.179, H35.20, H35.21, H35.22, H35.23, H35.30, H35.31, H35.32, H35.341, H35.342, H35.343, H35.349, H35.351, H35.352, H35.353, H35.359, H35.361, H35.362, H35.363, H35.369, H35.371, H35.372, H35.373, H35.379, H35.381, H35.382, H35.383, H35.389, H35.40, H35.411, H35.412, H35.413, H35.419, H35.421, H35.422, H35.423, H35.429, H35.431, H35.432, H35.433, H35.439, H35.441, H35.442, H35.443, H35.449, H35.451, H35.452, H35.453, H35.459, H35.461, H35.462, H35.463, H35.469, H35.50, H35.51, H35.52, H35.53, H35.54, H35.60, H35.61, H35.62, H35.63, H35.70, H35.711, H35.712, H35.713, H35.719, H35.721, H35.722, H35.723, H35.729, H35.731, H35.732, H35.733, H35.739, H35.81, H35.82, H35.89, H35.9, H36., O24.03, O24.13, O24.33, O24.83, O24.93, R73.01, R73.02, R73.09, R73.9 |
| Heart failure | 402.01, 402.11, 402.91, 404.01, 404.03, 404.11, 404.13, 404.91, 428.0, 428.1, 428.20, 428.21, 428.22, 428.23, 428.3, 428.30, 428.31, 428.32, 428.33, 428.40, 428.41, 428.42, 428.43, 428.5, 428.7, 428.9 | I11.0, I13.0, I13.2, I50.1, I50.20, I50.21, I50.22, I50.23, I50.30, I50.31, I50.32, I50.33, I50.40, I50.41, I50.42, I50.43, I50.9 |
| Hepatitis C | 070.54, 070.70, 070.71, V02.62, V12.09 | B18.2, B19.20, B19.21, Z22.52, Z86.19 |
| HIV | 042, V08, V65.44 | B20, Z21, Z71.7 |
| Hypertension | 401.0, 401.1, 401.9, 402.00, 402.01, 402.1, 402.10, 402.11, 402.2, 402.3, 402.4, 402.9, 402.90, 402.91, 403, 403.00, 403.01, 403.10, 403.11, 403.90, 403.91, 404.00, 404.01, 404.02, 404.03, 404.1, 404.10, 404.11, 404.12, 404.13, 404.90, 404.91, 404.92, 404.93, 405.0, 405.01, 405.09, 405.1, 405.11, 405.19, 405.2, 405.3, 405.4, 405.9, 405.91, 405.99 | I10., I11.0, I11.9, I12.0, I12.9, I13.0, I13.10, I13.11, I13.2, I15.0, I15.1, I15.2, I15.8, I15.9, N26.2 |
| Pneumonia | 480, 480.0, 480.1, 480.2, 480.3, 480.8, 480.9, 481, 482, 482.0, 482.1, 482.2, 482.3, 482.30, 482.31, 482.32, 482.39, 482.4, 482.40, 482.41, 482.42, 482.49, 482.8, 482.81, 482.82, 482.83, 482.84, 482.89, 482.9, 483., 483.0, 483.1, 483.8, 484, 484.1, 484.3, 484.5, 484.6, 484.7, 484.8, 485, 486 | J12, J12.0, J12.1, J12.2, J12.3, J12.8, J12.81, J12.89, J12.9, J13, J14, J15, J15.0, J15.1, J15.2, J15.20, J15.21, J15.211, J15.212, J15.29, J15.3, J15.4, J15.5, J15.6, J15.7, J15.8, J15.9, J16, J16.0, J16.8, J17, J18, J18.0, J18.1, J18.2, J18.8, J18.9 |
| Sleep Apnea | 327.20, 327.21, 327.22, 327.23, 327.24, 327.25, 327.26, 327.27, 327.29, 780.51, 780.53, 780.57 | G47.30, G47.31, G47.32, G47.33, G47.34, G47.35, G47.36, G47.37, G47.39 |
| Benzodiazepine Poisoning | 969.4 | T42.4X1A, T42.4X2A, T42.4X3A, T42.4X4A |
Appendix Table 3:
CPT/HCPCS Codes Used to Define Psychosocial Services
| Category | Codes |
|---|---|
| Any psychosocial service | 90791, 90792, 90832, 90833, 90834, 90836, 90837, 90838, 90839, 90840, 90846, 90847, 90849, 90853, 90863, 90875, 90876, E0202, E0203, G0396, G0397, G0443, G0445, G0446, G0447, G0503, G0504, G0505, G0506, G0507, G2011, G9475, H0001, H0002, H0003, H0004, H0005, H0006, H0007, H0008, H0009, H0010, H0011, H0012, H0013, H0014, H0015, H0016, H0017, H0018, H0019, H0022, H0026, H0031, H0032, H0035, H0036, H0037, H0038, H0039, H0040, H0046, H0047, H0050, H2000, H2001, H2010, H2011, H2012, H2013, H2014, H2015, H2016, H2017, H2018, H2019, H2020, H2021, H2027, H2034, H2035, H2036, OP912, OP913, OP914, OP915, P90804, P90806, P90808, P90810, P90812, P90814, P90816, P90818, P90821, P90823, P90826, P90828, P90845, P90846, Q5008, R90801, R90802, R90804, R90805, R90806, R90807, R90808, R90809, R90810, R90811, R90812, R90813, R90814, R90815, R90816, R90817, R90818, R90819, R90821, R90822, R90823, R90824, R90826, R90827, R90828, R90829, R90846, R90847, S0201, S90802, S90804, S90806, S90808, S90810, S90812, S90814, S90816, S90818, S90821, S90823, S90826, S90828, S9480, S9484, S9485, T1006, T1007, T1012, T1016, T1017, T1027, T1040, T2023, W1352, W1355, W1356, W1356L, Y9116, Y91169 |
Appendix Table 4:
HCPCS Codes and Drug Names Used to Define Medication Use
| Medication | HCPCS Codes and Drug Names |
|---|---|
| Buprenorphine | Buprenorphine Hydrochloride, Buprenorphine/naloxone, and their name brand equivalents (excluding Butrans, Belbuca, and Buprenex); HCPCS codes J0572-J0575 |
| Naltrexone | Naltrexone Hydrochloride, Naltrexone microspheres, Naltrexone-Bupropion, and their name brand equivalents; injectable naltrexone identified through HCPCS code J2315 |
| Methadone maintenance | HCPCS code H0020 |
| Benzodiazepine | Adinazolam, Alprazolam, Bentazepam, Bretazenil, Bromazepam, Brotizolam, Camazepam, Chlordiazepoxide, Cinazepam, Cinolazepam, Clobazam, Clonazepam, Clonazolam, Clorazepate, Clotiazepam, Cloxazolam, Delorazepam, Diazepam, Diclazepam, Estazolam, Ethy Carfluzepate, Etizolam, Ethyl loflazepate, Flubromazepam, Flubromazolam, Flunitrazepam, Flurazepam, Flutazolam, Flutoprazepam, Halazepam, Ketazolam, Loprazolam, Lorazepam, Lormetazepam, Medazepam, Mexazolam, Midazolam, Nifoxipam, Nimetazepam, Nitrazepam, Nordiazepam, Oxazepam, Phenazepam, Pinazepam, Prazepam, Premazepam, Pyrazolam, Quazepam, Rilmazefone, Temazepam, Thienalprazolam, Tetrazepam, Triazolam, and their name brand equivalents |
| Prescription Opioids | Levorphanol, Meperidine, Opium, Oxycodone, Hydromorphone, Fentanyl, Oxymorphone, Codeine, Morphine, Nalbuphine, Proposyphene, Tramadol, Hydrocodone, Pentazocine, Alfentanil, Morphine, Femifentanil, Sufentanil, and their name brand equivalents |
Appendix Table 5:
Sensitivity analysis of hazard of repeat overdose, with fixed effects for the healthcare organization that treated the index overdose
| Characteristics | Full sample (N = 2,962) |
|---|---|
|
| |
| HR (95% CI)* | |
| MOUD after overdose | 0.23 (0.07, 0.72)* |
| Index Overdose Characteristics | |
| Involved Heroin/Synthetic Opioids† | 1.93 (1.46, 2.55)* |
| Involved Benzodiazepines (BZ) | 0.84 (0.51, 1.38) |
| Involved stimulants | 0.57 (0.23, 1.4) |
| Resulted in Inpatient Hospitalization | 0.88 (0.67, 1.15) |
| Quarter of Index Overdose (1 = Q1 2013) | 1.02 (1, 1.03) |
| Age | |
| 12–24 | 1.03 (0.64, 1.65) |
| 25–39 | 1.19 (0.83, 1.71) |
| 40–55 | 1.04 (0.75, 1.44) |
| 56–64 | REF |
| Racial/Ethnic Group | |
| White | REF |
| Black | 0.8 (0.59, 1.08) |
| Other/Unknown | 0.99 (0.51, 1.9) |
| Gender | |
| Female | REF |
| Male | 1.21 (0.95, 1.54) |
| SUD Comorbidities †† | |
| Alcohol Use Disorder | 0.75 (0.55, 1.03) |
| Benzodiazepine Use Disorder | 0.99 (0.57, 1.72) |
| Cannabis Use Disorder | 1.21 (0.86, 1.69) |
| Opioid Use Disorder | 1.43 (1.12, 1.82)* |
| Stimulant Use Disorder | 0.82 (0.59, 1.13) |
| Psychiatric Comorbidities | |
| Bipolar Disorder | 0.94 (0.7, 1.26) |
| Major Depressive Disorder | 1.37 (1.06, 1.78)* |
| Anxiety Disorder | 0.8 (0.61, 1.04) |
| Personality Disorder | 1.34 (0.88, 2.04) |
| Schizophrenia | 0.95 (0.67, 1.34) |
| Medical Comorbidities | |
| Chronic Pain | 1.03 (0.77, 1.37) |
| Heart Failure | 1.14 (0.76, 1.71) |
| Hepatitis C | 1.25 (0.92, 1.7) |
| HIV | 1.15 (0.55, 2.38) |
| Hypertension | 1.25 (0.96, 1.64) |
| Pneumonia | 1.17 (0.81, 1.68) |
| Prescriptions Before Overdose | |
| Benzodiazepines (BZ) | |
| BZ in 6 Months Before Index | 1.15 (0.84, 1.58) |
| BZ at Time of Index | 0.95 (0.69, 1.29) |
| Prescription Opioids | |
| Opioids in 6 Months Before Index | 1.03 (0.76, 1.4) |
| Opioids at Time of Index | 1.08 (0.77, 1.51) |
| Health Service Utilization | |
| ED Visit in Prior 6 Months | 0.9 (0.7, 1.17) |
| Any psychosocial service in 6-month baseline | 0.96 (0.74, 1.25) |
| Index During Pandemic (Index >= March 2020) | 0.94 (0.62, 1.42) |
Note: CI = confidence interval; SUD = substance use disorder; BZ = benzodiazepines.
Statistically significant at p < 0.05.
Includes opioid poisonings with any diagnosis code for heroin or synthetic opioid poisoning, even if codes for methadone or other natural/semi-synthetic opioid poisoning were also present.
Reference for each comorbid condition is the absence of the condition.
Appendix Table 6:
Sensitivity analysis of hazard of repeat overdose with variable length of follow-up after index, with competing risks for death and disenrollment before 1 year
| Characteristics | Cause-specific HRs for repeat overdose, including fatal overdose (Total N = 4,097; N repeat overdose = 456) | Competing risk: Cause-specific HRs for death (N death = 320) | Competing risk: Cause-specific HRs for disenrollment before 1 year (N disenrollment before 1 year = 717) |
|---|---|---|---|
|
| |||
| HR (95% CI)* | HR (95% CI) | HR (95% CI) | |
| MOUD after overdose | 0.3 (0.11, 0.8)* | 0.3 (0.07, 1.21) | 0.27 (0.13, 0.58)* |
| Index Overdose Characteristics | |||
| Involved Heroin/Synthetic Opioids† | 1.91 (1.52, 2.41)* | 1.29 (0.95, 1.75) | 1.07 (0.89, 1.29) |
| Involved Benzodiazepines (BZ) | 0.75 (0.48, 1.16) | 0.76 (0.48, 1.22) | 0.99 (0.76, 1.3) |
| Involved stimulants | 0.66 (0.34, 1.3) | 0.93 (0.43, 2.01) | 1.15 (0.77, 1.7) |
| Resulted in Inpatient Hospitalization | 0.89 (0.71, 1.12) | 1.25 (0.99, 1.59) | 0.91 (0.76, 1.09) |
| Quarter of Index Overdose (1 = Q1 2013) | 1.02 (1.01, 1.03)* | 1 (0.99, 1.02) | 1.01 (1, 1.02) |
| Age | |||
| 12–24 | 1 (0.68, 1.46) | 0.15 (0.08, 0.3)* | 2.29 (1.62, 3.23)* |
| 25–39 | 1.05 (0.77, 1.43) | 0.47 (0.33, 0.66)* | 2.34 (1.71, 3.22)* |
| 40–55 | 0.98 (0.74, 1.3) | 0.71 (0.54, 0.93)* | 1.63 (1.19, 2.23)* |
| 56–64 | REF | REF | REF |
| Racial/Ethnic Group | |||
| White | REF | REF | REF |
| Black | 0.95 (0.75, 1.19) | 1.19 (0.9, 1.56) | 0.63 (0.51, 0.77)* |
| Other/Unknown | 1.05 (0.61, 1.79) | 0.91 (0.45, 1.85) | 1.25 (0.84, 1.86) |
| Gender | |||
| Female | REF | REF | REF |
| Male | 1.32 (1.08, 1.61)* | 1.14 (0.9, 1.45) | 1.34 (1.15, 1.56)* |
| SUD Comorbidities †† | |||
| Alcohol Use Disorder | 0.77 (0.59, 1.01) | 1.36 (1.02, 1.81)* | 1.23 (0.99, 1.51) |
| Benzodiazepine Use Disorder | 0.96 (0.59, 1.55) | 1.18 (0.61, 2.28) | 0.89 (0.58, 1.35) |
| Cannabis Use Disorder | 1.05 (0.78, 1.41) | 1.01 (0.66, 1.53) | 0.98 (0.77, 1.26) |
| Opioid Use Disorder | 1.68 (1.37, 2.07)* | 0.82 (0.63, 1.07) | 1.17 (0.98, 1.39) |
| Stimulant Use Disorder | 0.92 (0.7, 1.2) | 0.88 (0.6, 1.27) | 1.13 (0.9, 1.41) |
| Psychiatric Comorbidities | |||
| Bipolar Disorder | 1.04 (0.81, 1.33) | 0.89 (0.65, 1.2) | 0.92 (0.75, 1.13) |
| Major Depressive Disorder | 1.37 (1.1, 1.71)* | 0.74 (0.57, 0.97)* | 0.94 (0.79, 1.12) |
| Anxiety Disorder | 0.89 (0.71, 1.12) | 1.01 (0.77, 1.32) | 1.01 (0.85, 1.21) |
| Personality Disorder | 1.09 (0.75, 1.58) | 1.27 (0.81, 1.99) | 0.92 (0.67, 1.27) |
| Schizophrenia | 0.94 (0.7, 1.27) | 0.98 (0.68, 1.41) | 0.8 (0.61, 1.04) |
| Medical Comorbidities | |||
| Chronic Pain | 0.91 (0.71, 1.17) | 0.83 (0.59, 1.16) | 1.05 (0.87, 1.27) |
| Heart Failure | 1.17 (0.83, 1.66) | 1.63 (1.2, 2.22)* | 0.77 (0.49, 1.2) |
| Hepatitis C | 1.17 (0.89, 1.52) | 1.33 (0.99, 1.79) | 0.7 (0.52, 0.94)* |
| HIV | 1.27 (0.67, 2.43) | 0.35 (0.09, 1.44) | 0.27 (0.07, 1.09) |
| Hypertension | 1.26 (1, 1.59)* | 1.1 (0.84, 1.43) | 0.86 (0.71, 1.05) |
| Pneumonia | 1.35 (1, 1.83) | 1.92 (1.44, 2.54)* | 0.61 (0.42, 0.88)* |
| Prescriptions Before Overdose | |||
| Benzodiazepines (BZ) | |||
| BZ in 6 Months Before Index | 1.1 (0.84, 1.45) | 1.43 (1.05, 1.94)* | 0.92 (0.73, 1.17) |
| BZ at Time of Index | 0.97 (0.75, 1.27) | 1.16 (0.86, 1.56) | 0.94 (0.76, 1.16) |
| Prescription Opioids | |||
| Opioids in 6 Months Before Index | 0.92 (0.71, 1.2) | 1.49 (1.05, 2.1)* | 0.78 (0.64, 0.96)* |
| Opioids at Time of Index | 1.06 (0.79, 1.41) | 1.39 (0.96, 2) | 0.71 (0.57, 0.89)* |
| Health Service Utilization | |||
| ED Visit in Prior 6 Months | 1.03 (0.82, 1.29) | 1.13 (0.85, 1.5) | 1.05 (0.88, 1.25) |
| Any psychosocial service in 6-month baseline | 0.98 (0.78, 1.23) | 1 (0.75, 1.33) | 0.98 (0.81, 1.18) |
| Index During Pandemic (Index >= March 2020) | 0.91 (0.63, 1.3) | 1.21 (0.77, 1.88) | 0.2 (0.12, 0.33)* |
Note: CI = confidence interval; SUD = substance use disorder; BZ = Benzodiazepines.
Statistically significant at p < 0.05.
Includes opioid poisonings with any diagnosis code for heroin or synthetic opioid poisoning, even if codes for methadone or other natural/semi-synthetic opioid poisoning were also present.
Reference for each comorbid condition is the absence of the condition.
Appendix Table 7:
Sensitivity analysis of hazard of repeat overdose, not including index events during the COVID-19 pandemic (on or after March 1st, 2020)
| Characteristics | Sample (N = 2,651) |
|---|---|
|
| |
| HR (95% CI)* | |
| MOUD after overdose | 0.28 (0.09, 0.89)* |
| Index Overdose Characteristics | |
| Involved Heroin/Synthetic Opioids† | 2.19 (1.65, 2.91)* |
| Involved Benzodiazepines (BZ) | 0.76 (0.45, 1.27) |
| Involved stimulants | 0.58 (0.26, 1.33) |
| Resulted in Inpatient Hospitalization | 0.85 (0.64, 1.11) |
| Quarter of Index Overdose (1 = Q1 2013) | 1.02 (1, 1.03) |
| Age | |
| 12–24 | 0.77 (0.48, 1.24) |
| 25–39 | 1.01 (0.7, 1.46) |
| 40–55 | 0.92 (0.66, 1.28) |
| 56–64 | REF |
| Racial/Ethnic Group | |
| White | REF |
| Black | 0.9 (0.69, 1.18) |
| Other/Unknown | 1.04 (0.51, 2.12) |
| Gender | |
| Female | REF |
| Male | 1.21 (0.95, 1.53) |
| SUD Comorbidities †† | |
| Alcohol Use Disorder | 0.75 (0.53, 1.04) |
| Benzodiazepine Use Disorder | 0.82 (0.43, 1.53) |
| Cannabis Use Disorder | 1.14 (0.79, 1.62) |
| Opioid Use Disorder | 1.46 (1.14, 1.87)* |
| Stimulant Use Disorder | 0.95 (0.68, 1.33) |
| Psychiatric Comorbidities | |
| Bipolar Disorder | 1.07 (0.79, 1.44) |
| Major Depressive Disorder | 1.46 (1.13, 1.9)* |
| Anxiety Disorder | 0.76 (0.58, 1) |
| Personality Disorder | 1.21 (0.78, 1.89) |
| Schizophrenia | 1.01 (0.71, 1.44) |
| Medical Comorbidities | |
| Chronic Pain | 0.95 (0.7, 1.29) |
| Heart Failure | 1.06 (0.69, 1.64) |
| Hepatitis C | 1.21 (0.88, 1.66) |
| HIV | 1.04 (0.48, 2.25) |
| Hypertension | 1.2 (0.91, 1.59) |
| Pneumonia | 1.38 (0.96, 1.98) |
| Prescriptions Before Overdose | |
| Benzodiazepines (BZ) | |
| BZ in 6 Months Before Index | 1.19 (0.86, 1.65) |
| BZ at Time of Index | 0.97 (0.71, 1.33) |
| Prescription Opioids | |
| Opioids in 6 Months Before Index | 1.07 (0.78, 1.47) |
| Opioids at Time of Index | 1.12 (0.79, 1.6) |
| Health Service Utilization | |
| ED Visit in Prior 6 Months | 1 (0.76, 1.31) |
| Any psychosocial service in 6-month baseline | 1 (0.76, 1.31) |
Note: CI = confidence interval; SUD = substance use disorder; BZ = benzodiazepines.
Statistically significant at p < 0.05.
Includes opioid poisonings with any diagnosis code for heroin or synthetic opioid poisoning, even if codes for methadone or other natural/semi-synthetic opioid poisoning were also present.
Reference for each comorbid condition is the absence of the condition.
Footnotes
Conflict of Interest
Dr. Williams is the CMO and physician co-founder of Ophelia Health Inc. from which he receives equity, consulting fees, and travel reimbursement. Dr. Samples discloses that she has received consulting fees from the American Society of Addiction Medicine. No other authors have disclosures to report.
CRediT author statement
Andrew D. Tipping: Conceptualization, Methodology, Software, Formal Analysis, Investigation, Data Curation, Writing – Original Draft, Writing – Review & Editing, Visualization. Molly Nowels: Conceptualization, Methodology, Writing – Review & Editing. Clara Moore: Writing – Original Draft, Writing – Review & Editing. Hillary Samples: Conceptualization, Methodology, Writing – Review & Editing. Stephen Crystal: Conceptualization, Methodology, Supervision, Writing – Review & Editing. Arthur Robinson Williams: Conceptualization, Methodology, Writing – Review & Editing. Mark Olfson: Conceptualization, Methodology, Writing – Review & Editing. Jodi Heaps-Woodruff: Conceptualization, Methodology, Resources, Supervision, Writing – Review & Editing.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- Cleland CM, Bennett AS, Elliott L, Rosenblum A, Britton PC, & Wolfson-Stofko B (2020). Between- and within-person associations between opioid overdose risk and depression, suicidal ideation, pain severity, and pain interference. Drug and Alcohol Dependence, 206, 107734. 10.1016/j.drugalcdep.2019.107734 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Crystal S, Nowels M, Samples H, Olfson M, Williams AR, & Treitler P (2022). Opioid overdose survivors: Medications for opioid use disorder and risk of repeat overdose in Medicaid patients. Drug and Alcohol Dependence, 232, 109269. 10.1016/j.drugalcdep.2022.109269 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Department of Health and Human Services Press Office. (2021). Missouri Medicaid expansion brings quality essential health coverage to more than 275,000 Missourians. https://www.hhs.gov/about/news/2021/10/04/missouri-medicaid-expansion-brings-quality-essential-health-coverage.html
- Fairbairn N, Coffin PO, & Walley AY (2017). Naloxone for heroin, prescription opioid, and illicitly made fentanyl overdoses: Challenges and innovations responding to a dynamic epidemic. International Journal of Drug Policy, 46, 172–179. 10.1016/j.drugpo.2017.06.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Fudala PJ, Williford WO, Collins J, Raisch D, Goldsmith RJ, McNicholas L, & Tusel D (2003). Office-based treatment of opiate addiction with a sublingual-tablet formulation of buprenorphine and naloxone. The New England Journal of Medicine, 10. [DOI] [PubMed] [Google Scholar]
- Green CA, Perrin NA, Hazlehurst B, Janoff SL, DeVeaugh‐Geiss A, Carrell DS, Grijalva CG, Liang C, Enger CL, & Coplan PM (2019). Identifying and classifying opioid‐related overdoses: A validation study. Pharmacoepidemiology and Drug Safety, 28(8), 1127–1137. 10.1002/pds.4772 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Green CA, Perrin NA, Janoff SL, Campbell CI, Chilcoat HD, & Coplan PM (2017). Assessing the accuracy of opioid overdose and poisoning codes in diagnostic information from electronic health records, claims data, and death records: Identifying Opioid Overdoses. Pharmacoepidemiology and Drug Safety, 26(5), 509–517. 10.1002/pds.4157 [DOI] [PubMed] [Google Scholar]
- Hedegaard H, Miniño AM, Spencer MR, & Warner M (2021). Drug overdose deaths in the United States, 1999–2020. NCHS Data Brief, 428. 10.15620/cdc:112340 [DOI] [PubMed] [Google Scholar]
- Johnson RE, Eissenberg T, Stitzer ML, Strain EC, Liebson IA, & Bigelow GE (1995). A placebo controlled clinical trial of buprenorphine as a treatment for opioid dependence. Drug and Alcohol Dependence, 40(1), 17–25. 10.1016/0376-8716(95)01186-2 [DOI] [PubMed] [Google Scholar]
- Karmali RN, Ray GT, Rubinstein AL, Sterling SA, Weisner CM, & Campbell CI (2020). The role of substance use disorders in experiencing a repeat opioid overdose, and substance use treatment patterns among patients with a non-fatal opioid overdose. Drug and Alcohol Dependence, 209, 107923. 10.1016/j.drugalcdep.2020.107923 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Krawcyk N, Mojtabai R, Stewart EA, Fingerhood M, Agus D, Lyons BC, Weiner JP, & Saloner B (2020). Opioid agonist treatment and fatal overdose risk in a state-wide US population receiving opioid use disorder services. Addiction, 115, 1683–1694. 10.1111/add.14991 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Larochelle MR, Bernson D, Land T, Stopka TJ, Wang N, Xuan Z, Bagley SM, Liebschutz JM, & Walley AY (2018). Medication for opioid use disorder after nonfatal opioid overdose and association with mortality: A cohort study. Annals of Internal Medicine, 169(3), 137. 10.7326/M17-3107 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ling W, Casadonte P, Bigelow G, Kampman KM, Patkar A, Bailey GL, Rosenthal RN, & Beebe KL (2010). Buprenorphine implants for treatment of opioid dependence: A randomized controlled trial. JAMA, 304(14), 1576. 10.1001/jama.2010.1427 [DOI] [PubMed] [Google Scholar]
- Ling W, Charuvastra C, Collins JF, Batki S, Brown LS, Kintaudi P, Wesson DR, Mcnicholas L, Tusel DJ, Malkerneker U, Renner JA, Santos E, Casadonte P, Fye C, Stine S, Wang RIH, & Segal D (1998). Buprenorphine maintenance treatment of opiate dependence: A multicenter, randomized clinical trial. Addiction, 93(4), 475–486. 10.1046/j.1360-0443.1998.9344753.x [DOI] [PubMed] [Google Scholar]
- Mattson CL (2021). Trends and geographic patterns in drug and synthetic opioid overdose deaths—United States, 2013–2019. MMWR. Morbidity and Mortality Weekly Report, 70. 10.15585/mmwr.mm7006a4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Missouri Department of Mental Health. (2018, September 28). STR SOR and Medicaid Benefits. Dmh.Mo.Gov. https://dmh.mo.gov/media/pdf/str-sor-and-medicaid-benefits
- Missouri Department of Mental Health. (2023, January 1). Directory of Opioid (Methadone) Treatment Programs. Dmh.Mo.Gov. https://dmh.mo.gov/media/pdf/directory-opioid-methadone-treatment-programs
- Missouri overdose information. (2022). MIMH Addiction Science. https://www.mimhaddisci.org/missouri-overdose-data
- Missouri Senate. (2019, July 11). SB514—Modifies provisions relating to health care. Senate.Mo.Gov. https://www.senate.mo.gov/19info/BTS_Web/Bill.aspx?SessionType=R&BillID=5255853
- National Forensic Laboratory Information System. (2022). NFLIS publications. https://www.nflis.deadiversion.usdoj.gov/publicationsRedesign.xhtml
- Olfson M, Wall M, Wang S, Crystal S, & Blanco C (2018). Risks of fatal opioid overdose during the first year following nonfatal overdose. Drug and Alcohol Dependence, 190, 112–119. 10.1016/j.drugalcdep.2018.06.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Orgera K, & Tolbert J (2019, May 24). The opioid epidemic and Medicaid’s role in facilitating access to treatment. Kaiser Family Foundation. https://www.kff.org/medicaid/issue-brief/the-opioid-epidemic-and-medicaids-role-in-facilitating-access-to-treatment/ [Google Scholar]
- Roan JP, Bajaj N, Davis FA, & Kandinata N (2018). Opioids and chest wall rigidity during mechanical ventilation. Annals of Internal Medicine, 168(9), 678. 10.7326/L17-0612 [DOI] [PubMed] [Google Scholar]
- Saloner B, Daubresse M, & Caleb Alexander G (2017). Patterns of buprenorphine-naloxone treatment for opioid use disorder in a multistate population. Medical Care, 55(7), 669–676. 10.1097/MLR.0000000000000727 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Santo T, Clark B, Hickman M, Grebely J, Campbell G, Sordo L, Chen A, Tran LT, Bharat C, Padmanathan P, Cousins G, Dupouy J, Kelty E, Muga R, Nosyk B, Min J, Pavarin R, Farrell M, & Degenhardt L (2021). Association of Opioid Agonist Treatment With All-Cause Mortality and Specific Causes of Death Among People With Opioid Dependence: A Systematic Review and Meta-analysis. JAMA Psychiatry, 78(9), 979. 10.1001/jamapsychiatry.2021.0976 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Schumann H, Erickson T, Thompson TM, Zautcke JL, & Denton JS (2008). Fentanyl epidemic in Chicago, Illinois and surrounding Cook County. Clinical Toxicology, 46(6), 501–506. 10.1080/15563650701877374 [DOI] [PubMed] [Google Scholar]
- Sharma A, Kelly SM, Mitchell SG, Gryczynski J, O’Grady KE, & Schwartz RP (2017). Update on Barriers to Pharmacotherapy for Opioid Use Disorders. Current Psychiatry Reports, 19(6), 35. 10.1007/s11920-017-0783-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shover CL, Falasinnu TO, Dwyer CL, Santos NB, Cunningham NJ, Freedman RB, Vest NA, & Humphreys K (2020). Steep increases in fentanyl-related mortality west of the Mississippi River: Recent evidence from county and state surveillance. Drug and Alcohol Dependence, 216, 108314. 10.1016/j.drugalcdep.2020.108314 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Slavova S, Costich JF, Bunn TL, Luu H, Singleton M, Hargrove SL, Triplett JS, Quesinberry D, Ralston W, & Ingram V (2017). Heroin and fentanyl overdoses in Kentucky: Epidemiology and surveillance. International Journal of Drug Policy, 46, 120–129. 10.1016/j.drugpo.2017.05.051 [DOI] [PubMed] [Google Scholar]
- Slavova S, Quesinberry D, Costich JF, Pasalic E, Martinez P, Martin J, Eustice S, Akpunonu P, & Bunn TL (2020). ICD-10-CM–based definitions for emergency department opioid poisoning surveillance: Electronic health record case confirmation study. Public Health Reports, 135(2), 262–269. 10.1177/0033354920904087 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Substance Abuse and Mental Health Services Administration. (2023, May 10). Buprenorphine. SAMHSA.Gov. https://www.samhsa.gov/medications-substance-use-disorders/medicationscounseling-related-conditions/buprenorphine
- Torralva R, & Janowsky A (2019). Noradrenergic mechanisms in fentanyl-mediated rapid death explain failure of naloxone in the opioid crisis. Journal of Pharmacology and Experimental Therapeutics, 371(2), 453–475. 10.1124/jpet.119.258566 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Vandenbroucke JP, von Elm E, Altman DG, Gøtzsche PC, Mulrow CD, Pocock SJ, Poole C, Schlesselman JJ, & Egger M (2007). Strengthening the reporting of observational studies in epidemiology (STROBE): Explanation and elaboration. Epidemiology, 18(6), 805–835. 10.1097/EDE.0b013e3181577511 [DOI] [PubMed] [Google Scholar]
- Vivolo-Kantor A, Pasalic E, Liu S, Martinez PD, & Gladden RM (2021). Defining indicators for drug overdose emergency department visits and hospitalisations in ICD-10-CM coded discharge data. Injury Prevention, 27(Suppl 1), i56–i61. 10.1136/injuryprev-2019-043521 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Weiner SG, Baker O, Bernson D, & Schuur JD (2020). One-year mortality of patients after emergency department treatment for nonfatal opioid overdose. Annals of Emergency Medicine, 75(1), 13–17. 10.1016/j.annemergmed.2019.04.020 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Weiss RD, & Rao V (2017). The Prescription Opioid Addiction Treatment Study: What have we learned. Drug and Alcohol Dependence, 173, S48–S54. 10.1016/j.drugalcdep.2016.12.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Winograd RP, Wood CA, Stringfellow EJ, Presnall N, Duello A, Horn P, & Rudder T (2020). Implementation and evaluation of Missouri’s Medication First treatment approach for opioid use disorder in publicly-funded substance use treatment programs. Journal of Substance Abuse Treatment, 108, 55–64. 10.1016/j.jsat.2019.06.015 [DOI] [PubMed] [Google Scholar]