The American Psychiatric Association (APA) has updated its Privacy Policy and Terms of Use, including with new information specifically addressed to individuals in the European Economic Area. As described in the Privacy Policy and Terms of Use, this website utilizes cookies, including for the purpose of offering an optimal online experience and services tailored to your preferences.

Please read the entire Privacy Policy and Terms of Use. By closing this message, browsing this website, continuing the navigation, or otherwise continuing to use the APA's websites, you confirm that you understand and accept the terms of the Privacy Policy and Terms of Use, including the utilization of cookies.

×
ARTICLEFull Access

“Catching Chain” With Medicaid: The Impact of Medicaid Access on Opioid Overdose Mortality in Adults Released From State Detention

Abstract

Objective

To combat the heightened risk of opioid overdose death for individuals with criminal justice involvements, enhanced access to Medicaid remains paramount. This study examines the effect of a 2017 policy change in Pennsylvania that allowed for suspension, rather than termination, of Medicaid coverage while in prison on post‐release opioid overdose mortality risk (OOMR) for adults released from Pennsylvania prisons.

Methods

This retrospective cohort study utilizes administrative records from the Pennsylvania Department of Corrections for adults ≥18 years diagnosed with opioid use disorder (OUD) released in either 2015 or 2018. Death certificate data were used to compare OOMR within a year after release, both prior to and following the policy change. Demographic, treatment, and death characteristics were assessed with bivariate analyses. Multivariable logistic regressions were used to examine the association between qualification of Medicaid suspension and post‐release crude mortality and OOMR.

Results

Qualification for Medicaid suspension was not associated with a significant decrease in OOMR (OR = 0.82, 95% CI [0.47–1.46]) or crude mortality (OR = 1.02, 95% CI [0.67–1.57]) within 1 year after prison release for individuals diagnosed with OUD. In addition, the risk of opioid overdose mortality after release from prison for adults with an OUD increased from 2015 to 2018 (0.6%–1.7%; p < 0.0001), particularly from synthetic narcotics (57.1%–83.1%, p < 0.001), and drug overdose remained the leading cause of death (74.0%) within the initial year of release.

Conclusions

These findings underscore the need for ongoing research in overdose prevention tailored to criminal justice populations. This includes a deeper analysis of policies intended to enhance post‐release insurance continuity and their effect on individuals diagnosed with OUD after prison release.

HIGHLIGHTS

  • Qualifying for Medicaid suspension was not associated with a significant decrease in opioid overdose mortality within one year of prison release.

  • Drug overdose remained the leading cause of post‐release death for adults with an opioid use disorder diagnosis.

  • The risk of opioid overdose mortality after prison release increased between 2015 and 2018, particularly from synthetic narcotics.

Individuals released from prison have an increased risk of overdose death compared to their demographic nonincarcerated counterparts (1, 2, 3). The heightened risk of drug overdose death, in contrast to the general population (4, 5), disproportionately affect females (6) and whites (2, 6, 7, 8) in prison and is further elevated within the few weeks following release (1, 9). These overdose deaths in the post‐release period predominantly involve opioids (10) and remain the leading cause of death for individuals that return to communities from prison (1). Factors that contribute to overdose deaths after release from prison, include various demographics, such as, gender and race (6, 7, 11), as well as histories of mental health and substance use (6, 8), socioeconomic status (12), and incarceration related factors (2, 6, 7, 8, 13).

Admission to prison, in and of itself, can elevate the risk of an opioid overdose death. Individuals undergo abrupt changes in drug behavior upon correctional admission (14), which include, reductions in frequency of use (15), purity of substances consumed (15), as well as increases in solitary substance use (16). Disruptions in social networks—a critical tool for individuals that seek recovery from substance use—can also occur on an account of imprisonment, such as, rises in interpersonal relationships with companions that experienced overdoses (17, 18), and halted communications with established and trusted dealers, which coax individuals to purchase from unfamiliar sources upon release (11, 19). Other exposures of imprisonment, such as, restrictive housing (20) and higher security prisons and criminal justice supervision (20, 21), further increase the level of vulnerability an individual has to drug overdose mortality after release. Though, interruptions in healthcare, such as, reductions in substance use treatment and harm reduction services (12, 22, 23, 24, 25, 26, 27), often take precedence in post‐release discussions on overdose deaths bearing their profound effects on substance use outcomes.

Access to Medicaid—the principal payer of post‐release health care services (28)—remains critical for individuals with opioid use disorder (OUD) for continued health care upon release. State expansions in Medicaid has been associated with increased enrollment (29, 30, 31, 32, 33, 34) and streamlined access to various healthcare services upon release, such as treatment services for OUD (35, 36, 37), services for mental health (31, 32), and other healthcare services (30, 32, 38), thus supporting efforts to mitigate overdose death (22, 39). Moreover, these state expansions in Medicaid, as well as federal expansions, reduce financial barriers to access treatment services and may even assist with recidivism (32, 40).

Recognizing these benefits, Pennsylvania has made Medicaid more accessible for individuals involved in the criminal justice system. Prior to 2017, Medicaid coverage terminated upon entry into the criminal justice system and individuals had to reapply for Medicaid upon release. This often caused a lapse in insurance coverage at a particularly vulnerable time for the individual. In May of 2017, Act 76 was amended to authorize the suspension, rather than termination, of Medicaid benefits for imprisoned individuals, which streamlined the process of restoring access to Medicaid coverage after release. Supporting Information S1: Supplement A provides a chronological representation of various Medicaid policy changes pertinent to individuals with criminal justice involvement in Pennsylvania between 2011 and 2020.

Although previous studies have established that state Medicaid expansions have led to increased Medicaid enrollment and streamlined access to post‐release healthcare services, the effects of Medicaid expansions, such as coverage suspension during imprisonment, on opioid use, and more critically, opioid overdose mortality, are poorly understood. In addition, few studies focus on individuals with an OUD diagnosis, a demographic at heightened risk of opioid overdose mortality. Consequently, this study aims to investigate whether changing eligibility criteria, from Medicaid termination to Medicaid suspension, for adults exiting prison with an OUD diagnosis effects opioid overdose mortality risk (OOMR) within 1 year after release. We hypothesize that the policy shift will decrease the probability of post‐release opioid overdose mortality among released individuals diagnosed with OUD.

METHODS

Data Sources

This retrospective cohort study uses administrative records from Pennsylvania Department of Corrections (PADOC) of released adults with an OUD diagnosis between January 1, 2015 and December 31, 2015 or January 1, 2018 and December 31, 2018. PA‐DOC collects data on the state's prison population, which includes demographic, substance use, treatment, and custody characteristics. Personal identifiers (i.e., name [and possible aliases], date of birth, sex, and race) for each individual released during the period of interest were probabilistically linked to death records from the Center for Disease Control and Prevention (CDC) National Death Index (NDI). The NDI, a central computerized index of national death records, returns records with probabilistic scores to indicate potential matches of study populations and corresponding death characteristics (date of death, state of death, and causes of death).

Participants

The sample (n = 5984) included individuals aged ≥18 at release who had an OUD diagnosis and were released from a state correctional institutions (SCI) between January 1, 2015 and December 31, 2015, and January 1, 2018 and December 31, 2018, prior to and subsequent to the suspension of Medicaid throughout imprisonment in Pennsylvania. Figure 1 describes how the sample was constructed. The presence of OUD was established using the Texas Christian University Drug Screen II, the screening tool used by PA‐DOC to determine placement and level of care in treatment for those with OUDs and defined as a core score of three or greater—corresponding to a DSM III‐R diagnosis for drug dependency—combined with an opiate listed as any substance of choice. Deaths for individuals with OUD diagnoses (n = 150) were included when the date of death was within a year of release.

image

FIGURE 1. PRISM flow chart for inclusion and exclusion criteria of sample participants.

Measurements

The main outcome of interest was death from an opioid overdose. The presence of an opioid overdose death was identified with underlying cause of death (UCD) and multiple cause of death (MCD) codes from the International Classification of Disease Tenth Revision, which is further defined in Supporting Information S1: Supplement B. Consistent with prior studies (41, 42), UCDs drug poisoning—irrespective of whether the intent was classified as unintentional (X40–X44), suicide (X60–X64), homicide (X85), or undetermined (Y10–Y14)—were identified in conjunction with the presence of any opioid (MCD: T40.0–T40.4, T40.6). Categorical variables were also created to indicate drug poisonings by type of opioid: T40.0 (Opium); T40.1 (Heroin); T40.2 (natural and semi‐synthetic opioids [i.e., prescription opioids]), T40.3 (methadone), T40.4 (synthetic narcotics, excluding methadone [i.e., fentanyl]), or T40.6 (other and unspecified narcotics). In addition, prescription opioids were identified with MCD codes T40.2 and T40.3. Categories were not mutually exclusive, as deaths may have involved multiple substances, and were developed in correspondence to the methods used in the CDC National Center for Health Statistics.

Qualification for Medicaid suspension was determined by time served within PA‐SCIs, as the department, per state law, had the authority to suspend Medicaid for a period no >2 years. Individuals that served sentences >2 years were deemed ineligible to have their Medicaid benefits suspended throughout imprisonment. Drug‐defined offenses, defined in accordance with definitions from the Bureau of Justice Statistics (43), were defined as any violations of law that prohibit or regulate the possession, use, distribution, or manufacture of illegal drugs (Supporting Information S1: Supplement C). Receipt of behavioral health services was defined as admission to a therapeutic community, operated by the state prison system, where care was provided while in detention, irrespective of track, treatment length, level of care determination, whether the participation was a sentence requirement imposed by a judge, or whether the participation was prompted following screening results upon admittance. Medications for opioid use disorder was defined as a Food and Drug Administration approved medication (buprenorphine, methadone, or naltrexone) provided to an individual with OUD during imprisonment.

Statistical Analysis

Annual frequencies and proportions were calculated for extracted variables from the administratively linked dataset. Frequencies were computed for sex, race, pharmacological treatment, behavioral health service participation, and categorical groupings of age at release, which were then stratified by year of release (2015 vs. 2018). Demographic, treatment, and death characteristics were assessed with bivariate analyses, and multivariable logistic regressions were used to examine the association between qualification for Medicaid suspension and post‐release crude and OOMR, as well as to assess potential risk factors for crude and opioid overdose mortality within a year after prison release. Sensitivity analyses were also conducted to compare the mandatory minimum sentence length established by the court compared to actual time served to ensure proper identification of those that qualify for coverage suspension based on correctional length. All statistical analyses were conducted using SAS Version 9.4, and significance levels were set at 0.05. The study received approval from the Pennsylvania State University Internal Review Board with a waiver of informed consent.

RESULTS

Sample Characteristics

The sample, further described in Table 1, contained 5984 individuals aged ≥18 who were released from PA‐SCIs with an OUD diagnosis between January 1, 2015 and December 31, 2015, or January 1, 2018 and December 31, 2018. The sample was primarily white (77.9%), male (81.5%), and in their 30s (43.5%) at their latest date of release. The mean length of imprisonment was 36.6 ± 36.1 months and, throughout imprisonment, most participated in behavioral health services (62.4%) and few were provided pharmacological treatment (6.1%) and were imprisoned for a drug‐defined offense (26.9%). While PA‐SCIs commonly house fewer transient populations in comparison to other detention facilities, approximately half (49.1%) of the sample qualified for suspension of Medicaid based on time served.

TABLE 1. Characteristics of individuals (≥18) with OUD diagnosis released from SCI in Pennsylvania, 2015 and 2018; n (%) unless otherwise noted.a
YearTotal20152018p‐value
n598425213463
Died within 1 year after release150 (2.5%)47 (1.9%)103 (3.0%)<0.01**
Opioid overdose73 (1.2%)14 (0.6%)59 (1.7%)<0.001***
Sentence Qualificationb2939 (49.1%)1384 (54.9%)1555 (44.9%)0.17
Drug‐defined offensec1610 (26.9%)719 (28.5%)891 (25.7%)0.02*
Enrolled in SCI MOUDd program363 (6.1%)3 (0.1%)360 (10.4%)<0.001***
Participation in SCI therapeutic community2734 (62.4%)1824 (72.4%)1910 (55.2%)<0.001***
Sex
Male4876 (81.5%)2098 (83.2%)2778 (80.2%)<0.01**
Female1108 (18.5%)423 (16.8%)685 (19.8%)<0.01**
Race
White4663 (77.9%)1948 (77.3%)2715 (78.4%)0.30
Black880 (14.7%)373 (14.8%)507 (14.6%)0.88
Othere441 (7.4%)200 (7.9%)241 (7.0%)0.16
Age at release
18–291934 (32.3)948 (37.6%)986 (28.5%)<0.001***
30–392600 (43.5%)995 (39.5%)1605 (46.4%)<0.001***
40–49959 (16.0%)375 (14.9%)584 (16.9%)0.04*
50–59405 (6.8%)161 (6.4%)244 (7.1%)0.32
60+86 (1.4%)42 (1.7%)44 (1.3%)0.23

aMOUD, medications for opioid use disorder; OUD, opioid use disorder; SCI, state correctional institutions.

bSentence qualification was determined by the time served within a SCI and does not include any clinical or financial factors that are used in Medicaid qualification determinations.

cDrug‐defined offenses include any violations of laws that prohibited or regulated the possession, use, distribution, or manufacture of illegal drugs, which are in accordance with the definition of the Bureau of Justice Statistics (Bureau of Justice Statistics, 1994). More details available in Supporting Information S1: Supplement C.

dAn MOUD includes a medication approved by the Food and Drug Administration for the treatment of OUD (buprenorphine, methadone, or naltrexone).

eOther includes Chinese, Japanese, Hawaiian (includes part‐Hawaiian), Filipino, Other Asian or Pacific Islander, Indian (includes American, Alaskan, Canadian, Mexican Indian; Eskimo and Aleut), other non‐white (includes Cajun and Creole), and unknown, not stated, or not classifiable.

*p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001. Chi‐squares were run with fisher's exact tests with frequencies <5.

TABLE 1. Characteristics of individuals (≥18) with OUD diagnosis released from SCI in Pennsylvania, 2015 and 2018; n (%) unless otherwise noted.a
Enlarge table

Pooled analyses showed an increased death risk (1.9%–3.0%, p < 0.01) within the year after release, with increased risk of post‐release deaths attributed to opioid overdoses (0.6%–1.7%; p < 0.0001). The primary substance identified in opioid overdoses was synthetic opioids (57.1%–83.1%; p < 0.001). In addition, more individuals participated in pharmacological treatment (0.1%–10.4%; p < 0.001) and less participated in behavioral health services (72.4%–55.2%; p < 0.001) throughout their detentions in PA‐SCIs. Furthermore, more women (16.8% to 19.8%; p < 0.01) and individuals in their 30s with OUD diagnosis (aged 30–39, 39.5%–46.4%, p < 0.001) were released from PA‐SCIs in 2015 and 2018.

Characteristics of Deceased within a Year of Prison Release

Table 2 describes the individuals identified as deceased (n = 150) within a year of their release in the NDI database. The deceased sample was primarily white (86.0%), male (85.3%), in their 30s (43.5%), and (46.0%) died within 6 months following release. Drug overdose (74.0%) was the leading cause of death within a year of release, which was followed by motor vehicle (5.8%), and cardiovascular disease (3.3%). The most common opioid listed as a contributing cause of a drug overdose was synthetic opioids (78.1%), followed by heroin (13.7%), natural and semi‐synthetic opioids (4.1%), and other and unspecified opioids (4.1%). There were no deaths attributed to methadone or opium within the sample. While most deaths (91.3%) occurred within the Commonwealth of Pennsylvania, nearly all (98.0%) deaths occurred among states in the east coast of the United States.

TABLE 2. Characteristics of deceased individuals (≥18) with OUD diagnosis released from state correctional institutions in Pennsylvania, 2015 and 2018; n (%) unless otherwise noted.a
YearTotal20152018p‐value
n15047103
Leading cause of death
Drug overdoseb111 (74.0%)33 (70.2%)78 (75.7%)0.55
Motor vehicle8 (5.3%)2 (4.3%)6 (5.8%)1.00
Cardiovascular disease5 (3.3%)2 (4.3%)3 (2.9%)0.65
Other causes26 (17.3%)10 (21.3%)16 (15.5%)0.49
Opioid overdose73 (48.7%)14 (29.8%)59 (57.3%)<0.01**
Synthetic narcotics, excluding methadone57 (78.1%)8 (57.1%)49 (83.1%)<0.001***
Heroin10 (13.7%)6 (42.9%)4 (6.8%)0.07
Natural and semi‐synthetic opioids (i.e., prescription)3 (4.1%)0 (0.0%)3 (5.1%)0.55
Other and unspecified narcotics3 (4.1%)0 (0.0%)3 (5.1%)0.55
Death occurred within Pennsylvania137 (91.3%)44 (93.6)93 (90.3%)0.76
Length of sentence (≤2 years)69 (46.0%)26 (55.3%)43 (41.8%)0.16
Days between release and death
1–29 days22 (14.7%)4 (8.5%)18 (17.5%)0.21
30–179 days69 (46.0%)23 (48.9%)46 (46.0%)0.72
180–365 days59 (39.3%)20 (42.6%)37 (35.9%)0.59
Sex
Male128 (85.3%)42 (2.0%)86 (3.1%)<0.01*
Female22 (14.7%)5 (1.2%)17 (2.5%)0.18
Race
White129 (86.0%)41 (2.1%)88 (3.2%)0.02*
Black12 (8.0%)4 (1.1%)8 (1.6%)0.57
Other9 (6.0%)2 (1.0%)7 (2.9%)0.19
Age at release
18–2936 (24.0%)12 (25.5%)24 (23.3%)0.84
30–3970 (46.7%)23 (48.9%)47 (45.6%)0.73
40–4930 (20.0%)7 (14.9%)23 (22.3%)0.38
50–5910 (6.7%)2 (4.3%)8 (7.8%)0.73
60+4 (2.7%)3 (6.4%)1 (1.0%)0.09

aOUD, opioid use disorder.

bDrug overdose includes individuals with the following underlying causes of death codes: X40–X44, X60–X64, X85, Y10–Y14. Opioid overdose deaths include any of the underlying cause of death for an overdose (listed above) in combination with an opioid (T40.0–T40.4, T40.6) listed as a contributing cause of death. More details available in Supporting Information S1: Supplement B.

Significance was set at the 0.05 level for each test (*p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001). For categorical variables, chi‐squares were used with fisher's exact tests for frequencies <5. For continuous variables, t tests were used.

TABLE 2. Characteristics of deceased individuals (≥18) with OUD diagnosis released from state correctional institutions in Pennsylvania, 2015 and 2018; n (%) unless otherwise noted.a
Enlarge table

Post‐Release Mortality Risk and Medicaid Suspension

The risk of post‐release mortality, as depicted in Figures 2 and 3, exhibit fluctuations based on the qualification status for Medicaid suspension among individuals released in 2015 and 2018. In 2015, individuals, whether eligible or ineligible for benefit suspension, exhibited comparable risk of post‐release crude mortality (1.9% vs. 1.9%, p > 0.05) and opioid overdose mortality (0.6% vs. 0.5%, p > 0.05) after release from a PA‐SCI. Three years later, following the implementation of Medicaid suspension, despite an overall increase in both post‐release crude mortality and opioid overdose mortality, post‐release OOMR rose at a similar rate among individuals that qualified for Medicaid suspension (0.6%–1.5%, p < 0.05) compared to OOMR for individuals deemed ineligible (0.5%–1.9%; p < 0.01), who consequently had their benefits terminated.

image

FIGURE 2. Risk of opioid overdose mortality for released adults with an opioid use disorder diagnosis from PA‐SCIs, 2015 and 2018. PA‐SCIs, Pennsylvania state correctional institutions.

image

FIGURE 3. Risk of crude mortality for released adults with an opioid use disorder diagnosis from Pennsylvania state correctional institutions, 2015 and 2018.

Multivariable Logistic Regressions

The results of the multivariable analyses are reported in Table 3. Few covariates were significantly associated with risk of crude or opioid overdose death, including the primary variable of interested—the interaction between qualification for Medicaid suspension and release in 2018 (OOMR: OR = 0.82, 95% CI [0.47–1.46]; and crude mortality: OR = 1.02, 95% CI [0.67–1.57]). However, the direction of the association is consistent with the hypothesis that suspension of Medicaid is associated with reduced risk of post‐release opioid overdose death for individuals with OUD. In addition, Black race was associated with a statistically reduced risk of crude mortality (OR = 0.44, 95% CI [0.24–0.81)] compared to White individuals, and a release date in 2018 was associated with significantly increased risk of post‐release opioid overdose death (OR = 3.17, 95% CI [1.74–5.77]) in comparison to a release date in 2015.

TABLE 3. Multivariable logistic regression examining the association between qualification of Medicaid suspension and post‐release crude and opioid overdose mortality risk, and other risk factors for mortalities within a year after prison release.a
Model 1: Opioid overdose mortalityModel 2: Crude mortality
OR95% CIOR95% CI
LowerUpperLowerUpper
Age
30–39RefRef
18–290.670.371.210.730.491.11
40–491.160.632.151.250.801.93
50–590.920.362.391.010.511.98
60+0.940.137.032.140.756.07
Race
WhiteRefRef
Black0.750.361.530.440.240.81
Other0.650.231.830.620.311.24
Sex
MaleRefRef
Female0.780.411.520.700.441.12
Pharmacological treatment (i.e., MOUD)
Not provided an MOUDRefRef
Provided an MOUD0.480.151.550.770.371.63
Participation in TC
Nonparticipation in TCRefRef
Participation in TC1.110.671.850.780.551.12
Qualification for Medicaid suspension
UnqualifiedRefRef
Qualified0.880.531.461.030.721.47
Year of release
2015RefRef
20183.171.745.771.541.072.22
Qualification versus release in 20180.820.471.461.020.671.57

aMOUD, medications for opioid use disorder; TC, therapeutic community.

TABLE 3. Multivariable logistic regression examining the association between qualification of Medicaid suspension and post‐release crude and opioid overdose mortality risk, and other risk factors for mortalities within a year after prison release.a
Enlarge table

DISCUSSION

The risk of opioid overdose mortality after release from prison for adults with an OUD increased from 2015 to 2018, particularly from synthetic narcotics, and qualification for Medicaid suspension was not associated with a significant decrease in OOMR or crude mortality within 1 year after prison release for individuals diagnosed with OUD.

Risk of post‐release opioid overdose deaths for individuals with an OUD may be partially driven by the increased presence of fentanyl and other synthetic opioids in the illicit opioid market (44), as well as limited state harm reduction access (45, 46, 47). Consistent with prior studies (1, 6), our study found that drug overdoses have remained the leading cause of death for individuals released from prison and, furthermore, that individuals with a lower risk of post‐release overdose mortality are of black race (2, 6, 7, 8). In contrast with our findings, other studies have attributed pharmaceutical opioids (6) and heroin (2) as the primary opioids implicated in post‐release mortality from opioid overdoses for individuals released from SCI, whereas our findings indicate that synthetic opioids are the primary opioid implicated in post‐release opioid overdose mortalities for individuals with an OUD diagnosis. The change in opioids implicated likely mirrors the evolving pattern of the opioid epidemic, transitioning from prescription opioids to heroin and subsequently to fentanyl.

It is theorized that the mechanisms for which Medicaid suspension influence post‐release opioid overdose mortality is primarily through improvements in OUD treatment participation and treatment continuity (32, 35, 36, 37, 40), remarkably for treatments that involve agonist and partial agonist medications. These treatments generally lead to reduced illicit opioid use, especially when administered intravenously, thus reducing the likelihood of a fatal opioid overdose after release. The reduced economic strain associated with treatment access through Medicaid, as well as other Medicaid services (i.e. mental health treatment (31, 32) and treatments (30, 32, 38)), are likely to assist with mitigation of post‐release opioid overdose risk as well.

Further research is needed to advance understanding of each of these mechanisms involved in post‐release care and their impact on risk of opioid overdose mortality among those released from prison, particularly for more transient prison populations (6, 48). Since opioid‐related deaths constitute a small percentage of drug overdoses, it's possible the policy was significantly associated with decreases in non‐fatal overdoses after release or improvements in other post‐release outcomes like participation in different treatment services or medical services pertinent to the well‐being of individuals diagnosed with an OUD.

LIMITATIONS

To our knowledge, this is the first study to examine the impact of a state authorization to suspend Medicaid on opioid overdose mortality in the post‐release period for individuals with an OUD diagnosis; however, there are several limitations that deserve comment. The extent to which these findings can be generalized to other correctional facilities has not been established, particularly with the demonstrated differences in the demographic composition in Pennsylvania SCIs, in combination with the various state approaches to access and insure individuals with Medicaid.

OUD diagnoses were based on a screening instrument that was not confirmed independently, nor do we know the insurance status of released individuals or the departmental efficacies for dissemination of these state policies. The small sample size may have contributed to the failure to reject our null hypothesis, as it is possible that the study lacked sufficient power to detect it adequately, and the identification of opioid‐specific deaths was also limited to the death records provided to the NDI on death certificates. Misclassified or otherwise unknown etiologies, especially on specific drug involvements, may have underreported opioid‐related deaths. However, any misclassifications that may have occurred in the etiologies, should not vary between the groups. In addition, several highly weighted identifiers for data matching (i.e., social security number, state of birth, and state of residency) were not provided by PADOC, which limited probabilistic scores and may have further underreported opioid related deaths.

CLINICAL RELEVANCE

The findings underscore the need for ongoing research into overdose prevention approaches tailored for individuals within the criminal justice system, particularly those transitioning from incarceration. Although suspension policies hold promise, comprehensive studies that examine diverse outcomes are essential to determine the true effects of these policies for individuals released from prison. Further research is also crucial to understand the implications across diverse demographic cohorts, notably for individuals with documented concerns of opioid use, evaluate the suitability of implementing Medicaid suspension policies on a broader scale, and determine whether adjustments in implementation are necessary to enhance outcomes after release.

Penn State College of Medicine, Hershey, Pennsylvania, USA (L. Blumberger, W. Calo, G. Liu, D. L. Leslie); Penn State Cancer Institute, Hershey, Pennsylvania, USA (W. Calo); Penn State Harrisburg, Harrisburg, Pennsylvania, USA (D. J. Mallinson)
Send correspondence to Dr. Blumberger ()

The authors received no financial support for the research, authorship, and/or publication of this article, and the authors declare no conflicts of interest. The content of this manuscript is the authors' responsibility and does not necessarily represent the viewpoints of Pennsylvania State University. As this manuscript was created in collaboration with the Pennsylvania Department of Corrections and the Center for Disease Control and Prevention National Death Index, the submission of this manuscript is in compliance with the policies of each department on public dissemination of research findings, confidentiality, as well as other policies and procedures that govern the use of the data under the jurisdiction of the departments.

For individuals diagnosed with OUD who were released from Pennsylvania state prisons in 2015 and 2018, a significant decrease in risk of post‐release opioid overdose mortality was not observed following the implementation of Medicaid suspension. Additional investigation is necessary to further understand the impact of policies intended to enhance post‐release insurance continuity and their effects on individuals diagnosed with OUD after prison release.

REFERENCES

1 Binswanger IA, Stern MF, Deyo RA, Heagerty PJ, Cheadle A, Elmore JG, et al. Release from prison–a high risk of death for former inmates. N Engl J Med. 2007;356(2):157–165. https://doi.org/10.1056/nejmsa064115Google Scholar

2 Ranapurwala SI, Shanahan ME, Alexandridis AA, Proescholdbell SK, Naumann RB, Edwards D, et al. Opioid overdose mortality among former North Carolina inmates: 2000‐2015. Am J Publ Health. 2018;108(9):1207–1213. https://doi.org/10.2105/ajph.2018.304514Google Scholar

3 Rosen DL, Schoenbach VJ, Wohl DA. All‐cause and cause‐specific mortality among men released from state prison, 1980‐2005. Am J Publ Health. 2008;98(12):2278–2284. https://doi.org/10.2105/ajph.2007.121855Google Scholar

4 Friedman JR, Hansen H. Evaluation of increases in drug overdose mortality rates in the US by race and ethnicity before and during the COVID‐19 pandemic. JAMA Psychiatr. 2022;79(4):379–381. https://doi.org/10.1001/jamapsychiatry.2022.0004Google Scholar

5 Butelman ER, Huang Y, Epstein DH, Shaham Y, Goldstein RZ, Volkow ND, et al. Overdose mortality rates for opioids and stimulant drugs are substantially higher in men than in women: state‐level analysis. Neuropsychopharmacology. 2023;48(11):1639–1647. https://doi.org/10.1038/s41386‐023‐01601‐8Google Scholar

6 Binswanger IA, Blatchford PJ, Mueller SR, Stern MF. Mortality after prison release: opioid overdose and other causes of death, risk factors, and time trends from 1999 to 2009. Ann Intern Med. 2013;159(9):592–600. https://doi.org/10.7326/0003‐4819‐159‐9‐201311050‐00005Google Scholar

7 Pizzicato LN, Drake R, Domer‐Shank R, Johnson CC, Viner KM. Beyond the walls: risk factors for overdose mortality following release from the Philadelphia Department of Prisons. Drug Alcohol Depend. 2018;189:108–115. https://doi.org/10.1016/j.drugalcdep.2018.04.034Google Scholar

8 Binswanger IA, Stern MF, Yamashita TE, Mueller SR, Baggett TP, Blatchford PJ. Clinical risk factors for death after release from prison in Washington State: a nested case‐control study. Addiction. 2016;111(3):499–510. https://doi.org/10.1111/add.13200Google Scholar

9 Hartung DM, McCracken CM, Nguyen T, Kempany K, Waddell EN. Fatal and nonfatal opioid overdose risk following release from prison: a retrospective cohort study using linked administrative data. J Subst Use Addict Treat. 2023;147:208971. https://doi.org/10.1016/j.josat.2023.208971Google Scholar

10 O'Connor AW, Sears JM, Fulton‐Kehoe D. Overdose and substance‐related mortality after release from prison in Washington State: 2014‐2019. Drug Alcohol Depend. 2022;241:109655. https://doi.org/10.1016/j.drugalcdep.2022.109655Google Scholar

11 Brinkley‐Rubinstein L, Macmadu A, Marshall BDL, Heise A, Ranapurwala SI, Rich JD, et al. Risk of fentanyl‐involved overdose among those with past year incarceration: findings from a recent outbreak in 2014 and 2015. Drug Alcohol Depend. 2018;185:189–191. https://doi.org/10.1016/j.drugalcdep.2017.12.014Google Scholar

12 Wakeman SE, Bowman SE, McKenzie M, Jeronimo A, Rich JD. Preventing death among the recently incarcerated: an argument for naloxone prescription before release. J Addict Dis. 2009;28(2):124–129. https://doi.org/10.1080/10550880902772423Google Scholar

13 Mital S, Wolff J, Carroll JJ. The relationship between incarceration history and overdose in North America: a scoping review of the evidence. Drug Alcohol Depend. 2020;213:108088. https://doi.org/10.1016/j.drugalcdep.2020.108088Google Scholar

14 Bird SM, Hutchinson SJ. Male drugs‐related deaths in the fortnight after release from prison: Scotland, 1996‐99. Addiction. 2003;98(2):185–190. https://doi.org/10.1046/j.1360‐0443.2003.00264.xGoogle Scholar

15 Merrall EL, Kariminia A, Binswanger IA, Hobbs MS, Farrell M, Marsden J, et al. Meta‐analysis of drug‐related deaths soon after release from prison. Addiction. 2010;105(9):1545–1554. https://doi.org/10.1111/j.1360‐0443.2010.02990.xGoogle Scholar

16 Hagan H, Campbell JV, Thiede H, Strathdee SA, Ouellet L, Latka M, et al. Injecting alone among young adult IDUs in five US cities: evidence of low rates of injection risk behavior. Drug Alcohol Depend. 2007;91(Suppl 1):S48–S55. https://doi.org/10.1016/j.drugalcdep.2007.02.002Google Scholar

17 El‐Bassel N, Marotta PL, Goddard‐Eckrich D, Chang M, Hunt T, Wu E, et al. Drug overdose among women in intimate relationships: the role of partner violence, adversity and relationship dependencies. PLoS One. 2019;14(12):e0225854. https://doi.org/10.1371/journal.pone.0225854Google Scholar

18 Macmadu A, Frueh L, Collins AB, Newman R, Barnett NP, Rich JD, et al. Drug use behaviors, trauma, and emotional affect following the overdose of a social network member: a qualitative investigation. Int J Drug Pol. 2022;107:103792. https://doi.org/10.1016/j.drugpo.2022.103792Google Scholar

19 Carroll JJ, Marshall BDL, Rich JD, Green TC. Exposure to fentanyl‐contaminated heroin and overdose risk among illicit opioid users in Rhode Island: a mixed methods study. Int J Drug Pol. 2017;46:136–145. https://doi.org/10.1016/j.drugpo.2017.05.023Google Scholar

20 Brinkley‐Rubinstein L, Sivaraman J, Rosen DL, Cloud DH, Junker G, Proescholdbell S, et al. Association of restrictive housing during incarceration with mortality after release. JAMA Netw Open. 2019;2(10):e1912516. https://doi.org/10.1001/jamanetworkopen.2019.12516Google Scholar

21 Binswanger IA, Nguyen AP, Morenoff JD, Xu S, Harding DJ. The association of criminal justice supervision setting with overdose mortality: a longitudinal cohort study. Addiction. 2020;115(12):2329–2338. https://doi.org/10.1111/add.15077Google Scholar

22 Sordo L, Barrio G, Bravo MJ, Indave BI, Degenhardt L, Wiessing L, et al. Mortality risk during and after opioid substitution treatment: systematic review and meta‐analysis of cohort studies. BMJ. 2017;357:j1550. https://doi.org/10.1136/bmj.j1550Google Scholar

23 Lim S, Cherian T, Katyal M, Goldfeld KS, McDonald R, Wiewel E, et al. 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. 2023;118(3):459–467. https://doi.org/10.1111/add.16071Google Scholar

24 Green TC, Clarke J, Brinkley‐Rubinstein L, Marshall BDL, Alexander‐Scott N, Boss R, et al. Postincarceration fatal overdoses after implementing medications for addiction treatment in a statewide correctional system. JAMA Psychiatr. 2018;75(4):405–407. https://doi.org/10.1001/jamapsychiatry.2017.4614Google Scholar

25 Nunn A, Zaller N, Dickman S, Trimbur C, Nijhawan A, Rich JD. Methadone and buprenorphine prescribing and referral practices in US prison systems: results from a nationwide survey. Drug Alcohol Depend. 2009;105(1‐2):83–88. https://doi.org/10.1016/j.drugalcdep.2009.06.015Google Scholar

26 Scott CK, Dennis ML, Grella CE, Mischel AF, Carnevale J. The impact of the opioid crisis on U.S. state prison systems. Health Justice. 2021;9(1):17. https://doi.org/10.1186/s40352‐021‐00143‐9Google Scholar

27 Klemperer EM, Wreschnig L, Crocker A, King‐Mohr J, Ramniceanu A, Brooklyn JR, et al. The impact of the implementation of medication for opioid use disorder and COVID‐19 in a statewide correctional system on treatment engagement, postrelease continuation of care, and overdose. J Subst Use Addict Treat. 2023;152:209103. https://doi.org/10.1016/j.josat.2023.209103Google Scholar

28 Medicaid and CHIP Payment and Access Commission (MACPAC) . Medicaid and the Criminal Justice System. 2018.Google Scholar

29 Burns ME, Cook ST, Brown L, Tyska S, Westergaard RP. Increasing Medicaid enrollment among formerly incarcerated adults. Health Serv Res. 2021;56(4):643–654. https://doi.org/10.1111/1475‐6773.13634Google Scholar

30 Burns ME, Cook S, Brown LM, Dague L, Tyska S, Hernandez RomeroK, et al. Association between assistance with Medicaid enrollment and use of health care after incarceration among adults with a history of substance use. JAMA Netw Open. 2022;5(1):e2142688. https://doi.org/10.1001/jamanetworkopen.2021.42688Google Scholar

31 Cuddeback GS, Morrissey JP, Domino ME. Enrollment and service use patterns among persons with severe mental illness receiving expedited Medicaid on release from state prisons, county jails, and psychiatric hospitals. Psychiatr Serv. 2016;67(8):835–841. https://doi.org/10.1176/appi.ps.201500306.2016.1.testGoogle Scholar

32 Morrissey JP, Domino ME, Cuddeback GS. Expedited Medicaid enrollment, mental health service use, and criminal recidivism among released prisoners with severe mental illness. Psychiatr Serv. 2016;67(8):842–849. https://doi.org/10.1176/appi.ps.201500305Google Scholar

33 Grabert BK, Gertner AK, Domino ME, Cuddeback GS, Morrissey JP. Expedited Medicaid enrollment, service use, and recidivism at 36 months among released prisoners with severe mental illness. Psychiatr Serv. 2017;68(10):1079–1082. https://doi.org/10.1176/appi.ps.201600482Google Scholar

34 Blackburn J, Norwood C, Rusyniak D, Gilbert AL, Sullivan J, Menachemi N. Indiana's Section 1115 Medicaid waiver and interagency coordination improve enrollment for justice‐involved adults. Health Aff. 2020;39(11):1891–1899. https://doi.org/10.1377/hlthaff.2019.01823Google Scholar

35 Khatri UG, Howell BA, Winkelman TNA. Medicaid expansion increased medications for opioid use disorder among adults referred by criminal justice agencies. Health Aff. 2021;40(4):562–570. https://doi.org/10.1377/hlthaff.2020.01251Google Scholar

36 Winkelman TNA, Ford BR, Shlafer RJ, McWilliams A, Admon LK, Patrick SW. Medications for opioid use disorder among pregnant women referred by criminal justice agencies before and after Medicaid expansion: a retrospective study of admissions to treatment centers in the United States. PLoS Med. 2020;17(5):e1003119. https://doi.org/10.1371/journal.pmed.1003119Google Scholar

37 Gertner AK, Grabert B, Domino ME, Cuddeback GS, Morrissey JP. The effect of referral to expedited Medicaid on substance use treatment utilization among people with serious mental illness released from prison. J Subst Abuse Treat. 2019;99:9–15. https://doi.org/10.1016/j.jsat.2019.01.010Google Scholar

38 Wennerstrom A, Sugarman OK, Reilly B, Armstrong A, Whittington A, Bachhuber MA. Health services use among formerly incarcerated Louisiana Medicaid members within one year of release. PLoS One. 2023;18(5):e0285582. https://doi.org/10.1371/journal.pone.0285582Google Scholar

39 McLeod KE, Karim ME, Buxton JA, Martin RE, Scow M, Felicella G, et al. Use of community healthcare and overdose in the 30 days following release from provincial correctional facilities in British Columbia. Drug Alcohol Depend. 2021;229(Pt A):109113. https://doi.org/10.1016/j.drugalcdep.2021.109113Google Scholar

40 Fry CE, McGuire TG, Frank RG. Medicaid expansion's spillover to the criminal justice system: evidence from six urban counties. RSF. 2020;6(2):244–263. https://doi.org/10.7758/rsf.2020.6.2.11Google Scholar

41 Boslett AJ, Denham A, Hill EL. Using contributing causes of death improves prediction of opioid involvement in unclassified drug overdoses in US death records. Addiction. 2020;115(7):1308–1317. https://doi.org/10.1111/add.14943Google Scholar

42 Ruhm CJ. Nonopioid overdose death rates rose almost as fast as those involving opioids, 1999‐2016. Health Aff. 2019;38(7):1216–1224. https://doi.org/10.1377/hlthaff.2018.05522Google Scholar

43 Schlaffer M. Fact sheet: drug‐related crime. (163928). Rockville: NCJRS Photocopy Services; 1997. https://www.ojp.gov/ncjrs/virtual‐library/abstracts/fact‐sheet‐drug‐related‐crime‐0Google Scholar

44 Fogarty MF, Papsun DM, Logan BK. Analysis of fentanyl and 18 novel fentanyl analogs and metabolites by LC‐MS‐MS, and report of fatalities associated with methoxyacetylfentanyl and cyclopropylfentanyl. J Anal Toxicol. 2018;42(9):592–604. https://doi.org/10.1093/jat/bky035Google Scholar

45 Townsend T, Blostein F, Doan T, Madson‐Olson S, Galecki P, Hutton DW. Cost‐effectiveness analysis of alternative naloxone distribution strategies: first responder and lay distribution in the United States. Int J Drug Pol. 2020;75:102536. https://doi.org/10.1016/j.drugpo.2019.07.031Google Scholar

46 Levengood TW, Yoon GH, Davoust MJ, Ogden SN, Marshall BDL, Cahill SR, et al. Supervised injection facilities as harm reduction: a systematic review. Am J Prev Med. 2021;61(5):738–749. https://doi.org/10.1016/j.amepre.2021.04.017Google Scholar

47 Marshall BD, Milloy MJ, Wood E, Montaner JS, Kerr T. Reduction in overdose mortality after the opening of North America's first medically supervised safer injecting facility: a retrospective population‐based study. Lancet. 2011;377(9775):1429–1437. https://doi.org/10.1016/s0140‐6736(10)62353‐7Google Scholar

48 Binswanger IA, Blatchford PJ, Lindsay RG, Stern MF. Risk factors for all‐cause, overdose and early deaths after release from prison in Washington State. Drug Alcohol Depend. 2011;117(1):1–6. https://doi.org/10.1016/j.drugalcdep.2010.11.029Google Scholar