Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Jun 1.
Published in final edited form as: Drug Alcohol Depend. 2024 Mar 28;259:111274. doi: 10.1016/j.drugalcdep.2024.111274

Impact of jail-based methadone or buprenorphine treatment on non-fatal opioid overdose after incarceration

Teena Cherian 1, Sungwoo Lim 1, Monica Katyal 2, Keith S Goldfeld 3, Ryan McDonald 3, Ellen Wiewel 1, Maria Khan 3, Noa Krawczyk 3, Sarah Braunstein 1, Sean M Murphy 4, Ali Jalali 4, Philip J Jeng 4, Zachary Rosner 2, Ross MacDonald 3, Joshua D Lee 3
PMCID: PMC11111329  NIHMSID: NIHMS1984240  PMID: 38643529

Abstract

Background

Non-fatal overdose is a leading predictor of subsequent fatal overdose. For individuals who are incarcerated, the risk of experiencing an overdose is highest when transitioning from a correctional setting to the community. We assessed if enrollment in jail-based medications for opioid use disorder (MOUD) is associated with lower risk of non-fatal opioid overdoses after jail release among individuals with opioid use disorder (OUD).

Methods

This was a retrospective, observational cohort study of adults with OUD who were incarcerated in New York City jails and received MOUD or did not receive any MOUD (out-of-treatment) within the last three days before release to the community in 2011–2017. The outcome was the first non-fatal opioid overdose emergency department (ED) visit within 1 year of jail release during 2011–2017. Covariates included demographic, clinical, incarceration-related, and other characteristics. We performed multivariable cause-specific Cox proportional hazards regression analysis to compare the risk of non-fatal opioid overdose ED visits within 1 year after jail release between groups.

Results

MOUD group included 8,660 individuals with 17,119 incarcerations; out-of-treatment group included 10,163 individuals with 14,263 incarcerations. Controlling for covariates and accounting for competing risks, in-jail MOUD was associated with lower non-fatal opioid overdose risk within 14 days after jail release (adjusted HR=0.49, 95% confidence interval=0.33–0.74). We found no significant differences 15–28, 29–56, or 57–365 days post-release.

Conclusion

MOUD group had lower risk of non-fatal opioid overdose immediately after jail release. Wider implementation of MOUD in US jails could potentially reduce post-release overdoses, ED utilization, and associated healthcare costs.

Keywords: Opioid use disorder, medication for opioid use disorder, non-fatal overdose, emergency department, jail, urban population

1. Introduction

The opioid epidemic has been increasingly devastating communities in recent years. In 2020, roughly 2.7 million people in the United States (US) reported having an opioid use disorder (OUD) (Centers for Disease Control, 2022a) and 75% of drug overdose deaths involved an opioid (Centers for Disease Control, 2022b). Although drug addiction is a chronic medical condition requiring a public health approach to reduce harm, it continues to be criminalized (The Pew Charitable Trusts, 2022). More than half of those with OUD report involvement in the criminal legal system (Winkelman et al., 2018) and less than 10% of those with a drug dependency receive treatment while incarcerated (The Pew Charitable Trusts, 2022). In addition, there is racial bias in drug use enforcement. Although Black and White people in the US report similar rates of substance use, Black people are disproportionately incarcerated for drug charges at about six times the rate of White individuals (NAACP, 2023).

For individuals who are incarcerated, the risk of experiencing an overdose event is highest when transitioning from a correctional setting to the community, which could be due to reasons such as reduced drug tolerance and interruptions in OUD treatment (Binswanger et al., 2013; National Academies of Sciences, Engineering, and Medicine, 2019; Pizzicato et al., 2018; Ranapurwala et al., 2018). Non-fatal overdose is more common than fatal overdose (World Health Organization, 2021) and is a leading predictor of subsequent non-fatal and fatal overdose (Caudarella et al., 2016; Krawczyk et al., 2020; Saloner et al., 2020). Non-fatal overdoses are associated with a range of physical and mental morbidities including cardiac and renal conditions, post-traumatic stress disorder, and brain injury, as well as increased healthcare costs (Degenhardt et al., 2019; Schneider et al., 2021; U.S Department of Health and Human Services, 2019; Warner-Smith et al., 2002).

Several studies have demonstrated the benefits of methadone and buprenorphine, medications for OUD (MOUD), in prisons and jails. When provided during incarceration, MOUD has lowered mortality following release from jail (Bird et al., 2015; Degenhardt et al., 2014; Green et al., 2018; Huang et al., 2011; Lim et al., 2023; Marsden et al., 2017), reduced reincarceration (Westerberg et al., 2016) and self-reported drug use (Brinkley-Rubinstein et al., 2018; Heimer et al., 2006 ), improved engagement with community treatment (Brinkley-Rubinstein et al., 2018; Haas et al., 2021), and shown promise as a cost-effective intervention (Chatterjee et al., 2023). However, limited studies conducted in the US have examined the effect of MOUD provision in jail on non-fatal overdose events (Brinkley-Rubinstein et al., 2018; Haas et al., 2021). The impact of in-jail MOUD on non-fatal opioid overdoses, especially during the immediate time after release when risk of non-fatal overdose is particularly high, remains unclear. To address this gap in research, we examined whether enrollment in jail-based MOUD treatment is associated with lower risk of non-fatal opioid overdose among adults with OUD after release from New York City (NYC) jails to the community during 2011–2017.

2. Materials and methods

2.1. Study setting and design

This study was a retrospective observational cohort study and is part of a larger study to evaluate the impact of jail-based MOUD on health outcomes and OUD-related costs after release from NYC jails between May 1, 2011 and December 31, 2017 (National Institutes of Health, 2023).

NYC jails have the oldest and largest corrections-based opioid treatment program in the US. All health care service delivery to patients in NYC jails, including MOUD, is provided free of charge. Methadone and buprenorphine are the most commonly provided MOUD in NYC jails. During the medical intake evaluation at the time of incarceration, jail medical staff screen individuals to identify medical conditions and needs, including OUD. Since late-2017, all individuals with OUD are offered voluntary enrollment in the MOUD program and MOUD is made available for the duration of the incarceration. Prior to this, criminal legal status significantly determined eligibility.

2.2. Data sources

The data sources for this study were the NYC Health + Hospitals/Correctional Health Services’ (CHS) Electronic Medical Records (EMR), New York State’s Statewide Planning and Research Cooperative System (SPARCS) hospitalization and emergency department (ED) records, and NYC death certificate data. CHS data were linked with SPARCS data by performing deterministic data matching using a combination of partial identifying information such as first two letters of last name, last two letters of last name, first two letters of first name, last four digits of social security number, four digits of birth year, two digits of birth month, two digits of birth day, and first letter of sex available in SPARCS records. Additionally, CHS data were probabilistically matched with the NYC death certificate data using QualityStage Software (IBM, Armonk, NY, USA). An independent human review determined the matching quality to be acceptable (sensitivity = 97%; specificity = 96%). Additional details about the data matching process are described in a previous study (Lim et al., 2023).

2.3. Study population

The study sample included 79,115 incarceration events of 29,566 adults aged 18 years or older with diagnosis of OUD while incarcerated. We excluded: 1) incarceration events with non-community discharge (i.e., discharge to state prison or hospital) or death in custody, 2) incarceration events where no MOUD was provided at any point during the incarceration, likely reflecting non-current OUD, 3) events where only methadone-based opioid withdrawal treatment was provided for the last three days before discharge, as those individuals were on methadone doses falling between maintenance and out-of-treatment, and, 4) records of individuals who had subsequent incarceration events recorded after death, indicating potential recording or matching errors (Figure 1). The final sample included 31,382 incarceration events of 15,797 adults. The incarceration events were further categorized into the MOUD group (n = 17,119; received MOUD during last three days before release) and the out-of-treatment group (n = 14,263; received no MOUD during last three days before release). The selection of a three-day window for receipt of MOUD prior to jail release was based on the known half-life of methadone, which could be 8 to 59 hours (National Library of Medicine, 2023), after which tolerance to methadone may wane and increase overdose risk. Receiving MOUD within three days of discharge would confer some benefit against overdose risk, even if received for less than three days, for those in the MOUD group compared to those whose treatment ended earlier relative to jail release. Additional details on study setting, population, and MOUD eligibility criteria during incarceration are described elsewhere (Lim et al., 2023).

Figure 1.

Figure 1.

Open in a new tab

Flow chart of the study sample selection

MOUD: Medications for opioid use disorder

Note: The sum of individuals in MOUD and out-of-treatment groups (8,660+10,163=18,823) is greater than the unique number of individuals in the box before (15,797) because one individual with multiple incarcerations could have received different treatments at each incarceration event and be included in both MOUD and out-of-treatment groups. For example, if an individual with three incarceration events received MOUD during the first incarceration and out-of-treatment at the next two incarcerations, this person is counted once in the MOUD group and once in the out-of-treatment group.

2.4. Outcomes

The outcome was the first non-fatal opioid overdose event within 1 year of jail release during 2011–2017. We defined non-fatal opioid overdose using the Centers for Disease Control (CDC) surveillance case definition guidance for assessing drug overdose in ED and hospitalization discharge data (Vivolo-Kantor et al., 2021). We searched all available diagnosis fields in SPARCS ED data for ICD-9-CM codes (965.00, 965.01, 965.02, 9965.09, E850.0, E850.1, E850.2) and ICD-10 CM codes (T40.0X, T40.1X, T40.2X, T40.3X, T40.4X, T40.60, T40.69) to identify ED visits that occurred anywhere in New York State. ED visits that resulted in hospitalizations were only found in the hospitalization data of SPARCS. Therefore, we searched all diagnosis fields in the hospitalization data for the above ICD codes.

We considered first non-fatal, non-opioid drug overdose ED visit within 1 year of discharge, first ED visit due to reasons other than non-fatal drug overdoses within 1 year of discharge, death, and reincarceration events as competing risks. We defined non-fatal drug overdose visits using the CDC surveillance case definition in which all available diagnosis fields in SPARCS ED and hospitalization data included ICD-9-CM codes (960–979, E850-E858, E950.0-E950.5, E980.0-E980.5, E962.0) and ICD-10-CM codes (T36-T5) (Vivolo-Kantor et al., 2021). Of these, we removed non-fatal opioid overdose events to determine non-fatal, non-opioid drug overdose ED visits.

2.5. Exposure

The exposure variable was receipt of MOUD during the last three days before jail discharge. Those in the out-of-treatment group received no MOUD during this time.

2.6. Person-time

Person-time was the time that individuals spent in the community after release from jail, for up to 365 days following release. For each discharge from jail, we counted the number of days in the community to first non-fatal opioid overdose ED visit, first non-fatal, non-opioid drug overdose visit, first ED visit not due to non-fatal drug overdose, death, or reincarceration event, whichever occurred first. Those who experienced none of the five events had a person-time of 365 days, unless the 365-day period following jail discharge occurred after the end of the study period of December 31, 2017, in which case the follow-up period was right-censored.

2.7. Covariates

Covariates included demographics (age at jail release, sex, race, Hispanic ethnicity, highest level of education, marital status); housing (experiencing homelessness at the time of incarceration); clinical (alcohol use/dependence/use disorder, asthma, bipolar or related disorder, cerebrovascular disease/accident history, chronic kidney disease/kidney failure, cocaine use/dependence/use disorder, congestive heart failure, depression and depressive symptoms, diabetes mellitus, hepatitis C, HIV, hypertension, liver disease, myocardial infarction history, neoplasm diagnosis, non-asthma pulmonary diagnoses/disease, peripheral vascular disease, personality disorder, post-traumatic stress disorder [PTSD], trauma, and stress-related disorder, schizophrenia or psychotic disorder, seizure history/seizure disorder, serious mental illness); behavioral (injection drug use, current smoking); incarceration-related (length of incarceration, felony versus other charges, number of previous incarceration events, year of discharge from jail); and healthcare utilization characteristics (total number of ED visits within 1 year prior to jail discharge date). Clinical characteristics were assessed and recorded by CHS clinicians during the intake evaluation of each incarceration. Additional details on these variables and how they were collected can be found elsewhere (Lim et al., 2023).

2.8. Statistical analysis

We produced descriptive statistics for sociodemographic, clinical, behavioral, incarceration-related, and other characteristics of the study sample and compared them in the MOUD and out-of-treatment groups using standardized mean difference (SMD). We then calculated crude rates for non-fatal opioid overdose ED visits, non-opioid drug overdoses, and other ED visits. We calculated these crude rates for three follow-up time periods (2 weeks, 1 month, and 1 year after release). Finally, we performed multivariable cause-specific Cox proportional hazards regression analysis, adjusting for the demographic covariates listed above and other covariates where the SMD was greater than 0.10. Length of incarceration is correlated with charge severity and therefore was not included as a covariate in the model. To address bias due to informative censoring, we included four competing risk outcomes: non-fatal, non-opioid drug overdose ED visits; other ED visits; death; and reincarceration events. In the cause-specific hazards model, we modeled all these outcomes in a competing risk framework. We estimated hazard rates using separate models, while changing the event of interest in each case and treating each event as censoring events in the other models. To address correlation between multiple discharge events for the same person, we included individual-level random effects. Additionally, we tested the proportionality assumption for the Cox proportional hazard model using weighted Schoenfeld residuals (Grambsch and Therneau, 1994) and found violation of this assumption. For non-exposure variables, we addressed the violation by conducting stratification analysis. For the exposure variable, we divided the data into four follow-up intervals: 1–14 days, 15–28 days, 29–56 days, and 57–365 days after jail discharge, and calculated time-specific hazard ratios (HR) for MOUD. All statistical analyses were conducted using SAS version 9.4 and R version 4.2.2. Statistical significance was determined using a two-sided P-value < 0.05.

3. Results

Table 1 shows sociodemographic, clinical, and incarceration-related characteristics of eligible incarceration events for adults with OUD who were incarcerated in NYC jails (n = 31,382). The study population had an average age at discharge of 42 years. Most individuals were male (82.5%), non-Hispanic Black or Hispanic (74.0%), single or never married (78.8%), and had an education level of high school degree or less (79.7%). Experiencing homelessness at the time of incarceration was reported by 17.2% individuals. Of those incarcerated, 24.9% and 34.5% had alcohol use disorder and cocaine use disorder, respectively. The most prevalent mental health illnesses were personality disorder (13.3%), followed by depression and depressive symptoms (12.8%), and PTSD, trauma, and stress-related disorder (9.2%). Serious mental illness was designated in 5.3% of individuals. Diagnosis of liver disease and hepatitis C infection were recorded at 16.6% and 35.8% of incarceration events, respectively. In addition, 45.5% reported injection drug use and 81.7% reported current smoking. The average length of stay in jail was 60 days with 3.9% of individuals spending less than three days in jail. More than half the study population had a misdemeanor charge. The maximum number of ED visits in the MOUD group was 180 while it was 149 in the out-of-treatment group.

Table 1.

Sociodemographic, clinical, and incarceration-related characteristics of eligible incarceration events

Total (n=31,382) MOUD (n=17,119) Out-of-treatment (n=14,263) Standardized difference (95% CI)
n % n % n %
Age
 18–24 1,556 5.0 501 2.9 1,055 7.4 0.22 (0.19–0.24)
 25–44 15,934 50.8 8,711 50.9 7,223 50.6
 45–64 13,636 43.5 7,798 45.6 5,838 40.9
 65+ 256 0.8 109 0.6 147 1.0
Average age at discharge, years (SD) 42 (11) 43 (10) 41 (11) 0.14 (0.12–0.17)
Sex
 Female 5,497 17.5 3,939 23.0 1,558 10.9 0.33 (0.30–0.35)
 Male 25,885 82.5 13,180 77.0 12,705 89.1
Race and ethnicity
 Hispanic 13,684 43.6 7,236 42.2 6,448 45.2 0.11 (0.09–0.13)
 Non-Hispanic Black 9,545 30.4 5,149 30.1 4,396 30.8
 Non-Hispanic White 7,523 24.0 4,438 25.9 3,085 21.6
 Another race or ethnicitya 630 2.0 296 1.7 334 2.3
Education
 Less than high school degree 13,462 42.9 7,369 43.1 6,093 42.7 0.02 (0.001–0.05)
 High school degree or equivalent 11,532 36.8 6,322 36.9 5,210 36.5
 Some college/trade school 5,380 17.1 2,869 16.8 2,511 17.6
 Unknown 1,008 3.2 559 3.3 449 3.2
Marital status
 Single, never married 24,718 78.8 13,543 79.1 11,175 78.4 0.03 (0.01–0.06)
 Married/partnered 4,939 15.7 2,606 15.2 2,333 16.4
 Other 1,725 5.5 970 5.7 755 5.3
Housing Characteristics
 Homelessness 5,411 17.2 3,264 19.1 2,147 15.1 0.11 (0.09–0.13)
Clinical Characteristics
 Alcohol use/dependence/use disorder 7,803 24.9 4,272 25.0 3,531 24.8 0.01 (−0.02–0.03)
 Asthma 5,134 16.4 2,692 15.7 2,442 17.1 0.04 (0.02–0.06)
 Bipolar or related disorder 2,007 6.4 1,066 6.2 941 6.6 0.02 (−0.01–0.04)
 Cerebrovascular disease/accident history 17 0.1 9 0.05 8 0.06 0.02 (−0.01–0.04)
 Chronic kidney disease/kidney failure * * * * * * 0.002 (−0.02–0.02)
 Cocaine use/dependence/use disorder 10,814 34.5 6,499 38.0 4,315 30.3 0.16 (0.14–0.19)
 Congestive heart failure 79 0.3 40 0.2 39 0.3 0.01 (−0.01–0.03)
 Depression and depressive symptoms 4,029 12.8 2,171 12.7 1,858 13.0 0.01 (−0.01–0.03)
 Diabetes mellitus 1,818 5.8 966 5.6 852 6.0 0.01 (−0.01–0.04)
 Hepatitis C 11,243 35.8 7,196 42.0 4,047 28.4 0.29 (0.27–0.31)
 HIV 2,486 7.9 1,479 8.6 1,007 7.1 0.06 (0.04–0.08)
 Hypertension 3,420 10.9 1,899 11.1 1,521 10.7 0.01 (−0.01–0.04)
 Liver disease 5194 16.6 3,379 19.7 1,815 12.7 0.2 (0.17–0.21)
 Myocardial infarction history 14 0.04 * * * * 0.02 (−0.01–0.04)
 Neoplasm diagnosisb 428 1.4 222 1.3 206 1.4 0.01 (−0.01–0.04)
 Non-asthma pulmonary diagnoses/disease 181 0.6 92 0.5 89 0.6 0.01 (−0.01–0.03)
 Peripheral vascular disease 80 0.3 46 0.3 34 0.2 0.01 (−0.02–0.03)
 Personality disorder 4187 13.3 1,927 11.3 2,260 15.9 0.13 (0.11–0.16)
 PTSD, trauma, and stress-related disorder 2,886 9.2 1,542 9.0 1,344 9.4 0.01 (−0.01–0.04)
 Schizophrenia or psychotic disorder 1,606 5.1 778 4.5 828 5.8 0.06 (0.04–0.08)
 Seizure history / seizure disorder 1,893 6.0 1,156 6.8 737 5.2 0.07 (0.05–0.09)
 Serious mental illnessc 1,647 5.3 871 5.1 776 5.4 0.02 (−0.01–0.04)
Behavioral Characteristics
 Self-reported injection drug use 14,283 45.5 8,674 50.7 5,609 39.3 0.23 (0.21–0.25)
 Current smoking 25,622 81.7 14,147 82.6 11,475 80.5 0.06 (0.03–0.08)
Legal Characteristics
 Less than three days in jail 1,219 3.9 318 1.9 901 6.3 0.23 (0.20–0.25)
 Average length of incarceration, days (SD) 60 (79) 44 (60) 79 (93) 0.45 (0.43–0.47)
 Highest charge severity at/near admission
  Felony 13,234 42.2 4,481 26.2 8,753 61.4 0.90 (0.87–0.92)
  Misdemeanor 16,409 52.3 12,131 70.9 4,278 30.0
  Other 1,739 5.5 507 3.0 1,232 8.6
 Average previous NYC jail admission events since 2011, count (SD) 3 (4) 4 (5) 2 (3) 0.43 (0.40–0.45)
 Jail discharge year
  2011 2,676 8.5 1,576 9.2 1,100 7.7 0.19 (0.17–0.22)
  2012 5,327 17.0 2,956 17.3 2,371 16.6
  2013 5,304 16.9 2,683 15.7 2,621 18.4
  2014 4,973 15.9 2,529 14.8 2,444 17.1
  2015 4,662 14.9 2,359 13.8 2,303 16.2
  2016 4,196 13.4 2,277 13.3 1,919 13.5
  2017 4,244 13.5 2,739 16.0 1,505 10.6
Healthcare utilization
 Median number of emergency department visits for 1 year prior to discharge date, (IQR) 0 (0–2) 0 (0–3) 0 (0–2) 0.14 (0.11–0.16)
Open in a new tab
a

Another race or ethnicity includes individuals who reported no or unknown Hispanic ethnicity, Asian or Pacific Islander, American Indian/Alaskan, other-unspecified race, or unknown race.

b

Excludes basal cell carcinoma

c

Based on criteria from the New York State Office of Mental Health (until mid-2017) and Diagnostic and Statistical Manual of Mental Disorders fifth edition (DSM-5) (after mid-2017), based mostly on diagnoses of depressive disorders, bipolar and related disorders, schizophrenia spectrum and other psychotic disorders, and post-traumatic disorder but also on observance of severe functional impairment or clinical distress as a result of a DSM-5 diagnosis.

*

Cells that could lead to the inadvertent disclosure of private data are marked with an asterisk (*)

CHS: Correctional Health Services; CI: Confidence Interval; HIV: human immunodeficiency virus; IQR: Interquartile range; MOUD: medications for opioid use disorder; PTSD: post-traumatic stress disorder; SD: standard deviation

Based on the observed SMD, we concluded there were important imbalances between the out-of-treatment and MOUD groups with respect to a number of characteristics. The MOUD group had more females (SMD=0.33), homelessness (SMD=0.11), cocaine use disorder (SMD=0.16), reported injection drug use (SMD=0.23), personality disorder (SMD=0.13), liver disease (SMD=0.20), hepatitis C (SMD=0.29), misdemeanor charges (SMD=0.90), average previous jail admissions (SMD=0.43), and ED visits for 1 year prior to discharge date (SMD=0.14).

3.1. Crude rates of non-fatal overdose events

Of the 31,382 jail discharge events, 32% (n = 10,065) resulted in at least one ED visit within 1 year of discharge date. Non-fatal opioid overdose visits represented 5.6% of these ED visits. The crude rate of first non-fatal opioid overdose ED visit was lower in the MOUD group within 14 days after jail discharge (106 visits per 1000 person-years versus 138 per 1,000-person years; RR: 0.85; 95% confidence interval [CI] = 0.60–1.21; P = 0.37) (Table 2). The crude rate of first non-fatal opioid overdose in the first month after jail discharge was not different between MOUD and out-of-treatment groups (87 visits per 1,000 person-years versus 89 per 1,000-person years; RR: 1.08; 95% CI = 0.80–1.44; P = 0.63). Lastly, those in the MOUD group experienced significantly more non-fatal opioid overdose events within a year of jail discharge compared with the out-of-treatment group (38 visits per 1,000 person-years versus 29 per 1,000-person years; RR: 1.35; 95% CI = 1.14–1.59; P = <0.001).

Table 2.

Number and crude rates of first visit by type of ED visit, time since jail discharge, and receipt of in-jail MOUD

MOUD (n=17,119)
 Out-of-treatment (n=14,263)
Type of ED visit Time since jail discharge Number of visits Rate, visits per 1,000 person-years Number of visits Rate, visits per 1000 person-years Crude rate ratio (95% CI) P-value
Non-fatal opioid overdose visit 14 days 66 106 73 138 0.85 (0.60–1.21) 0.37
1 month 103 87 91 89 1.08 (0.80–1.44) 0.63
1 year 315 38 244 29 1.35 (1.14–1.59) <0.001
Non-fatal, non-opioid drug overdose visit 14 days 65 104 37 70 1.52 (1.01–2.28) 0.046
1 month 91 77 63 61 1.26 (0.91–1.75) 0.16
1 year 346 42 239 28 1.46 (1.24–1.73) <0.0001
ED visit not due to non-fatal drug overdose 14 days 1,736 2946 1,222 2421 1.22 (1.13–1.32) <0.0001
1 month 2,488 2291 1,718 1803 1.28 (1.20–1.36) <0.0001
1 year 5536 885 4334 661 1.34 (1.29–1.39) <0.0001
Open in a new tab

CI: Confidence Interval; ED: emergency department; MOUD: medications for opioid use disorder

3.2. Adjusted HR for non-fatal overdose events by MOUD

Table 3 presents unadjusted and adjusted hazard ratios from the cause-specific model. After accounting for potential confounding, competing risks, and individual-level random effects, we observed that in-jail MOUD was associated with reduced risk of non-fatal opioid overdose within 1–14 days of jail release (adjusted HR: 0.49; 95% CI = 0.33–0.74). Unlike the first 2 weeks, we found no associations within 15–28, 29–56, or 57–365 days after release. Per our multivariable, random effect, cause-specific regression analysis, in-jail MOUD was not associated with risk of non-fatal, non-opioid drug overdose for any time periods after jail release. Lastly, we found that the risk of an ED visit due to factors other than non-fatal drug overdose was higher in the in-jail MOUD group within the 15–28 days (HR: 1.17; 95% CI = 1.01–1.35) and 57–365 days (HR: 1.12; 95 % = 1.03–1.22) following jail discharge.

Table 3.

Cause-specific hazards model for risk of first non-fatal opioid overdose event with competing risksa (n=31,382)

Outcome Time since jail discharge Unadjusted hazard ratio (95% CI) Adjusted hazard ratio (95% CI)
Non-fatal, opioid overdose 1–14 days 0.67 (0.46–0.99) 0.49 (0.33–0.74)
15–28 days 1.29 (0.60–2.79) 0.96 (0.44–2.09)
29–56 days 1.71 (0.82–3.57) 1.29 (0.61–2.73)
57–365 days 0.81 (0.49–1.32) 0.64 (0.38–1.07)
Non-fatal, non-opioid drug overdose 1–14 days 1.76 (1.05–2.97) 1.37 (0.79–2.39)
15–28 days 1.52 (0.69–3.31) 1.29 (0.58–2.89)
29–56 days 1.44 (0.72–2.91) 1.13 (0.55–2.33)
57–365 days 1.73 (1.07–2.77) 1.43 (0.87–2.35)
ED visit not due to non-fatal drug overdose 1–14 days 1.04 (0.96–1.13) 1.01 (0.92–1.10)
15–28 days 1.28 (1.13–1.45) 1.17 (1.01–1.35)
29–56 days 1.14 (1.02–1.28) 1.13 (0.99–1.29)
57–365 days 1.19 (1.10–1.28) 1.12 (1.03–1.22)
Death 1–14 days 0.33 (0.15–0.72) 0.32 (0.14–0.73)
15–28 days 0.08 (0.01–0.58) 0.05 (0.007–0.43)
29–56 days 0.16 (0.02–1.30) 0.21 (0.02–1.99)
57–365 days 1.37 (0.86–2.18) 1.18 (0.71–1.98)
Recidivism 1–14 days 1.37 (1.25–1.50) 1.15 (1.04–1.28)
15–28 days 1.29 (1.16–1.44) 1.14 (1.01–1.28)
29–56 days 1.16 (1.06–1.27) 1.02 (0.93–1.13)
57–365 days 1.12 (1.06–1.18) 1.03 (0.97–1.09)
Open in a new tab
a

A random effect was estimated, including a Gaussian frailty term to the model. Multivariable models include demographic (age, sex, race and ethnicity, marital status, the highest level of education), clinical (personality disorder diagnosis, cocaine use disorder, liver disease, Hepatitis C virus diagnosis), behavioral (injecting drug use), housing (homelessness), health-care utilization (total numbers of emergency department visits during 1 year prior to discharge date), and legal characteristics variables (felony charges versus other charges, count of previous incarceration events, year of discharge from jail).

CI: Confidence Interval; ED: emergency department; MOUD: medications for opioid use disorder

4. Discussion

Our findings suggest that among individuals who are incarcerated and have OUD, those who were enrolled in a jail-based MOUD treatment program during three days prior to jail discharge had lower risk of experiencing a non-fatal opioid overdose ED visit within 2 weeks after jail release compared with those who received no MOUD during the three days prior to release. Our results are similar to another retrospective study in Connecticut, which used administrative data to show that those who received MOUD while incarcerated had a reduced likelihood of non-fatal overdose due to any drug following jail release, although specific follow-up time periods after jail discharge were not examined (Haas et al., 2021). A Rhode Island study that used self-reported non-fatal overdoses found similar results (Brinkley-Rubinstein, 2018).

Research exploring the impact of in-jail MOUD on non-overdose related ED visits post-incarceration is limited. Contrary to this study, a study in California showed reduced likelihood of an ED visit due to any reason among those who received MOUD while incarcerated (Will et al., 2022). The two studies have different geographic settings– highly dense urban city versus less populated county. Individuals in the MOUD group may visit EDs for methadone doses if their opioid treatment programs (OTPs) are not open, particularly on weekends, as people can be released from jail any time. This may be especially common in a large, urban setting like NYC where housing is geographically dispersed. Other differences between the two studies include study time periods (2011–2017 versus 2019–2021) and regional healthcare landscapes.

The null association between in-jail MOUD and non-fatal overdose after the first two weeks post-jail release in our study may reflect challenges in accessing and continuing care at OTPs. Community reentry after incarceration is often a fraught time, with overlapping concerns about housing stability, employment, food, and social relationships, and possibly criminal legal situations which do not necessarily cease upon release. This may be true for people who had contacts with health and other programming prior to their incarceration, including OTP participation. Though the OTPs of referral for individuals in this study were not examined and efforts are made by correctional health staff to make referrals appropriate to each patient’s experience and situation, OTPs are highly regulated which may create access barriers. Further, patient experience with these OTPs with regards to satisfaction and perceptions about continuity of MOUD were also unknown, but may create other barriers (Sanders et al., 2013; Vail et al., 2021). Understanding and removing such barriers, including expansion of dispensation of methadone to settings other than OTPs such as mobile vans, office-based clinics, and pharmacies could improve continuity of care in all settings (Gibbons 2022; Krawczyk et al., 2023).

Our findings have implications for policymakers and clinicians. Firstly, MOUD should be widely available in correctional facilities for those with OUD and discharge plans should facilitate transition to community-based OUD treatment programs. Providing this care while incarcerated ensures coordination and continuity of care, which may also promote access to other medical services post-incarceration, thereby reducing fragmented care for this at-risk population. Additionally, healthcare practitioners in the ED could play a critical role in preparing patients to avoid subsequent non-fatal and fatal overdose events (Adams, 2018; Soares et al., 2022). When patients present to an ED with a non-fatal opioid overdose event, ED healthcare providers should discuss substance use without shame or stigma, educate patients on overdose prevention strategies and treatment programs, provide harm reduction services, and ensure that those engaged in MOUD are linked to a community treatment program.

Our study has several strengths. To our knowledge, this is the first study to examine risk of non-fatal opioid overdose as the main outcome while modeling other outcomes in a competing risk framework. We assessed risk at specific time periods after jail release to identify when risk is highest and how patterns change over the year after jail release. In addition, we included many covariates related to health and incarceration-related characteristics that could explain differences in MOUD and out-of-treatment groups. Furthermore, we used administrative data from a large urban jail system, as well as from state and city health registries.

Our study has some limitations. First, ED visit data might be underestimating the true incidence of non-fatal opioid overdose events as not everyone who experiences an overdose event presents at an ED. Individuals may be having non-fatal opioid overdose events at home or elsewhere and receive treatment by bystanders or emergency medical services (EMS) but choose not to be transported to an ED or prefer to seek care at an outpatient facility. Future research should supplement ED data with EMS and other outpatient data to understand the true number of non-fatal opioid overdose events in this population. Additionally, we do not have data on incarcerations occurring outside of NYC among persons released from NYC jails; however, since our follow-up time is immediately after jail release to the community and since the vast majority are from NYC or surrounding counties, we do not believe that bias is large enough to change our main findings.

5. Conclusion

The growing opioid epidemic has drastic consequences that are especially pronounced among those who are incarcerated. Our study shows that MOUD in jails could potentially reduce post-release non-fatal overdoses, which might also reduce ED utilization and associated healthcare costs. Offering MOUD in jail could be lifesaving for many people with OUD after release and ensuring access to evidence-based treatment upon jail release might result in improved quality of life and social stability for this vulnerable population.

Highlights.

  • Medication for opioid use disorder program in jail lowered non-fatal overdose risk

  • Examined non-fatal overdose risk while addressing various competing risk events

  • Assessed non-fatal overdose risk at specific time periods after jail release

  • Used data from a large urban jail system and state and city health registries

Funding sources

This work was supported by the National Institutes of Health/National Institute on Drug Abuse (R01DA045042-01A1).

Footnotes

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.

Declarations of interest: None

Declaration of competing interests

None to declare.

Disclaimer

The raw data used to produce this publication was purchased from or provided by the New York State Department of Health (NYSDOH). However, the calculations, metrics, conclusions derived, and views expressed herein are those of the author(s) and do not reflect the conclusions or views of NYSDOH. NYSDOH, its employees, officers, and agents make no representation, warranty or guarantee as to the accuracy, completeness, currency, or suitability of the information provided here.

References

  1. Adams JM, 2018. Increasing Naloxone Awareness and Use: The Role of Health Care Practitioners. JAMA 319(20):2073–2074. [DOI] [PubMed] [Google Scholar]
  2. Binswanger IA, Blatchford PJ, Mueller SR, Stern MF, 2013. Mortality after prison release: opioid overdose and other causes of death, risk factors, and time trends from 1999 to 2009. Ann. Intern. Med 159(9):592–600. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bird SM, Fischbacher CM, Graham L, Fraser A, 2015. Impact of opioid substitution therapy for Scotland’s prisoners on drug-related deaths soon after prisoner release. Addiction. 110:1617–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Brinkley-Rubinstein L, McKenzie M, Macmadu A, Larney S, Zaller N, Dauria E, Rich J, 2018. A randomized, open label trial of methadone continuation versus forced withdrawal in a combined US prison and jail: Findings at 12 months post-release. Drug Alcohol Depend. 184:57–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Caudarella A, Dong H, Milloy MJ, Kerr T, Wood E, Hayashi K, 2016. Non-fatal overdose as a risk factor for subsequent fatal overdose among people who inject drugs. Drug Alcohol Depend. 162:51–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Centers for Disease Control, 2022a. Opioid Use Disorder. Available at: https://www.cdc.gov/dotw/opioid-use-disorder/index.html (accessed 6 April 2023).
  7. Centers for Disease Control, 2022b. Death Rate Maps & Graphs. Available at: https://www.cdc.gov/drugoverdose/deaths/index.html (accessed 6 April 2023).
  8. Chatterjee A, Weitz M, Savinkina A, Macmadu A, Madushani RWMA, Potee RA, Ryan D, Murphy SM, Walley AY, Linas BP, 2023. Estimated Costs and Outcomes Associated With Use and Nonuse of Medications for Opioid Use Disorder During Incarceration and at Release in Massachusetts. JAMA Netw. Open. 6(4):e237036. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Degenhardt L, Larney S, Kimber J, Gisev N, Farrell M, Dobbins T, Weatherburn DJ, Gibson A, Mattick R, Butler T, Burns L, 2014. The impact of opioid substitution therapy on mortality postrelease from prison: retrospective data linkage study. Addiction.109:1306–17. [DOI] [PubMed] [Google Scholar]
  10. Degenhardt L, Grebely J, Stone J, Hickman M, Vickerman P, Marshall BDL, Bruneau J, Altice FL, Henderson G, Rahimi-Movaghar A, Larney S, 2019. Global patterns of opioid use and dependence: harms to populations, interventions, and future action. Lancet. 394(10208):1560–1579. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Gibbons J, Stuart E, Saloner B, 2022. Methadone on Wheels-A New Option to Expand Access to Care Through Mobile Units. JAMA Psychiatry. 79(3):187–188. [DOI] [PubMed] [Google Scholar]
  12. Grambsch PM, Therneau T, 1994. Proportional hazards tests and diagnostics based on weighted residuals. Biometrika 81:515–26. [Google Scholar]
  13. Green TC, Clarke J, Brinkley-Rubinstein L, Marshall BDL, Alexander-Scott N, Boss R, Rich JD, 2018. Postincarceration Fatal Overdoses After Implementing Medications for Addiction Treatment in a Statewide Correctional System. JAMA Psychiatry. 75(4):405–407. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Haas A, Viera A, Doernberg M, Barbour R, Tong G, Grau LE, Heimer R, 2021. Post-incarceration outcomes for individuals who continued methadone treatment while in Connecticut jails, 2014–2018. Drug Alcohol Depend. 227:108937. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Heimer R, Catania H, Newman RG, Zambrano J, Brunet A, Ortiz AM, 2006. Methadone maintenance in prison: evaluation of a pilot program in Puerto Rico. Drug Alcohol Depend. 83(2):122–9. [DOI] [PubMed] [Google Scholar]
  16. Huang YF, Kuo HS, Lew-Ting CY, Tian F, Yang CH, Tsai TI, Gange SJ, Nelson KE, 2011. Mortality among a cohort of drug users after their release from prison: an evaluation of the effectiveness of a harm reduction program in Taiwan. Addiction. 106:1437–45. [DOI] [PubMed] [Google Scholar]
  17. Krawczyk N, Eisenberg M, Schneider KE, Richards TM, Lyons BC, Jackson K, Ferris L, Weiner JP, Saloner B, 2020. Predictors of Overdose Death Among High-Risk Emergency Department Patients With Substance-Related Encounters: A Data Linkage Cohort Study. Ann. Emerg. Med 75(1):1–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Krawczyk N, Joudrey P, Simon R, Russel D, Frank D, 2023. Recent modifications to the US methadone treatment system are a Band-Aid—not a solution—to the nation’s broken opioid use disorder treatment system. Health Affairs Scholar. 1(1). Available at: 10.1093/haschl/qxad018 (accessed 24 January 2024) [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Lim S, Cherian T, Katyal M, Goldfeld KS, McDonald R, Wiewel E, Khan M, Krawczyk N, Braunstein S, Murphy SM, Jalali A, Jeng PJ, MacDonald R, Lee JD, 2023. Association between jail-based methadone or buprenorphine treatment for opioid use disorder and overdose mortality after release from New York City jails 2011–17. Addiction. 118(3):459–467. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Marsden J, Stillwell G, Jones H, Cooper A, Eastwood B, Farrell M, Lowden T, Maddalena N, Metcalfe C, Shaw J, Hickman M, 2017. Does exposure to opioid substitution treatment in prison reduce the risk of death after release? A national prospective observational study in England. Addiction. 112:1408–18. [DOI] [PubMed] [Google Scholar]
  21. NAACP, 2023. Criminal Justice Fact Sheet. https://naacp.org/resources/criminal-justice-fact-sheet (accessed 29 August 2023).
  22. National Academies of Sciences, Engineering, and Medicine, 2019. Medications for opioid use disorder save lives. Washington, DC: The National Academies Press. doi: 10.17226/25310. [DOI] [PubMed] [Google Scholar]
  23. National Institutes of Health, 2023. Report Project Details. Impact of jail-based methadone on overdose, recidivism, HIV and health outcomes, and costs in New York City, 2011–2017. Available at: https://reporter.nih.gov/projectdetails/9596037 (accessed 3 April 2023).
  24. National Library of Medicine, 2023. Methadose Drug Label Information. Available at: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=808a9d0b-720b-4034-a862-5122ff514608 (Accessed 6 February 2024).
  25. Pizzicato LN, Drake R, Domer-Shank R, Johnson CC, Viner KM, 2018. Beyond the walls: Risk factors for overdose mortality following release from the Philadelphia Department of Prisons. Drug Alcohol Depend. 189:108–115. [DOI] [PubMed] [Google Scholar]
  26. Ranapurwala SI, Shanahan ME, Alexandridis AA, Proescholdbell SK, Naumann RB, Edwards D Jr., Marshall SW., 2018. Opioid Overdose Mortality Among Former North Carolina Inmates: 2000–2015. Am. J. Public Health. 108(9):1207–1213. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Saloner B, Chang HY, Krawczyk N, Ferris L, Eisenberg M, Richards T, Lemke K, Schneider KE, Baier M, Weiner JP, 2020. Predictive Modeling of Opioid Overdose Using Linked Statewide Medical and Criminal Justice Data. JAMA Psychiatry. 77(11):1155–1162. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Sanders JJ, Roose RJ, Lubrano MC, Lucan SC, 2013. Meaning and methadone: patient perceptions of methadone dose and a model to promote adherence to maintenance treatment. J Addict Med 7(5):307–13. [DOI] [PubMed] [Google Scholar]
  29. Schneider KE, Tomko C, Nestadt DF, Silberzahn BE, White RH, Sherman SG, 2021. Conceptualizing overdose trauma: The relationships between experiencing and witnessing overdoses with PTSD symptoms among street-recruited female sex workers in Baltimore, Maryland. Int. J. Drug Policy. 92:102859. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Soares WE 3rd., Melnick ER, Nath B, D’Onofrio G, Paek H, Skains RM, Walter LA, Casey MF, Napoli A, Hoppe JA, Jeffery MM, 2022. Emergency Department Visits for Nonfatal Opioid Overdose During the COVID-19 Pandemic Across Six US Health Care Systems. Ann. Emerg. Med 79(2):158–167. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. The Pew Charitable Trusts, 2022. Drug Arrests Stayed High Even as Imprisonment Fell From 2009 to 2019. https://www.pewtrusts.org/-/media/assets/2022/02/drug-arrests-stayed-high-even-as-imprisonment-fell-from-2009-to-2019.pdf (accessed 29 August 2023).
  32. U.S. Department of Health and Human Services, 2019. Non-Fatal Opioid Overdose and Associated Health Outcomes: Final Summary Report. Available at: https://aspe.hhs.gov/reports/non-fatal-opioid-overdose-associated-health-outcomes-final-summary-report-0#results (accessed 7 April 2023).
  33. Vail W, Faro E, Watnick D, Giftos J, Fox AD, 2021. Does incarceration influence patients’ goals for opioid use disorder treatment? A qualitative study of buprenorphine treatment in jail. Drug Alcohol Depend. 222:108529. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Vivolo-Kantor A, Pasalic E, Liu S, Martinez PD, Gladden RM, Overdose Morbidity Team, 2021. Defining indicators for drug overdose emergency department visits and hospitalisations in ICD-10-CM coded discharge data. Inj. Prev 27(S1):i56–i61. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Warner-Smith M, Darke S, Day C, 2002. Morbidity associated with non-fatal heroin overdose. Addiction. 97(8):963–7. [DOI] [PubMed] [Google Scholar]
  36. Westerberg VS, McCrady BS, Owens M, Guerin P, 2016. Community-Based Methadone Maintenance in a Large Detention Center is Associated with Decreases in Inmate Recidivism. J. Subst. Abuse Treat. 70:1–6. [DOI] [PubMed] [Google Scholar]
  37. Will J, Abare M, Olson M, Chyorny A, Wilhelm-Leen E, 2022. Emergency department utilization by individuals with opioid use disorder who were recently incarcerated. J. Subst. Abuse. Treat. 141:108838. [DOI] [PubMed] [Google Scholar]
  38. Winkelman TNA, Chang VW, Binswanger IA, 2018. Health, Polysubstance Use, and Criminal Justice Involvement Among Adults With Varying Levels of Opioid Use. JAMA Netw. Open. 1(3):e180558. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. World Health Organization, 2021. Opioid overdose. Available at: https://www.who.int/news-room/fact-sheets/detail/opioid-overdose (accessed 7 April 2023).

RESOURCES