Skip to main content
NCBI home page
JAMA Network logoLink to JAMA Network
. 2024 Jul 24;7(7):e2421740. doi: 10.1001/jamanetworkopen.2024.21740

Medication for Opioid Use Disorder After Serious Injection-Related Infections in Massachusetts

Simeon D Kimmel 1,2,✉, Alexander Y Walley 1, Laura F White 3, Shapei Yan 1, Christine Grella 4,5, Adam Majeski 1, Michael D Stein 6, Amy Bettano 7, Dana Bernson 7, Mari-Lynn Drainoni 2,6,8, Jeffrey H Samet 1, Marc R Larochelle 1
PMCID: PMC11270137  PMID: 39046742

Key Points

Question

After serious injection-related infections (SIRIs), what characteristics are associated with receipt of medication for opioid use disorder (MOUD)?

Findings

In this cohort study of 9757 individuals who survived a SIRI hospitalization in Massachusetts, 47.7% received MOUD in the 12 months after discharge. Treatment with MOUD before the SIRI hospitalization was associated with both receipt of any MOUD treatment and treatment rates in the 12 months after a SIRI hospitalization.

Meaning

In this study, MOUD was underused and prior MOUD treatment was associated with subsequent treatment after SIRI hospitalization.

This cohort study of individuals who survived a serious injection-related infection hospitalization in Massachusetts examines treatment with medications for opioid use disorder and its association with demographic and clinical characteristics.

Abstract

Importance

Serious injection-related infections (SIRIs) cause significant morbidity and mortality. Medication for opioid use disorder (MOUD) improves outcomes but is underused. Understanding MOUD treatment after SIRIs could inform interventions to close this gap.

Objectives

To examine rehospitalization, death rates, and MOUD receipt for individuals with SIRIs and to assess characteristics associated with MOUD receipt.

Design, Setting, and Participants

This retrospective cohort study used the Massachusetts Public Health Data Warehouse, which included all individuals with a claim in the All-Payer Claims Database and is linked to individual-level data from multiple government agencies, to assess individuals aged 18 to 64 years with opioid use disorder and hospitalization for endocarditis, osteomyelitis, epidural abscess, septic arthritis, or bloodstream infection (ie, SIRI) between July 1, 2014, and December 31, 2019. Data analysis was performed from November 2021 to May 2023.

Exposure

Demographic and clinical factors potentially associated with posthospitalization MOUD receipt.

Main Outcomes and Measures

The main outcome was MOUD receipt measured weekly in the 12 months after hospitalization. We used zero-inflated negative binomial regression to examine characteristics associated with any MOUD receipt and rates of treatment in the 12 months after hospitalization. Secondary outcomes were receipt of any buprenorphine formulation, methadone, and extended-release naltrexone examined individually.

Results

Among 9757 individuals (mean [SD] age, 43.3 [12.1] years; 5701 [58.4%] male) who survived a SIRI hospitalization, 4652 (47.7%) received MOUD, 6518 (66.8%) were rehospitalized, and 1068 (10.9%) died within 12 months. Of those treated with MOUD in the 12 months after hospitalization, the mean (SD) number of MOUD initiations during follow-up was 3.0 (1.8), with 985 of 4652 individuals (21.2%) receiving treatment at least 80% of the time. MOUD treatment after SIRI hospitalization was significantly associated with MOUD in the prior 6 months (buprenorphine: adjusted odds ratio [AOR], 16.63; 95% CI, 14.04-19.71; methadone: AOR, 28.61; 95% CI, 22.80-35.90; or naltrexone: AOR, 2.05; 95% CI, 1.58-2.66). Prior buprenorphine (incident rate ratio [IRR], 1.16; 95% CI, 1.09-1.22) or methadone (IRR, 1.90; 95% CI, 1.79-2.02) use was associated with higher treatment rates after hospitalization, and prior naltrexone use (IRR, 0.85; 95% CI, 0.77-0.95) was associated with lower rates.

Conclusions and Relevance

This study found that in the year after a SIRI hospitalization in Massachusetts, mortality and rehospitalization were common, and only half of patients received MOUD. Treatment with MOUD before a SIRI was associated with posthospitalization MOUD initiation and time receiving MOUD. Efforts are needed to initiate MOUD treatment during SIRI hospitalizations and subsequently retain patients in treatment.

Introduction

Serious injection-related infections (SIRIs), including endocarditis, osteomyelitis, epidural abscess, septic arthritis, and bloodstream infections, have more than doubled in the past 2 decades, causing morbidity and mortality for people who inject drugs.1,2,3 Most SIRIs are associated with opioid use disorder (OUD), despite regional variation.4,5,6 One-year mortality rates are elevated after SIRIs, reaching nearly 30% for endocarditis.2,7,8,9,10,11,12,13 Typically, SIRIs are treated with 2 or more weeks of intravenous antibiotics in hospitals, at post–acute care facilities, or at home, costing the health system more than $700 million annually.14,15

Although individuals with SIRIs seek care for their infections, hospitalizations are opportunities to initiate medications for opioid use disorder (MOUD).16,17,18,19 Methadone and buprenorphine, 2 types of MOUD, are associated with decreased opioid overdose and all-cause mortality.20 After SIRIs, MOUD treatment is associated with decreased rehospitalizations and mortality7,13,20,21,22,23,24 but are underused and with benefits attenuated by poor retention.7,21,25 Thus, there is an urgent need to understand how to improve MOUD initiation and retention after hospital discharge.

With an OUD prevalence of nearly 5%, one of the highest opioid-related hospitalization rates in the country, and a tradition of MOUD delivery,26,27,28,29,30 Massachusetts is a suitable site to investigate MOUD treatment after SIRIs.31 We used the Massachusetts Public Health Data Warehouse (PHD) to identify individuals with OUD hospitalized for SIRIs and examined MOUD receipt and treatment rates in the 12 months after discharge.32 We sought to describe MOUD treatment and its association with demographic and clinical characteristics to inform efforts to improve care. Identifying disparities by race, sex, housing status, and MOUD history or disposition, for example, may shape interventions to improve MOUD uptake and engagement.

Methods

Study Design and Data Source

This retrospective cohort study used the Massachusetts PHD, which included all Massachusetts residents with a claim in the All-Payer Claims Database (APCD).33 The APCD records were linked longitudinally at the individual level to 23 government datasets, 11 of which were used in this study. Studies that use the PHD have examined OUD prevalence and MOUD administration after overdose and endocarditis, among other conditions.7,31,34,35 Individual counts smaller than 10 are suppressed for privacy. The Boston Medical Center/Boston University Medical Campus Institutional Review Board determined this study was not human subjects research and therefore waived the need for study approval and informed consent. We followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines.36

Cohort Selection

At the time of this analysis, the PHD included data from January 1, 2014, to December 31, 2020, for the 11 datasets used. The cohort included individuals aged 18 to 64 years with OUD at the time of hospitalization for a serious infection (ie, SIRI) between July 1, 2014, and December 31, 2019, allowing 6 months before the index hospitalization for covariate identification and 12 months of MOUD-receipt assessment (eFigure 1 in Supplement 1). We excluded individuals older than 64 years to limit potential misclassification of infections due to noninjection factors (eg, dialysis and valve disease) and individuals younger than 18 because they receive MOUD and inject opioids less frequently.37,38

First, we identified discharges with a serious infection diagnosis, including endocarditis, osteomyelitis, septic arthritis, epidural abscess, or bloodstream infection (eg, bacteremia or fungemia), using International Classification of Diseases, Ninth Revision (ICD-9) and International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) (eTable 2 in Supplement 1). Second, we identified individuals whose infections were likely related to opioid injection based on record of OUD in the 6 months before hospitalization based on OUD billing codes, methadone or buprenorphine treatment, admission to a Bureau of Substance Addiction Services (BSAS) treatment facility, or an opioid-related ambulance trip. Buprenorphine, including long-acting formulations, was identified from Prescription Monitoring Program data or procedure codes. Methadone was identified from the BSAS data or procedure codes (eTables 3 and 4 in Supplement 1).

The first qualifying SIRI hospitalization during the study period was defined as the index hospitalization and triggered cohort inclusion. Because the focus was posthospitalization MOUD treatment, we excluded individuals who died during the hospitalization or were discharged to hospice (eFigure 2 in Supplement 1). For those with multiple qualifying hospitalizations, we examined MOUD treatment after the first to preserve individual-level analyses.

Main Outcomes and Measures

The main outcome was MOUD receipt measured weekly in the year after discharge for a SIRI hospitalization. Any week with at least 1 day of treatment with methadone, buprenorphine, or extended-release naltrexone was considered a week with treatment. Buprenorphine and methadone receipt were identified as described above. We identified extended-release naltrexone from pharmacy or procedure codes in the APCD. We used days’ supply and dispensing data for prescriptions (including long-acting buprenorphine) and date of claim plus 27 days for extended-release naltrexone to classify weeks receiving MOUD. Because the PHD does not include MOUD from inpatient or correctional pharmacies, receipt during hospitalizations or during incarceration could not be examined, and these weeks were excluded.

For secondary analyses, we examined weekly receipt of methadone, buprenorphine, and extended-release naltrexone as individual outcomes in separate models. We also reported the number of MOUD initiations, defined as MOUD receipt during the follow-up period after at least 2 eligible weeks without any MOUD, excluding incarceration or rehospitalization.

For descriptive purposes, we also report on rehospitalization, incarceration, and all-cause and opioid-related mortality from death records. As with previous research in the PHD, opioid-related deaths were determined by the medical examiner or standardized assessment by the Massachusetts Department of Public Health.35

Exposures and Confounders

To describe the cohort and examine associations with MOUD receipt, we included multiple demographic and clinical characteristics. We identified age and insurance status at time of index hospitalization from the APCD. To examine racial disparities in MOUD treatment after SIRIs, data on race and ethnicity were obtained from the APCD and corroborated across multiple data sources in the PHD using a tiered method. We identified anxiety, depression, alcohol use disorder, and stimulant use disorder using ICD-9 or ICD-10 codes in the APCD (eTable 4 in Supplement 1) or admission to a BSAS-licensed facility for the relevant disorder. We calculated a modified Elixhauser Comorbidity Index using the presence of 31 comorbidities identified with ICD-9 or ICD-10 codes during the index hospitalization or in the prior 6 months.39 Because alcohol use, drug use, anxiety, and depression were separate variables in this study, these were excluded from the total score. We identified receipt of the opioid analgesic benzodiazepine in the 6 months before a SIRI discharge using Prescription Monitoring Program data and receipt of a naloxone prescription using pharmacy records. We identified methadone, buprenorphine, and naltrexone (extended-release or oral) in the 6 months before hospitalization as described earlier. We identified homelessness in the 6 months before discharge for the SIRI using multiple sources in the PHD (eTable 6 in Supplement 1). Additionally, we examined hospitalization length and discharge disposition (eg, routine, skilled nursing facility [SNF], or rehabilitation facility) and infection type.

Statistical Analysis

We used χ2 tests to compare baseline cohort characteristics by MOUD receipt in the follow-up period. We calculated the proportion receiving MOUD for weeks leading up to and after the index hospitalization and repeated analyses stratified by discharge home and to a SNF or rehabilitation facility. We calculated the proportion of weeks with MOUD in the 1-year follow-up period, excluding weeks with acute care hospitalization, with incarceration, or after death. We used the Cramer V test to examine correlations of covariates and created a composite variable for anxiety and/or depression based on correlation greater than 0.4. Zero-inflated negative binomial regression is a 2-part model for use in the presence of excess zeros. Because no MOUD receipt was common, we used this approach to examine factors associated with (1) receipt of any MOUD (model 1) and (2) weeks of MOUD receipt in the follow-up period (model 2). In model 2, we included an offset term for number of weeks in which MOUD receipt could be observed (ie, excluding weeks incarcerated, hospitalized, or after death) to model MOUD treatment rates (eFigure 3 in Supplement 1). For secondary analyses, we used the same approach to examine receipt of methadone, buprenorphine, or extended-release naltrexone as individual outcomes. Two-sided P < .05 was considered statistically significant. We performed analyses using SAS Studio, version 3.81 (SAS Institute Inc). Data analysis was performed from November 2021 to May 2023.

Results

We identified 65 723 hospitalizations for serious infections, 19 540 (29.7%) of which included record of OUD during the hospitalization or in the 6 months prior. We excluded 781 hospitalizations resulting in death or hospice discharge. As the first qualifying hospitalization triggered cohort inclusion, 9002 subsequent SIRI hospitalizations were excluded. The resulting study cohort included 9757 individuals (mean [SD] age, 43.3 [12.1] years; 5701 [58.4%] male and 4056 [41.6%] female; 55 [0.6%] American Indian, other non-Hispanic race [no additional description available from the database], or unknown; 39 [0.4%] non-Hispanic Asian or Pacific Islander; 807 [8.3%] non-Hispanic Black; 1239 (12.7) Hispanic; and 7617 [78.1%] non-Hispanic White) with an index SIRI hospitalization (eFigure 2 in Supplement 1). Of these individuals, during 12 months of potential follow-up, 6518 (66.8%) were rehospitalized, 459 (4.7%) were incarcerated in prison, 205 (2.1%) died of opioid overdose, and 1068 (10.9%) died of any cause. Within 3 months of discharge, 4654 (47.7%) were rehospitalized, 166 (1.7%) were incarcerated, and 378 (3.9%) died, including 49 (0.5%) of an opioid overdose.

MOUD Receipt Before and After SIRI Hospitalization

In the 12 months after an index SIRI hospitalization, 4652 of 9757 individuals (47.7%) received any MOUD. Of those who received MOUD, the mean (SD) number of MOUD initiations during follow-up was 3.0 (1.8). In the week before hospitalization, of 9116 individuals, 1735 (19.0%) were treated with MOUD: 722 (7.9%) with buprenorphine, 989 (10.8%) with methadone, and 24 (0.3%) with extended-release naltrexone. Three months after discharge, of 8750 individuals, 2048 (23.4%) were treated with MOUD, with buprenorphine treatment increasing to 835 (9.5%), methadone to 1177 (13.5%), and extended-release naltrexone to 36 (0.4%). At 12 months after discharge, of 8275 individuals, 2015 (24.4%) were treated with MOUD, with buprenorphine treatment decreasing to 750 (9.1%), methadone treatment increasing to 1234 (14.9%), and extended-release naltrexone treatment stable at 31 (0.4%) (Figure 1). Results were similar when stratified by discharge to home or an SNF (eFigure 4 in Supplement 1). Among the 4652 who received MOUD after SIRI hospitalization, 1894 (40.7%) had more than 50% of follow-up while receiving MOUD treatment and 985 (21.2%) had more than 80% of follow-up (Figure 2).

Figure 1. Weekly Methadone, Buprenorphine, and Naltrexone Receipt 6 Months Before and 12 Months After Serious Injection-Related Infection (SIRI) Between January 1, 2014, and December 31, 2019, in Massachusetts.

Figure 1.

Open in a new tab

Index hospitalization occurred between July 1, 2014, and December 31, 2019. MOUD indicates medication for opioid use disorder.

aSuppressed due to count less than 10.

Figure 2. Percentage of Weeks With Receipt of Medication for Opioid Use Disorder (MOUD) After Serious Injection-Related Infection Between July 1, 2014, and December 31, 2019, in Massachusetts.

Figure 2.

Open in a new tab

Percentage of weeks receiving MOUD treatment was calculated using total weeks of MOUD in the numerator and weeks during which MOUD receipt was observable in our data as the denominator. Weeks while hospitalized, during incarceration, and after death were excluded.

Participant Characteristics by MOUD Receipt

As indicated in Table 1, those who received MOUD in the 12 months after infection, compared with those not treated with MOUD, were significantly more likely to have received treatment with buprenorphine (1885 [40.5%] vs 349 [6.8%], P < .001), methadone (1386 [29.8%] vs 178 [3.5%], P < .001), or naltrexone (273 [5.9%] vs 156 [3.1%], P < .001) in the 6 months before the SIRI. Additionally, they were more likely to have filled a naloxone prescription (710 [15.3%] vs 316 [6.2%], P < .001), more likely have experienced homelessness (1267 [27.2%] vs 749 [14.7%], P < .001) or an opioid overdose (690 [14.8%] vs 655 [12.8%], P = .004), and less likely to be prescribed opioids (1550 [33.3%] vs 3133 [61.4%], P < .001) or benzodiazepines (1256 [27.0%] vs 1709 [33.5%], P < .001) in the 6 months before SIRI. Those who received MOUD were more likely young (aged 18-34 years: 1772 [38.1%] vs 1132 [22.2%], P < .001), less likely non-Hispanic Black (278 [6.0%] vs 529 [10.4%], P < .001), more likely to have Medicaid (3114 [66.9%] vs 2421 [47.4%], P < .001), and less likely to have commercial (527 [11.3%] vs 918 [18.0%]) or Medicare insurance (828 [17.8%] vs 1559 [30.5%]) (P < .001 for both). They were more likely to have anxiety or depression (3370 [72.4%] vs 3233 [63.3%], P < .001), more likely to have alcohol (1594 [34.3%] vs 1536 [30.1%], P < .001) or stimulant use disorder (1111 [23.9%] vs 660 [12.9%], P < .001), and less likely to have high Elixhauser Comorbidity Index scores (≥3) (2169 [46.6%] vs 3359 [65.8%], P < .001).

Table 1. Characteristics of Individuals With and Without MOUD Receipt After Serious Injection-Related Infections in Massachusetts, July 1, 2014, to December 31, 2019a.

Characteristic Total (N = 9757) Any MOUD in 12-mo follow-up (n = 4652) No MOUD in 12-mo follow-up (n = 5105) P value
Age group, y
18-34 2904 (29.8) 1772 (38.1) 1132 (22.2) <.001
35-49 3339 (34.2) 1791 (38.5) 1548 (30.3)
50-64 3514 (36.0) 1089 (23.4) 2425 (47.5)
Sex
Male 5701 (58.4) 2644 (56.8) 3057 (59.9) .002
Female 4056 (41.6) 2008 (43.2) 2048 (40.1)
Race and ethnicity
American Indian, other,b or unknown 55 (0.6) Data suppressed 39 (0.8) <.001
Asian or Pacific Islander, non-Hispanic 39 (0.4) Data suppressed 31 (0.6)
Black, non-Hispanic 807 (8.3) 278 (6.0) 529 (10.4)
Hispanic 1239 (12.7) 620 (13.3) 619 (12.1)
White, non-Hispanic 7617 (78.1) 3730 (80.2) 3887 (76.1)
Insurance
Commercial 1445 (14.8) 527 (11.3) 918 (18.0) <.001
Medicaid 5535 (56.7) 3114 (66.9) 2421 (47.4)
Medicare 2387 (24.5) 828 (17.8) 1559 (30.5)
Self-pay 248 (24.5) 137 (2.9) 111 (1.8)
Other or unknown 142 (1.5) 46 (1.0) 96 (1.9)
Homelessnessc 2016 (20.7) 1267 (27.2) 749 (14.7) <.001
Anxiety or depressionc 6063 (67.7) 3370 (72.4) 3233 (63.3) <.001
Alcohol use disorderc 3130 (32.1) 1594 (34.3) 1536 (30.1) <.001
Stimulant use disorderc 1771 (18.2) 1111 (23.9) 660 (12.9) <.001
Opioid prescription, priorc 4683 (48.0) 1550 (33.3) 3133 (61.4) <.001
Benzodiazepine prescription, priorc 2965 (30.4) 1256 (27.0) 1709 (33.5) <.001
Naloxone prescription, priorc 1026 (10.5) 710 (15.3) 316 (6.2) <.001
Modified Elixhauser Comorbidity Index scored
0 1044 (10.7) 638 (13.7) 406 (8.0) <.001
1 1573 (16.1) 931 (20.0) 642 (12.6)
2 1612 (16.5) 914 (19.7) 698 (13.7)
≥3 5528 (56.7) 2169 (46.6) 3359 (65.80
MOUD, priorc
Buprenorphine 2234 (22.9) 1885 (40.5) 349 (6.8) <.001
Methadone 1564 (16.0) 1386 (29.8) 178 (3.5) <.001
Naltrexone (oral or extended release) 429 (4.4) 273 (5.9) 156 (3.1) <.001
Opioid overdosec 1345 (13.8) 690 (14.8) 655 (12.8) .004
Length of hospitalization, median (IQR), d 7 (4-11) 7 (4-11) 7 (4-13) <.001
Disposition site
Home or routine discharge 2566 (26.3) 1358 (29.2) 1208 (23.7) <.001
Home with services 1708 (17.5) 567 (12.2) 1141 (22.4)
Transferred, acute care hospital 866 (8.9) 436 (9.4) 430 (8.4)
SNF or rehabilitation facility 2884 (29.6) 1290 (27.7) 1594 (31.2)
Patient-directed discharge 1333 (13.7) 774 (16.6) 559 (11.0)
Other 400 (4.1) 227 (4.90 173 (3.4)
Infection type
Endocarditis 1707 (17.5) 1037 (22.3) 670 (13.1) <.001
Epidural abscess 688 (7.1) 342 (7.4) 346 (6.8)
Septic arthritis 1046 (10.7) 593 (12.8) 453 (8.9)
Osteomyelitis 1932 (19.8) 761 (16.4) 1171 (22.9)
Bloodstream infection 4384 (44.9) 1919 (41.3) 2465 (48.3)
Open in a new tab

Abbreviations: MOUD, medication for opioid use disorder; SIRI, serious injection-related infection; SNF, skilled nursing facility.

a

Data are presented as number (percentage) of patients unless otherwise indicated.

b

No additional description available from the database.

c

Homeless status; anxiety and depression; alcohol use disorder; stimulant use disorder; prior prescriptions for opioids, benzodiazepine, and naloxone; prior MOUD; and opioid overdose were assessed in the 6 months before SIRI hospitalization. All other characteristics were ascertained at time of hospitalization.

d

Modified Elixhauser Comorbidity Index score excludes alcohol, drug use, and anxiety or depression, as these characteristics were examined separately.

Zero-Inflated Negative Binomial Model Results: Any MOUD Receipt

In adjusted models, several characteristics were significantly associated with receipt of any MOUD treatment in the year after discharge for a SIRI (Table 2): having received buprenorphine (adjusted odds ratio [AOR], 16.63; 95% CI, 14.04-19.71), methadone (AOR, 28.61; 95% CI, 22.80-35.90), or naltrexone (AOR, 2.05; 95% CI, 1.58-2.66) during the 6 months before SIRI; endocarditis (AOR, 1.83; 95% CI, 1.55-2.15) or septic arthritis (AOR, 1.80; 95% CI, 1.48-2.18) compared with a bloodstream infection; having anxiety or depression (AOR, 1.39; 95% CI, 1.22-1.58); having filled a naloxone prescription (AOR, 1.43; 95% CI, 1.17-1.76); experiencing homelessness (AOR, 1.46; 95% CI, 1.25-1.69); or having Medicaid (AOR, 1.56; 95% CI, 1.32-1.85). Characteristics negatively associated with any MOUD receipt included the following: age of 50 to 64 years compared with 18 to 34 years (AOR, 0.51; 95% CI, 0.43-0.60); filling an opioid prescription in the 6 months before the SIRI (AOR, 0.56; 95% CI, 0.49-0.64); non-Hispanic Black race (AOR, 0.72; 95% CI: 0.59-0.89); having an Elixhauser Comorbidity Index score of 3 or more (AOR, 0.62; 95% CI, 0.51-0.76); and discharge home with services (AOR, 0.65; 95% CI, 0.54-0.78), discharge to an SNF or rehabilitation facility (AOR, 0.79; 95% CI, 0.67-0.92), or patient-directed discharge (AOR, 0.76; 95% CI, 0.63-0.92) compared with routine discharge.

Table 2. Zero-Inflated Negative Binomial Model Results: Association of Characteristics With Any MOUD Receipt and Weeks Treated With MOUD Among Individuals With Serious Injection-Related Infections, Massachusetts, July 1, 2014, to December 31, 2019a.

Characteristic Any MOUD (N = 9757) Total weeks receiving MOUD (n = 4652)
AOR (95% CI) P value IRR (95% CI) P value
Age group, y
18-34 1 [Reference] NA 1 [Reference] NA
35-49 0.87 (0.75-1.00) .05 1.09 (1.03-1.15) .004
50-64 0.51 (0.43-0.60) <.001 1.06 (0.99-1.11) .12
Sex
Female 1.09 (0.96-1.22) .17 1.01 (0.96-1.06) .81
Male 1 [Reference] NA 1 [Reference] NA
Race and ethnicity
American Indian, other,b or unknown 0.44 (0.24-0.80) .01 1.01 (0.72-1.41) .96
Black non-Hispanic 0.72 (0.59-0.89) .002 0.94 (0.84-1.04) .23
Hispanic 1.01 (0.85-1.19) .90 0.99 (0.92-1.06) .75
White non-Hispanic 1 [Reference] NA 1 [Reference] NA
Insurance
Commercial 1 [Reference] NA 1 [Reference] NA
Medicaid 1.56 (1.32-1.85) <.001 1.11 (1.02-1.20) .01
Medicare 0.98 (0.81-1.19) .82 1.04 (0.95-1.14) .40
Self-pay 1.15 (0.78-1.68) .48 1.15 (0.98-1.34) .09
Other or unknown 0.70 (0.43-1.13) .14 0.84 (0.65-1.08) .18
Homelessnessc 1.46 (1.25-1.69) <.001 0.94 (0.89-1.00) .06
Anxiety or depressionc 1.39 (1.22-1.58) <.001 1.05 (0.99-1.11) .10
Alcohol use disorderc 0.92 (0.81-1.05) .21 0.96 (0.91-1.01) .14
Stimulant use disorderc 1.06 (0.9-1.24) .49 0.97 (0.91-1.03) .29
Opioid prescription, priorc 0.56 (0.49-0.64) <.001 0.90 (0.86-0.95) <.001
Benzodiazepine prescription, priorc 0.88 (0.77-1.01) .08 1.03 (0.98-1.10) .25
Naloxone prescription, priorc 1.43 (1.17-1.76) <.001 1.02 (0.95-1.10) .50
Modified Elixhauser Comorbidity Index scored
0 1 [Reference] NA 1 [Reference] NA
1 0.98 (0.79-1.21) .85 1.07 (0.98-1.17) .11
2 0.87 (0.70-1.09) .22 1.10 (1.01-1.20) .03
≥3 0.62 (0.51-0.76) <.001 1.02 (0.94-1.10) .64
MOUD, priorc
Buprenorphine 16.63 (14.04-19.71) <.001 1.16 (1.09-1.22) <.001
Methadone 28.61 (22.80-35.90) <.001 1.90 (1.79-2.02) <.001
Naltrexone (oral or extended release) 2.05 (1.58-2.66) <.001 0.85 (0.77-0.95) .004
No MOUD 1 [Reference] NA 1 [Reference] NA
Opioid overdosec 0.98 (0.83-1.16) .80 0.96 (0.89-1.03) .25
Disposition site
Home or routine discharge 1 [Reference] NA 1 [Reference] NA
Home with services 0.65 (0.54-0.78) <.001 1.04 (0.95-1.13) .39
Transferred, acute care hospital 0.88 (0.71-1.10) .26 0.96 (0.87-1.05) .35
SNF or rehabilitation facility 0.79 (0.67-0.92) .002 0.94 (0.88-1.00) .07
Patient-directed discharge 0.76 (0.63-0.92) .004 0.86 (0.80-0.93) <.001
Other 1.10 (0.83-1.48) .50 1.02 (0.90-1.15) .78
Infection type
Endocarditis 1.83 (1.55-2.15) <.001 1.03 (0.96-1.10) .40
Epidural abscess 1.24 (0.98-1.56) .07 1.05 (0.95-1.16) .33
Septic arthritis 1.80 (1.48-2.18) <.001 0.98 (0.91-1.06) .64
Osteomyelitis 1.05 (0.89-1.23) .57 1.05 (0.91-1.06) .64
Bloodstream infection 1 [Reference] NA 1 [Reference] NA
Open in a new tab

Abbreviations: AOR, adjusted odds ratio; IRR, incident rate ratio; MOUD, medication for opioid use disorder; NA, not applicable; SNF, skilled nursing facility.

a

Zero-inflated negative binomial models with a 2-step modeling approach. The first step, receipt of any MOUD, is presented as AOR, and the second step, total treatment time, is presented as IRR.

b

No additional description available from the database.

c

Homeless status; anxiety and depression; alcohol use disorder; stimulant use disorder; prior prescriptions for opioids, benzodiazepine, and naloxone; prior MOUD; and prior opioid overdose were assessed in the 6 months before SIRI hospitalization. All other characteristics were ascertained at time of hospitalization.

d

Modified Elixhauser Comorbidity Index score excludes alcohol, drug use, and anxiety or depression because these characteristics were examined separately.

Zero-Inflated Negative Binomial Model Results: MOUD Treatment Rates

Among the 4652 individuals receiving MOUD treatment, the following characteristics were significantly associated with higher rate ratios of MOUD treatment during follow-up: buprenorphine (incident rate ratio [IRR], 1.16; 95% CI, 1.09-1.22) or methadone (IRR, 1.90; 95% CI, 1.79-2.02) treatment in the 6 months before hospitalization and ages of 35 to 49 years (IRR, 1.09; 95% CI, 1.03-1.15) compared with 18 to 34 years. Characteristics associated with lower treatment rate ratios were treatment with oral or extended-release naltrexone (IRR, 0.85; 95% CI, 0.77-0.95) and opioid prescription (IRR, 0.90; 95% CI, 0.86-0.95) in the 6 months before the index SIRI hospitalization and patient-directed discharge compared with routine discharge (IRR, 0.86; 95% CI, 0.80-0.93).

Sensitivity Analyses

Analyses with buprenorphine, methadone, and extended-release naltrexone as separate outcomes had similar results (eTables 7-9 in Supplement 1). Notable associations were found for prior treatment with the same class of MOUD, with AORs of 15.94 (95% CI, 13.87-18.32) for buprenorphine, 31.55 (95% CI, 26.69-37.31) for methadone, and 10.38 (95% CI, 7.70-14.00) for naltrexone. In contrast, prior treatment with methadone was negatively associated with any buprenorphine receipt (AOR, 0.60; 95% CI, 0.51-0.72) and lower buprenorphine rate ratios (IRR, 0.84; 95% CI, 0.74-0.95). For models with extended-release naltrexone as the outcome, prior methadone treatment was negatively associated with any naltrexone receipt (AOR, 0.63; 95% CI, 0.43-0.93).

Discussion

In this cohort of 9757 individuals with OUD and SIRIs, more than half did not receive MOUD in the year after hospitalization despite evidence that these medications reduce rehospitalization and mortality.7,13,21,22,23 The period after SIRI hospitalization may be a critical opportunity to deliver substance use treatment, as 1-year mortality was 10.9%. Low MOUD treatment rates after SIRI hospitalization is likely a failure to initiate and link to MOUD treatment during hospitalizations. Improved clinician effort and systems such as multidisciplinary addiction consult services are needed to deliver MOUD to this population during a period when motivation may be high.16,40,41,42

Approximately 1 in 4 individuals received MOUD 3 months after discharge, a low percentage but higher than some previous estimates after SIRIs, which range from 5% to 25%.7,9,21,22 The characteristic most associated with receiving MOUD in the year after a SIRI was prior MOUD receipt, with strongest associations for methadone.43,44 Low-barrier access to MOUD in the community is needed to not only reduce risks in the first place but also improve uptake of MOUD after complications. However, initiation is only one challenge; retention requires greater attention. Only 40% of those who received MOUD after a SIRI were treated for more than half of the subsequent year. Of the 3 MOUD types, only methadone treatment rates increased during follow-up. Methadone regulations may deter some from initiating the medication, but it seems to be associated with improved retention after SIRI.44,45,46,47 Long-acting injectable buprenorphine may improve treatment retention after SIRIs, but research is ongoing for this population.48,49,50

Several other characteristics were associated with MOUD receipt after discharge: Medicaid, which may provide better MOUD coverage; homelessness, which may increase contact with services with addiction treatment; and anxiety or depression, which may reflect care seeking or likelihood of receiving diagnoses once in care. People with endocarditis and septic arthritis were also more likely to receive posthospitalization MOUD, perhaps reflecting an ability to marshal treatment during lengthier hospitalizations. Programs such as addiction consultations that connect hospitalized people who inject drugs to MOUD treatment are critically needed.16,19,23,51

Disposition to SNFs was associated with decreased MOUD receipt, perhaps reflecting admissions that exclude people receiving MOUD or poor linkage after SNF discharge.52,53 Being discharged home with services was also associated with less MOUD treatment, which may reflect similar stigma among home service agencies. Patient-directed discharge was associated with decreased receipt of MOUD, perhaps because those not receiving MOUD for withdrawal more commonly leave54,55,56,57 or because these discharges are often chaotic, making outpatient linkage challenging.58 These findings underscore the importance of improving MOUD delivery for inpatients and after hospitalization.

Finally, non-Hispanic Black patients were less likely to receive MOUD, corroborating the literature about addiction treatment inequities and racism.59,60,61,62,63 Because Black individuals are experiencing increasing overdose rates, ensuring access to effective addiction treatment is essential.64,65 Multipronged efforts are needed to address interpersonal and structural racism in OUD care.62,66,67

Total time receiving MOUD (eg, treatment rates), in addition to initiation or uninterrupted retention, may be an important OUD treatment measure, especially when discontinuation definitions vary.68 In this study, treatment rates were lower for the following: those who received opioid analgesics, those who had a patient-directed discharge, and those who received any form of naltrexone before hospitalization. Homelessness was not significantly associated with a lower treatment rate. Nonetheless, because lack of housing presents barriers to medication storage and appointment attendance,69,70,71 supportive housing may improve adherence.72 Integrating harm reduction, infectious disease and OUD treatment, mobile health, peer support, telephones, and housing assistance may improve MOUD treatment rates.40,50,73,74,75,76,77,78,79,80

Strengths and Limitations

This study has several strengths, including use of a longitudinal linked dataset,7,12 inclusion of all SIRIs in Massachusetts over 4.5 years, and granular weekly MOUD indicators. The modeling approach accounted for treatment interruptions.

This study also has limitations. First, we expected some misclassification of OUD status, including attribution of infections to OUD caused by stimulant injection or other comorbidities. Second, the data do not include MOUD receipt during inpatient hospitalizations or incarceration, limiting assessments of their role in MOUD treatment. In addition, MOUD provided as part of clinical trials was not available, likely affecting few individuals. Third, post–acute care admission dates were not neatly delineated in the PHD. Fourth, this study was subject to residual confounding. We were limited by available data and could not examine all characteristics (eg, methadone dose, transportation, distance to MOUD treatment, and changes to housing). Fifth, generalizability to other locations or times may be limited due to differences in drug supply or treatment landscape.

Conclusions

Among Massachusetts patients with SIRIs, fewer than 1 in 2 received MOUD in the year after the infection even though these medications are associated with improved mortality and decreased rehospitalization. Among those who received MOUD, only 4 in 10 were treated for at least 50% of the year after hospitalization. Prior MOUD treatment experience was associated with increased MOUD treatment. To improve MOUD initiation and retention after a SIRI, dedicated efforts are needed to integrate MOUD into SIRI follow-up care and reform the addiction treatment delivery system to make services more accessible and appealing.

Supplement 1.

eFigure 1. Cohort construction, covariate, and outcome assessment for individuals with serious injection-related infections (SIRI)

eFigure 2. Consort diagram, cohort construction

eFigure 3. Illustrative examples of MOUD assessment following SIRI hospitalization with exclusions for hospitalization, incarceration, and death

eFigure 4. Weekly methadone, buprenorphine and naltrexone receipt 6 months before and 12 months after hospitalization for serious injection related infection for those discharged home or to skilled nursing facilities or rehabilitation, Massachusetts, January 1, 2014, to December 31, 2020

eTable 1. Sources of data in Massachusetts Public Health Data Warehouse in study

eTable 2. ICD-9 and ICD-10 codes for serious infections

eTable 3. Identification of opioid use disorder (OUD)

eTable 4. ICD-9 and ICD-10 codes used to identify opioid use disorder (OUD)

eTable 5. Codes used to identify covariates

eTable 6. Datasets used to identify homelessness in the Massachusetts Public Health Data Warehouse

eTable 7. Zero-inflated negative binomial model results: characteristics associated with any buprenorphine receipt and weeks treated with buprenorphine among individuals with serious injection-related infections, Massachusetts, July 1, 2014, to December 31, 2019

eTable 8. Zero-inflated negative binomial model results: characteristics associated with any methadone receipt and weeks treated with methadone among individuals with serious injection-related infections, Massachusetts, July 1, 2014, to December 31, 2019

eTable 9. Zero-inflated negative binomial model results: characteristics associated with any extended-release naltrexone receipt and weeks treated with extended-release naltrexone among individuals with serious injection-related infections, Massachusetts, July 1, 2014, to December 31, 2019

jamanetwopen-e2421740-s001.pdf (631.7KB, pdf)
Supplement 2.

Data Sharing Statement

jamanetwopen-e2421740-s002.pdf (15.8KB, pdf)

References

  • 1.Ronan MV, Herzig SJ. Hospitalizations related to opioid abuse/dependence and associated serious infections increased sharply, 2002-12. Health Aff (Millwood). 2016;35(5):832-837. doi: 10.1377/hlthaff.2015.1424 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Njoroge LW, Al-Kindi SG, Koromia GA, ElAmm CA, Oliveira GH. Changes in the association of rising infective endocarditis with mortality in people who inject drugs. JAMA Cardiol. 2018;3(8):779-780. doi: 10.1001/jamacardio.2018.1602 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Wurcel AG, Anderson JE, Chui KKH, et al. Increasing infectious endocarditis admissions among young people who inject drugs. Open Forum Infect Dis. 2016;3(3):ofw157. doi: 10.1093/ofid/ofw157 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Serota DP, Bartholomew TS, Tookes HE. Evaluating differences in opioid and stimulant use-associated infectious disease hospitalizations in Florida, 2016-2017. Clin Infect Dis. 2021;73(7):e1649-e1657. doi: 10.1093/cid/ciaa1278 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Schranz A, Barocas JA. Infective endocarditis in persons who use drugs: epidemiology, current management, and emerging treatments. Infect Dis Clin North Am. 2020;34(3):479-493. doi: 10.1016/j.idc.2020.06.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Levy S, Bassler J, Gagnon K, et al. Methamphetamines and serious injection-related infections: opioid use care continuum and opportunities to end Alabama’s drug crisis. Open Forum Infect Dis. 2022;10(1):ofac708. doi: 10.1093/ofid/ofac708 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Kimmel SD, Walley AY, Li Y, et al. Association of treatment with medications for opioid use disorder with mortality after hospitalization for injection drug use-associated infective endocarditis. JAMA Netw Open. 2020;3(10):e2016228. doi: 10.1001/jamanetworkopen.2020.16228 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Thakarar K, Rokas KE, Lucas FL, et al. Mortality, morbidity, and cardiac surgery in Injection Drug Use (IDU)-associated versus non-IDU infective endocarditis: the need to expand substance use disorder treatment and harm reduction services. PLoS One. 2019;14(11):e0225460. doi: 10.1371/journal.pone.0225460 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Rodger L, Glockler-Lauf SD, Shojaei E, et al. Clinical characteristics and factors associated with mortality in first-episode infective endocarditis among persons who inject drugs. JAMA Netw Open. 2018;1(7):e185220. doi: 10.1001/jamanetworkopen.2018.5220 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Bor DH, Woolhandler S, Nardin R, Brusch J, Himmelstein DU. Infective endocarditis in the U.S., 1998-2009: a nationwide study. PLoS One. 2013;8(3):e60033. doi: 10.1371/journal.pone.0060033 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Leahey PA, LaSalvia MT, Rosenthal ES, Karchmer AW, Rowley CF. High morbidity and mortality among patients with sentinel admission for injection drug use-related infective endocarditis. Open Forum Infect Dis. 2019;6(4):ofz089. doi: 10.1093/ofid/ofz089 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Larochelle MR, Bernstein R, Bernson D, et al. Touchpoints—opportunities to predict and prevent opioid overdose: a cohort study. Drug Alcohol Depend. 2019;204:107537. doi: 10.1016/j.drugalcdep.2019.06.039 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Brothers TD, Lewer D, Jones N, et al. Opioid agonist treatment and risk of death or rehospitalization following injection drug use-associated bacterial and fungal infections: a cohort study in New South Wales, Australia. PLoS Med. 2022;19(7):e1004049. doi: 10.1371/journal.pmed.1004049 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Baddour LM, Wilson WR, Bayer AS, et al. ; American Heart Association Committee on Rheumatic Fever, Endocarditis, and Kawasaki Disease of the Council on Cardiovascular Disease in the Young, Council on Clinical Cardiology, Council on Cardiovascular Surgery and Anesthesia, and Stroke Council . Infective endocarditis in adults: diagnosis, antimicrobial therapy, and management of complications: a scientific statement for healthcare professionals from the American Heart Association. Circulation. 2015;132(15):1435-1486. doi: 10.1161/CIR.0000000000000296 [DOI] [PubMed] [Google Scholar]
  • 15.Kim JH, Fine DR, Li L, et al. Disparities in United States hospitalizations for serious infections in patients with and without opioid use disorder: a nationwide observational study. PLoS Med. 2020;17(8):e1003247. doi: 10.1371/journal.pmed.1003247 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Weinstein ZM, Wakeman SE, Nolan S. Inpatient addiction consult service: expertise for hospitalized patients with complex addiction problems. Med Clin North Am. 2018;102(4):587-601. doi: 10.1016/j.mcna.2018.03.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Barocas JA, Savinkina A, Adams J, et al. Clinical impact, costs, and cost-effectiveness of hospital-based strategies for addressing the US opioid epidemic: a modelling study. Lancet Public Health. 2022;7(1):e56-e64. doi: 10.1016/S2468-2667(21)00248-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Englander H, Dobbertin K, Lind BK, et al. Inpatient addiction medicine consultation and post-hospital substance use disorder treatment engagement: a propensity-matched analysis. J Gen Intern Med. 2019;34(12):2796-2803. doi: 10.1007/s11606-019-05251-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Stein MD, Phillips KT, Herman DS, et al. Skin-cleaning among hospitalized people who inject drugs: a randomized controlled trial. Addiction. 2021;116(5):1122-1130. doi: 10.1111/add.15236 [DOI] [PubMed] [Google Scholar]
  • 20.Leshner AI, Mancher M, eds. Medications for Opioid Use Disorder Save Lives. National Academies Press; 2019. doi: 10.17226/25310 [DOI] [PubMed] [Google Scholar]
  • 21.Barocas JA, Morgan JR, Wang J, McLoone D, Wurcel A, Stein MD. Outcomes associated with medications for opioid use disorder among persons hospitalized for infective endocarditis. Clin Infect Dis. 2021;72(3):472-478. doi: 10.1093/cid/ciaa062 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Marks LR, Munigala S, Warren DK, Liang SY, Schwarz ES, Durkin MJ. Addiction medicine consultations reduce readmission rates for patients with serious infections from opioid use disorder. Clin Infect Dis. 2019;68(11):1935-1937. doi: 10.1093/cid/ciy924 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Serota DP, Rosenbloom L, Hervera B, et al. Integrated infectious disease and substance use disorder care for the treatment of injection drug use-associated infections: a prospective cohort study with historical control. Open Forum Infect Dis. 2022;10(1):ofac688. doi: 10.1093/ofid/ofac688 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Stein MD, Anderson B. Injection frequency mediates health service use among persons with a history of drug injection. Drug Alcohol Depend. 2003;70(2):159-168. doi: 10.1016/S0376-8716(02)00344-7 [DOI] [PubMed] [Google Scholar]
  • 25.Rosenthal ES, Karchmer AW, Theisen-Toupal J, Castillo RA, Rowley CF. Suboptimal addiction interventions for patients hospitalized with injection drug use-associated infective endocarditis. Am J Med. 2016;129(5):481-485. doi: 10.1016/j.amjmed.2015.09.024 [DOI] [PubMed] [Google Scholar]
  • 26.Renner JA, Rubin ML. Engaging heroin addicts in treatment. Am J Psychiatry. 1973;130(9):976-980. doi: 10.1176/ajp.130.9.976 [DOI] [PubMed] [Google Scholar]
  • 27.Alford DP, LaBelle CT, Kretsch N, et al. Collaborative care of opioid-addicted patients in primary care using buprenorphine: five-year experience. Arch Intern Med. 2011;171(5):425-431. doi: 10.1001/archinternmed.2010.541 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Shanahan CW, Beers D, Alford DP, Brigandi E, Samet JH. A transitional opioid program to engage hospitalized drug users. J Gen Intern Med. 2010;25(8):803-808. doi: 10.1007/s11606-010-1311-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Trowbridge P, Weinstein ZM, Kerensky T, et al. Addiction consultation services—linking hospitalized patients to outpatient addiction treatment. J Subst Abuse Treat. 2017;79:1-5. doi: 10.1016/j.jsat.2017.05.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Wakeman SE, Metlay JP, Chang Y, Herman GE, Rigotti NA. Inpatient addiction consultation for hospitalized patients increases post-discharge abstinence and reduces addiction severity. J Gen Intern Med. 2017;32(8):909-916. doi: 10.1007/s11606-017-4077-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Barocas JA, White LF, Wang J, et al. Estimated prevalence of opioid use disorder in Massachusetts, 2011-2015: a capture-recapture analysis. Am J Public Health. 2018;108(12):1675-1681. doi: 10.2105/AJPH.2018.304673 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Commonwealth of Massachusetts . Chapter 55: An Act Requiring Certain Reports for Opiate Overdoses. Published August 5, 2015. Accessed September 16, 2020. https://malegislature.gov/laws/sessionlaws/acts/2015/chapter55
  • 33.Massachusetts Department of Public Health . Public Health Data Warehouse (PHD): Overview. Accessed February 1, 2023. https://www.mass.gov/info-details/public-health-data-warehouse-phd-overview
  • 34.Commonwealth of Massachusetts . Outside Section 48: Public Health Data 1. FY2018 Budget Summary. Published 2017. Accessed October 3, 2023. https://budget.digital.mass.gov/bb/gaa/fy2018/os_18/h48.htm
  • 35.Larochelle MR, Bernson D, Land T, et al. Medication for opioid use disorder after nonfatal opioid overdose and association with mortality: a cohort study. Ann Intern Med. 2018;169(3):137-145. doi: 10.7326/M17-3107 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP; STROBE Initiative . The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: guidelines for reporting observational studies. Int J Surg. 2014;12(12):1495-1499. doi: 10.1016/j.ijsu.2014.07.013 [DOI] [PubMed] [Google Scholar]
  • 37.Bagley SM, Larochelle MR, Xuan Z, et al. Characteristics and receipt of medication treatment among young adults who experience a nonfatal opioid-related overdose. Ann Emerg Med. 2020;75(1):29-38. doi: 10.1016/j.annemergmed.2019.07.030 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Rajan S, Ruggles KV, Guarino H, Mateu-Gelabert P. Heroin use and drug injection among youth also misusing prescription drugs. Am J Health Behav. 2018;42(1):144-155. doi: 10.5993/AJHB.42.1.14 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Elixhauser A, Steiner C, Harris DR, Coffey RM. Comorbidity measures for use with administrative data. Med Care. 1998;36(1):8-27. doi: 10.1097/00005650-199801000-00004 [DOI] [PubMed] [Google Scholar]
  • 40.Seval N, Eaton E, Springer SA. Beyond antibiotics: a practical guide for the infectious disease physician to treat opioid use disorder in the setting of associated infectious diseases. Open Forum Infect Dis. 2019;7(1):ofz539. doi: 10.1093/ofid/ofz539 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Kimmel SD, Phillips KT, Anderson BJ, Stein MD. Characteristics associated with motivation to stop substance use and improve skin and needle hygiene among hospitalized patients who inject drugs. Subst Abus. 2022;43(1):878-883. doi: 10.1080/08897077.2021.2007520 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Pollini RA, O’Toole TP, Ford D, Bigelow G. Does this patient really want treatment? factors associated with baseline and evolving readiness for change among hospitalized substance using adults interested in treatment. Addict Behav. 2006;31(10):1904-1918. doi: 10.1016/j.addbeh.2006.01.003 [DOI] [PubMed] [Google Scholar]
  • 43.Uebelacker LA, Bailey G, Herman D, Anderson B, Stein M. Patients’ beliefs about medications are associated with stated preference for methadone, buprenorphine, naltrexone, or no medication-assisted therapy following inpatient opioid detoxification. J Subst Abuse Treat. 2016;66:48-53. doi: 10.1016/j.jsat.2016.02.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Timko C, Schultz NR, Cucciare MA, Vittorio L, Garrison-Diehn C. Retention in medication-assisted treatment for opiate dependence: a systematic review. J Addict Dis. 2016;35(1):22-35. doi: 10.1080/10550887.2016.1100960 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Hser YI, Evans E, Huang D, et al. Long-term outcomes after randomization to buprenorphine/naloxone versus methadone in a multi-site trial. Addiction. 2016;111(4):695-705. doi: 10.1111/add.13238 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Hser YI, Saxon AJ, Huang D, et al. Treatment retention among patients randomized to buprenorphine/naloxone compared to methadone in a multi-site trial. Addiction. 2014;109(1):79-87. doi: 10.1111/add.12333 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Wakeman SE, Larochelle MR, Ameli O, et al. Comparative effectiveness of different treatment pathways for opioid use disorder. JAMA Netw Open. 2020;3(2):e1920622. doi: 10.1001/jamanetworkopen.2019.20622 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Lee JD, Malone M, McDonald R, et al. Comparison of treatment retention of adults with opioid addiction managed with extended-release buprenorphine vs daily sublingual buprenorphine-naloxone at time of release from jail. JAMA Netw Open. 2021;4(9):e2123032. doi: 10.1001/jamanetworkopen.2021.23032 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Morgan JR, Walley AY, Murphy SM, et al. Characterizing initiation, use, and discontinuation of extended-release buprenorphine in a nationally representative United States commercially insured cohort. Drug Alcohol Depend. 2021;225:108764. doi: 10.1016/j.drugalcdep.2021.108764 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Bart G, Korthuis PT. Exemplar Hospital Initiation Trial to Enhance Treatment Engagement (Protocol 0098). National Institute on Drug Abuse, Clinical Trials Network. Accessed June 6, 2024. https://www.drugabuse.gov/about-nida/organization/cctn/ctn/research-studies/exemplar-hospital-initiation-trial-to-enhance-treatment-engagement-exhit-entre-nih-heal-initiative
  • 51.Marks LR, Liang SY, Muthulingam D, et al. Evaluation of partial oral antibiotic treatment for persons who inject drugs and are hospitalized with invasive infections. Clin Infect Dis. 2020;71(10):e650-e656. doi: 10.1093/cid/ciaa365 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Kimmel SD, Rosenmoss S, Bearnot B, et al. Northeast postacute medical facilities disproportionately reject referrals for patients with opioid use disorder. Health Aff (Millwood). 2022;41(3):434-444. doi: 10.1377/hlthaff.2021.01242 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Kimmel SD, Rosenmoss S, Bearnot B, Larochelle M, Walley AY. Rejection of patients with opioid use disorder referred for post-acute medical care before and after an anti-discrimination settlement in Massachusetts. J Addict Med. 2021;15(1):20-26. doi: 10.1097/ADM.0000000000000693 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Pollini RA, Paquette CE, Drvar T, et al. A qualitative assessment of discharge against medical advice among patients hospitalized for injection-related bacterial infections in West Virginia. Int J Drug Policy. 2021;94:103206. doi: 10.1016/j.drugpo.2021.103206 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Simon R, Snow R, Wakeman S. Understanding why patients with substance use disorders leave the hospital against medical advice: a qualitative study. Subst Abus. 2020;41(4):519-525. doi: 10.1080/08897077.2019.1671942 [DOI] [PubMed] [Google Scholar]
  • 56.Nolan NS, Marks LR, Liang SY, Durkin MJ. Medications for opioid use disorder associated with less against medical advice discharge among persons who inject drugs hospitalized with an invasive infection. J Addict Med. 2021;15(2):155-158. doi: 10.1097/ADM.0000000000000725 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Eaton EF, Westfall AO, McClesky B, et al. In-hospital illicit drug use and patient-directed discharge: barriers to care for patients with injection-related infections. Open Forum Infect Dis. 2020;7(3):ofaa074. doi: 10.1093/ofid/ofaa074 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Southern WN, Nahvi S, Arnsten JH. Increased risk of mortality and readmission among patients discharged against medical advice. Am J Med. 2012;125(6):594-602. doi: 10.1016/j.amjmed.2011.12.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Priest KC, King CA, Englander H, Lovejoy TI, McCarty D. Differences in the delivery of medications for opioid use disorder during hospitalization by racial categories: a retrospective cohort analysis. Substance Abuse. 2022;43(1):1251-1259. doi: 10.1080/08897077.2022.2074601 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60.Peeler M, Gupta M, Melvin P, et al. Racial and ethnic disparities in maternal and infant outcomes among opioid-exposed mother-infant dyads in Massachusetts (2017-2019). Am J Public Health. 2020;110(12):1828-1836. doi: 10.2105/AJPH.2020.305888 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Lagisetty PA, Ross R, Bohnert A, Clay M, Maust DT. Buprenorphine treatment divide by race/ethnicity and payment. JAMA Psychiatry. 2019;76(9):979-981. doi: 10.1001/jamapsychiatry.2019.0876 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 62.Jackson DS, Nguemeni Tiako MJ, Jordan A. Disparities in addiction treatment: learning from the past to forge an equitable future. Med Clin North Am. 2022;106(1):29-41. doi: 10.1016/j.mcna.2021.08.008 [DOI] [PubMed] [Google Scholar]
  • 63.Hall OT, Jordan A, Teater J, et al. Experiences of racial discrimination in the medical setting and associations with medical mistrust and expectations of care among black patients seeking addiction treatment. J Subst Abuse Treat. 2022;133:108551. doi: 10.1016/j.jsat.2021.108551 [DOI] [PubMed] [Google Scholar]
  • 64.Larochelle MR, Slavova S, Root ED, et al. Disparities in opioid overdose death trends by race/ethnicity, 2018-2019, from the HEALing Communities Study. Am J Public Health. 2021;111(10):1851-1854. doi: 10.2105/AJPH.2021.306431 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 65.Friedman J, Mann NC, Hansen H, et al. Racial/ethnic, social, and geographic trends in overdose-associated cardiac arrests observed by US emergency medical services during the COVID-19 pandemic. JAMA Psychiatry. 2021;78(8):886-895. doi: 10.1001/jamapsychiatry.2021.0967 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 66.Kunins HV. Structural racism and the opioid overdose epidemic: the need for antiracist public health practice. J Public Health Manag Pract. 2020;26(3):201-205. doi: 10.1097/PHH.0000000000001168 [DOI] [PubMed] [Google Scholar]
  • 67.Farahmand P, Arshed A, Bradley MV. Systemic racism and substance use disorders. Psychiatr Ann. 2020;50(11):494-498. doi: 10.3928/00485713-20201008-01 [DOI] [Google Scholar]
  • 68.Dong H, Stringfellow EJ, Russell WA, Bearnot B, Jalali MS. Impact of alternative ways to operationalize buprenorphine treatment duration on understanding continuity of care for opioid use disorder. Int J Ment Health Addict. Published online December 16, 2022. doi: 10.1007/s11469-022-00985-w [DOI] [Google Scholar]
  • 69.McLaughlin MF, Li R, Carrero ND, Bain PA, Chatterjee A. Opioid use disorder treatment for people experiencing homelessness: a scoping review. Drug Alcohol Depend. 2021;224:108717. doi: 10.1016/j.drugalcdep.2021.108717 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 70.Swartz N, Adnan T, Peréa F, Baggett TP, Chatterjee A. “Sick and tired of being sick and tired”: exploring initiation of medications for opioid use disorder among people experiencing homelessness. J Subst Abuse Treat. 2022;138:108752. doi: 10.1016/j.jsat.2022.108752 [DOI] [PubMed] [Google Scholar]
  • 71.Fine DR, Lewis E, Weinstock K, Wright J, Gaeta JM, Baggett TP. Office-based addiction treatment retention and mortality among people experiencing homelessness. JAMA Netw Open. 2021;4(3):e210477. doi: 10.1001/jamanetworkopen.2021.0477 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 72.Fine DR, Dickins KA, Adams LD, et al. Drug overdose mortality among people experiencing homelessness, 2003 to 2018. JAMA Netw Open. 2022;5(1):e2142676. doi: 10.1001/jamanetworkopen.2021.42676 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 73.Brooks HL, O’Brien DC, Salvalaggio G, Dong K, Hyshka E. Uptake into a bedside needle and syringe program for acute care inpatients who inject drugs. Drug Alcohol Rev. 2019;38(4):423-427. doi: 10.1111/dar.12930 [DOI] [PubMed] [Google Scholar]
  • 74.Serota DP, Tookes HE, Hervera B, et al. Harm reduction for the treatment of patients with severe injection-related infections: description of the Jackson SIRI Team. Ann Med. 2021;53(1):1960-1968. doi: 10.1080/07853890.2021.1993326 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 75.Sharma M, Lamba W, Cauderella A, Guimond TH, Bayoumi AM. Harm reduction in hospitals. Harm Reduct J. 2017;14(1):32. doi: 10.1186/s12954-017-0163-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 76.Phillips KT, Stewart C, Anderson BJ, Liebschutz JM, Herman DS, Stein MD. A randomized controlled trial of a brief behavioral intervention to reduce skin and soft tissue infections among people who inject drugs. Drug Alcohol Depend. 2021;221:108646. doi: 10.1016/j.drugalcdep.2021.108646 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 77.Jakubowski A, Norton BL, Hayes BT, et al. Low-threshold buprenorphine treatment in a syringe services program: program description and outcomes. J Addict Med. 2022;16(4):447-453. doi: 10.1097/ADM.0000000000000934 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 78.National Institute on Drug Abuse. Exemplar Hospital Initiation Trial to Enhance Treatment Engagement (EXHIT ENTRE). Accessed June 6, 2024. https://nida.nih.gov/about-nida/organization/cctn/ctn/research-studies/exemplar-hospital-initiation-trial-to-enhance-treatment-engagement-exhit-entre-nih-heal-initiative [DOI] [PMC free article] [PubMed]
  • 79.Hill K, Kuo I, Shenoi SV, Desruisseaux MS, Springer SA. Integrated care models: HIV and substance use. Curr HIV/AIDS Rep. 2023;20(5):286-295. doi: 10.1007/s11904-023-00667-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 80.Bart G, Korthuis PT, Donohue JM, et al. Exemplar Hospital initiation trial to Enhance Treatment Engagement (EXHIT ENTRE): protocol for CTN-0098B a randomized implementation study to support hospitals in caring for patients with opioid use disorder. Addict Sci Clin Pract. 2024;19(1):29. doi: 10.1186/s13722-024-00455-9 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement 1.

eFigure 1. Cohort construction, covariate, and outcome assessment for individuals with serious injection-related infections (SIRI)

eFigure 2. Consort diagram, cohort construction

eFigure 3. Illustrative examples of MOUD assessment following SIRI hospitalization with exclusions for hospitalization, incarceration, and death

eFigure 4. Weekly methadone, buprenorphine and naltrexone receipt 6 months before and 12 months after hospitalization for serious injection related infection for those discharged home or to skilled nursing facilities or rehabilitation, Massachusetts, January 1, 2014, to December 31, 2020

eTable 1. Sources of data in Massachusetts Public Health Data Warehouse in study

eTable 2. ICD-9 and ICD-10 codes for serious infections

eTable 3. Identification of opioid use disorder (OUD)

eTable 4. ICD-9 and ICD-10 codes used to identify opioid use disorder (OUD)

eTable 5. Codes used to identify covariates

eTable 6. Datasets used to identify homelessness in the Massachusetts Public Health Data Warehouse

eTable 7. Zero-inflated negative binomial model results: characteristics associated with any buprenorphine receipt and weeks treated with buprenorphine among individuals with serious injection-related infections, Massachusetts, July 1, 2014, to December 31, 2019

eTable 8. Zero-inflated negative binomial model results: characteristics associated with any methadone receipt and weeks treated with methadone among individuals with serious injection-related infections, Massachusetts, July 1, 2014, to December 31, 2019

eTable 9. Zero-inflated negative binomial model results: characteristics associated with any extended-release naltrexone receipt and weeks treated with extended-release naltrexone among individuals with serious injection-related infections, Massachusetts, July 1, 2014, to December 31, 2019

jamanetwopen-e2421740-s001.pdf (631.7KB, pdf)
Supplement 2.

Data Sharing Statement

jamanetwopen-e2421740-s002.pdf (15.8KB, pdf)

Articles from JAMA Network Open are provided here courtesy of American Medical Association

RESOURCES