Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Jan 17.
Published before final editing as: J Subst Use Addict Treat. 2025 Aug 25:209797. doi: 10.1016/j.josat.2025.209797

Factors Associated with Mortality Following Hospitalization Among Veterans with Opioid Use Disorder

Michael A Incze 1,2, Sophia Huebler 2,3,4, Jacob D Baylis 2,4, Andrea Stofko 2, A Taylor Kelley 1,4, Ingrid A Binswanger 5,6,7,8, Gavin Bart 9, Andrew J Saxon 10,11, Ajay Manhapra 12,13, Adam J Gordon 2,4
PMCID: PMC12488471  NIHMSID: NIHMS2109208  PMID: 40865641

Abstract

Introduction:

Hospitalizations are common among people with opioid use disorder (OUD). While hospitalizations represent opportunities to engage patients and offer treatment, they are also destabilizing events associated with an increased risk of death in the post-hospitalization period.

Methods:

We conducted a retrospective cohort study within the Veterans Health Administration including all Veterans with OUD who experienced at least one medical hospitalization between January 2011 and December 2021. We examined which patient-level clinical and demographic factors were associated with all-cause and opioid-related mortality within 0–30 and 0–365 days following an index medical hospitalization.

Results:

The cohort included 90,920 Veterans with OUD who experienced one or more medical hospitalizations during the study period. Median age was 58 years, and 93% were male. Older age (adjusted Odds Ratio [aOR] range 30d: 1.50–2.66; 1y: 1.58–3.28), higher medical complexity (aOR range 30d: 2.11–6.23; 1y: 1.96–7.34), multiple substance use disorders (SUD; aOR 30d: 1.81 (95%CI 1.44,2.27) 1y: 1.48 [95%CI 1.36,1.62]), and length of hospitalization (aOR 30d: 6.78 [95%CI 4.85,9.47] 1y: 3.45 [95%CI 2.96,4.01]) were associated with increased all-cause mortality following hospitalization. Homelessness (aOR 30d: 0.75 [95%CI 0.63,0.90]; 1y: 0.85 [95%CI 0.80,0.91]), depression (aOR 1y: 0.89 [95%CI 0.84,0.95]), bipolar disorder (aOR 1y: 0.88 [95%CI 0.82,0.94]), buprenorphine receipt (aOR 1y: 0.79 [95%CI 0.69,0.91]), and service connection (aOR 30d: 0.76 [95%CI 0.60,0.97] 1y: 0.64 [95%CI 0.59,0.70]) were associated with reduced all-cause mortality. Younger age (aOR range 30d: 3.21–5.24; 1y: 2.71–2.38), homelessness (aOR 1y: 1.40 [95%CI 1.20,1.63]), and multiple SUD (aOR 1y: 1.78 [95%CI 1.33,2.38]) were among factors associated with increased opioid-related mortality after hospitalization. Black race (aOR 1y: 0.61 [95%CI 0.50,0.74]) and higher service connection (aOR 30d: 0.41 [95%CI 0.21,0.81]; 1y: 0.53 [95%CI 0.43–0.66]) were associated with reduced opioid-related mortality after hospitalization.

Conclusions:

Several patient-level factors were associated with increased all-cause mortality (e.g., length of hospital stay), reduced all-cause mortality (e.g., homelessness), increased opioid-related mortality (e.g., multiple SUD), and reduced opioid-related mortality (e.g., service connection) after hospitalization. This information provides a roadmap for future development and study of tailored supports and risk stratification tools to enhance post-hospitalization transitional care for patients with OUD.

Keywords: Opioid Use Disorders, Opioid Related Disorder, Transitional Care, Patient Navigation, Discharge Planning, Risk Assessment

Introduction

The ongoing overdose public health crisis continues to exact an immense toll on the lives and health of Americans.(Garnett MF, 2024) In 2022 alone, opioid overdose accounted for an estimated 3.1 million years of life lost in the US.(Hebert & Hill, 2024) Despite more than a decade of continued investment in expanding access to opioid use disorder (OUD) treatment, fewer than 25% of individuals with OUD access addiction care.(Krawczyk et al., 2022) Hospitals are increasingly recognized as important venues in which to intervene on OUD and other substance use disorders and offer treatment.(Honora Englander et al., 2019; H. Englander et al., 2019; Englander et al., 2024; Laura R. Marks et al., 2019; Ober et al., 2023; Salvalaggio et al., 2022; Suen et al., 2022) There is a growing body of evidence demonstrating that integrating addiction treatment interventions into hospital settings has strong positive impacts on OUD treatment initiation, linkage to follow up care, reduced substance use, and other important health-related outcomes.(Callister et al., 2022; Deng et al., 2022; Honora Englander et al., 2019; Englander et al., 2022; H. Englander et al., 2019; Englander et al., 2024; Harris et al., 2021; L. R. Marks et al., 2019; McNeely et al., 2024; Ober et al., 2025; Salvalaggio et al., 2022; Trowbridge et al., 2017; Wakeman et al., 2017; Weinstein et al., 2020; Weinstein et al., 2018; Wilson et al., 2022)

While medical hospitalizations represent opportunities to engage people with OUD and offer treatment, they can also be destabilizing events. Medical, psychosocial, environmental, and OUD-related factors may synergize in the post-hospitalization period to interrupt transitions to follow-up care. These factors include residual symptoms and physical debilitation from acute medical illness, new medications and their attendant side effects, lack of stable housing and reliable transportation, insurance-related barriers, and opioid cravings, among others. Prior single site and regional studies of hospitalized patients with OUD have demonstrated mortality rates ranging from 8% to over 33% in the year following hospitalization, as well as an increased risk of rapid hospital re-admission.(Hochstatter et al., 2023; King et al., 2022; Moreno et al., 2019; Wilson et al., 2022) One study examining a national sample of hospitalized Veterans with OUD found that while all-cause mortality within one year post-discharge was 7.8% (comparing favorably to certain non-Veteran groups with OUD), there was a nearly 3-fold increased risk of dying in the two-week period immediately following discharge compared to later in the same year.(Incze et al., 2025)

These findings underscore the need to develop interventions that help individuals with OUD rapidly link to medical and OUD care following hospitalization. While several models have been explored to support rapid linkage to follow-up OUD and medical care after hospitalization, the optimal components, design, and implementation approach for OUD care transition interventions are unknown.(M. A. Incze, S. Huebler, D. Chen, et al., 2024; M. A. Incze, S. Huebler, K. Szczotka, et al., 2024; M.A. Incze et al., 2024; Incze et al., 2023; James et al., 2023; Krawczyk et al., 2024; Smith et al., 2021; Snow et al., 2019; Taylor et al., 2023) Identifying patient-level factors that are associated with mortality following hospitalization among patients with OUD can help both to identify patients who are at the highest risk of death in the immediate post-discharge period as well as to inform the development of tailored interventions to provide optimal post-hospitalization support. Yet, while prior work has generally characterized factors associated with mortality among people with OUD such as access to medication treatment and stable housing,(Fine et al., 2020; Larochelle et al., 2018) little is known about which factors influence mortality risk specific to the immediate post-hospitalization period.

We conducted a national study of hospitalized Veterans with OUD to identify clinical and demographic factors associated with mortality risk within 30 days and one year following a medical hospitalization. To our knowledge, this study is the first of its kind using national data from the US Department of Veterans Affairs (VA) and may provide a roadmap for future study and development of tailored supports to enhance post-hospital transitional care for patients with OUD.

Methods

Study Design and Data Sources

We conducted a retrospective cohort study of Veterans with OUD who experienced one or more medical hospitalizations within the VA between January 1, 2011, and December 31, 2021. Data to characterize and develop our cohort, including hospitalizations, medication dispensation, healthcare encounters, demographics, and medical diagnoses were gathered from the Veterans Health Administration’s Corporate Data Warehouse within the VA’s Veterans Informatics and Computing Infrastructure (VINCI) framework. These data were linked to mortality data from the VA Mortality Data Repository. The University of Utah Institutional Review Board approved this study. We followed STROBE guidelines for reporting study results.(Vandenbroucke et al., 2007)

Study Setting and Population

The study population included all Veterans aged ≥18 years with a documented OUD diagnosis who experienced one or more medical hospitalizations within the VA system during the study period. OUD diagnosis was determined by ICD-9 and ICD-10 coding, had to be documented within two years prior to hospitalization, and excluded codes for OUD in remission to focus our evaluation on individuals with active OUD (Appendix, Table 1). The one-year observation window started at the day of discharge from the index medical hospitalization, identified from the health record as the first inpatient medical admission lasting under 45 days during the study period where the patient was alive at discharge. Each member of the cohort was included only once (i.e., only the first qualifying hospitalization for each individual was included in the analysis; subsequent hospitalizations during the study period were not counted). We excluded non-medical hospitalizations (e.g., psychiatric, inpatient substance use treatment) to focus our assessment on care transitions from the medical hospital setting – which is unique with respect to scope of care, resources, and clinical factors affecting post-discharge care linkage.

Main Outcomes

Our primary outcome was all-cause mortality within two pre-specified time intervals (0–30 days and 0–365 days) following index hospital discharge. Our secondary outcome was opioid-related mortality as defined by ICD coding in the death record (Appendix, Table 2). Individuals who died within 30 days of index hospitalization were also included in the one-year analysis. We examined clinical and demographic factors among our cohort that were associated with both all-cause and opioid-related mortality. Demographic factors included age, sex, self-reported race/ethnicity, rural/urban county of residence, marital status, VA service connection, and housing status (i.e., homelessness or housing instability vs not). Clinical factors included year of index hospitalization, length of hospital stay, medical complexity, co-occurring mental health and substance use disorders, and receipt of medications for OUD.

Determination of housing status/homelessness followed established methods using ICD-10 and VA clinical administrative codes (Appendix, Table 3).(Tsai et al., 2022) Receipt of buprenorphine (only formulations FDA-approved for OUD treatment) and naloxone were identified by examining inpatient and outpatient dispensation records within the VA. Our data did not account for direct administration or dispensation of these medications, nor did it account for medications received outside of the VA. For each patient, we calculated the Charlson Comorbidity Index (CCI) at a single time point at the start of the observation period as a measure of overall medical complexity.(Charlson et al., 1987) Service connection (i.e., the degree to which a Veteran’s medical condition or disability can be linked to military service) served as an indicator of eligibility to receive VA benefits and priority access to care.(2023) Data on race and ethnicity were based on self-report and included to examine the association of these factors with our study outcomes.

Statistical Analysis

Data analysis occurred between May 2024 and March 2025 using SQL Server Management Studio (SSMS) and R version 4.4.1. We used descriptive statistics to characterize our sample. Our primary and secondary outcomes were coded as binary indicators of mortality. We used multivariate logistic regressions to simultaneously examine the association of all available clinical and demographic variables with mortality following an index medical hospitalization. Separate models were fitted for each combination of mortality outcome (all-cause and opioid-related) and timeframe (0–30 days and 0–365 days). We coded missing values as an ‘Unknown’ category for each categorical variable. Adjusted odds ratios (aOR), derived from exponentiated coefficients of the model, and 95% Confidence Intervals (95%CI) were reported for each factor, illustrating its positive or negative association with all-cause and opioid-related mortality. P values were 2-sided, and statistical significance was set at p<0.05.

Results

Our cohort included 90,920 Veterans with OUD who experienced one or more medical hospitalizations during the study period. Median (Interquartile Range) age was 58 (50,64) years, and 93% were male. There were 7,040 (7.7%) deaths from any cause within the one-year observation window, including 810 (0.8%) opioid-related deaths. Opioid-related deaths represented 11.5% of all deaths within the one-year observation window. There were 906 (1.0%) total deaths within 30 days of discharge, including 88 (<0.01%) opioid-related deaths. Opioid-related deaths represented 9.7% of deaths within 30 days of hospital discharge. Among the total cohort, 4,470 (4.9%) individuals received a prescription for naloxone in the VA within 6 months prior to the index hospitalization, and 3,970 (4.4%) received a prescription for buprenorphine within the VA within 30 days prior to index hospital admission. Demographic and clinical characteristics of our cohort are summarized in Table 1.

Table 1:

Sample Demographic and Clinical Characteristics

Characteristic Overall N = 90,920 Survived N = 83,880 Death During Observation Window N = 7,040
Median Age (IQR) 58.0 (50.0, 64.0) 58.0 (49.0, 63.0) 62.0 (57.0, 67.0)
Sex (%)
 Male 84,711 (93%) 77,942 (93%) 6,769 (96%)
 Female 6,111 (6.7%) 5,843 (7.0%) 268 (3.8%)
 Unknown 98 (0.1%) 95 (0.1%) 3 (<0.1%)
Race (%)
 White 63,714 (70%) 58,707 (70%) 5,007 (71%)
 Black 22,142 (24%) 20,505 (24%) 1,637 (23%)
 Other 2,343 (2.6%) 2,166 (2.6%) 177 (2.5%)
 Unknown 2,721 (3.0%) 2,502 (3.0%) 219 (3.1%)
Ethnicity(%)
 Hispanic/Latine 5,127 (5.6%) 4,738 (5.6%) 389 (5.5%)
 Non-Hispanic/Latine 83,839 (92%) 77,327 (92%) 6,512 (93%)
 Unknown 1,954 (2.1%) 1,815 (2.2%) 139 (2.0%)
Marital Status (%)
 Married 26,228 (29%) 24,215 (29%) 2,013 (29%)
 Single 64,476 (71 %) 59,465 (71 %) 5,011 (71%)
 Unknown 216 (0.2%) 200 (0.2%) 16 (0.2%)
Rural (%)
 Urban 67,588 (74%) 62,281 (74%) 5,307 (75%)
 Rural 23,281 (26%) 21,549 (26%) 1,732 (25%)
 Unknown 51 (<0.1%) 50 (<0.1%) 1 (<0.1%)
Veterans experiencing homelessness (%) 24,750 (27%) 23,086 (28%) 1,664 (24%)
Charlson Co-Morbidity Index Score (%)
 0 50,809.0 (55.9%) 48,422.0 (57.7%) 2,387.0 (33.9%)
 1–5 32,996.0 (36.3%) 29,703.0 (35.4%) 3,293.0 (46.8%)
 5–10 6,685.0 (7.4%) 5,454.0 (6.5%) 1,231.0 (17.5%)
 10+ 430.0 (0.5%) 301.0 (0.4%) 129.0 (1.8%)
Length of Index Stay in Weeks (Mean (sd)) 0.7(1.0) 0.7(1.0) 0.9(1.1)
Service Connected Disability (%)
 No service related disability 10,907 (12%) 10,007 (12%) 900 (13%)
 <50% 35,282 (39%) 31,602 (38%) 3,680 (52%)
 50–100% 44,731 (49%) 42,271 (50%) 2,460 (35%)
Mental Health Diagnoses (%)
 Any Mental Health 79,969 (88%) 73,938 (88%) 6,031 (86%)
 Adjustment Disorder 32,231 (35%) 29,932 (36%) 2,299 (33%)
 Anxiety Disorders 48,069 (53%) 44,653 (53%) 3,416 (49%)
 Bipolar Disorder 19,540 (21%) 18,319 (22%) 1,221 (17%)
 Depressive Disorders 69,839 (77%) 64,673 (77%) 5,166 (73%)
 Post-Traumatic Stress Disorder 40,278 (44%) 37,472 (45%) 2,806 (40%)
 Schizophrenia 8,445 (9.3%) 7,877 (9.4%) 568 (8.1%)
Co-Occurring Substance Use Disorders (%)
 Alcohol Use Disorder 52,831 (58%) 48,842 (58%) 3,989 (57%)
 Cannabis Use Disorder 30,709 (34%) 28,574 (34%) 2,135 (30%)
 Cocaine Use Disorder 41,686 (46%) 38,596 (46%) 3,090 (44%)
 Sedative Use Disorder 31,636 (35%) 29,315 (35%) 2,321 (33%)
 Stimulant Use Disorder 31,391 (35%) 29,184 (35%) 2,207 (31%)
 Multiple Use Disorders 63,333 (70%) 58,376 (70%) 4,957 (70%)
Medication for OUD Receipt
 Buprenorphine Episode of Care at Time of Admission 3,970 (4.4%) 3,733 (4.5%) 237 (3.4%)
 Naloxone Received Within 6 Months Prior to Admission 4,470 (4.9%) 4,000 (4.8%) 470 (6.7%)
Open in a new tab

Factors Associated with Increased All-Cause Mortality

Table 2 illustrates which clinical and demographic factors were associated with increased all-cause mortality within 30-day and one-year time intervals after hospitalization. Older age was associated with increased all-cause mortality during both time intervals after hospitalization. Compared to the 41–64 age category, the 65–80 age category (30d: aOR=1.50 [95%CI 1.28,1.76]; 1y: aOR=1.58 [95%CI 1.48,1.68]) and >80yo category (30d: aOR=2.66 [95%CI 1,86,3.79]; 1y: aOR= 3.28 [95%CI 2.82,3.82]) were both associated with increased all-cause mortality risk, whereas the 25–40 age category was associated with reduced all-cause mortality within one year (aOR 0.62 [95%CI 0.55,0.69]).

Table 2.

Clinical and demographic characteristics associated with all-cause mortality within 30 days and 365 days of medical hospitalization.

Variable 30-day Mortality 365-day Mortality
Odds Ratio 95% CI P-Value Odds Ratio 95% CI P-Value
Age (Reference: 41–64yo)
 18–24 0.66 (0.21, 2.09) 0.48 0.49 (0.31, 0.76) 0.002
 25–40 0.65 (0.47, 0.90) 0.009 0.62 (0.55, 0.69) <0.001
 65–80 1.5 (1.28, 1.76) <0.001 1.58 (1.48, 1.68) <0.001
 80+ 2.66 (1.86, 3.79) <0.001 3.28 (2.82, 3.82) <0.001
Sex (Ref: Male)
 Female 0.62 (0.42, 0.92) 0.017 0.76 (0.67, 0.87) <0.001
 Unknown 1.91 (0.26, 14.21) 0.53 0.73 (0.22, 2.41) 0.61
Self-Reported Race (Ref: White)
 Black 0.71 (0.59, 0.85) <0.001 0.72 (0.67, 0.77) <0.001
 Other 0.82 (0.52, 1.30) 0.4 0.97 (0.82, 1.14) 0.69
Ethnicity (Ref: Non-Hispanic/Latino)
 Hispanic/Latine 0.94 (0.70, 1.26) 0.67 0.91 (0.82, 1.02) 0.11
 Unknown 0.95 (0.58, 1.55) 0.82 0.92 (0.74, 1.14) 0.45
Urban Status (Ref: Urban)
 Rural 1.01 (0.86, 1.18) 0.89 0.95 (0.90, 1.01) 0.13
 Unknown 0 (0.00, NA) 0.97 0.37 (0.05, 2.73) 0.33
Marital Status (Ref: Married)
 Single 0.85 (0.73, 0.99) 0.037 1.03 (0.97, 1.10) 0.28
 Unknown 0 (0.00, NA) 0.95 0.86 (0.43, 1.72) 0.68
Housing Status (Ref: Housed)
 Homeless 0.75 (0.63, 0.90) 0.002 0.85 (0.80, 0.91) <0.001
Year of Intervention
 Starting Year (Ref: 2011) 1.05 (1.03, 1.08) <0.001 1.02 (1.02, 1.03) <0.001
 Years Since Diagnosis 0.99 (0.97, 1.01) 0.45 1 (0.99, 1.01) 0.85
Length of Hospital Stay (Weeks)
Length of Stay 6.78 (4.85, 9.47) <0.001 3.45 (2.96, 4.01) <0.001
Service Connection (Ref: 1–49%)
No Service Related Disability 1.5 (1.20, 1.88) <0.001 1.19 (1.10, 1.28) <0.001
 50–100% 0.76 (0.60, 0.97) 0.025 0.64 (0.59, 0.70) <0.001
Charlson Comorbidity Index (Ref: 0)
 Score 1–5 2.11 (1.80, 2.47) <0.001 1.96 (1.85, 2.07) <0.001
 Score 5–10 4.15 (3.40, 5.06) <0.001 3.86 (3.57, 4.17) <0.001
 Score 10+ 6.23 (3.79, 10.26) <0.001 7.34 (5.90, 9.13) <0.001
Mental Health Diagnosis
 Depressive Disorder 0.92 (0.78, 1.09) 0.34 0.89 (0.84, 0.95) <0.001
 Anxiety Disorder 0.92 (0.80, 1.07) 0.29 0.95 (0.90, 1.01) 0.08
 Post-Traumatic Stress Disorder 1.13 (0.97, 1.33) 0.12 1.18 (1.11, 1.25) <0.001
 Bipolar Disorder 0.87 (0.72, 1.06) 0.16 0.88 (0.82, 0.94) <0.001
 Schizophrenia 0.98 (0.76, 1.27) 0.91 0.94 (0.85, 1.03) 0.2
Co-Occuring Substance Use Disorder
 Sedative Use Disorder 1.08 (0.85, 1.38) 0.52 1.13 (1.03, 1.24) 0.013
 Stimulant Use Disorder 0.95 (0.72, 1.24) 0.68 0.91 (0.82, 1.01) 0.076
 Multiple Use Disorder 1.81 (1.44, 2.27) <0.001 1.48 (1.36, 1.62) <0.001
 Alcohol Use Disorder 0.9 (0.76, 1.07) 0.22 0.98 (0.92, 1.05) 0.64
 Cannabis Use Disorder 0.94 (0.79, 1.12) 0.49 0.93 (0.87, 0.99) 0.032
 Cocaine Use Disorder 0.92 (0.74, 1.13) 0.42 1.02 (0.94, 1.11) 0.58
Medication Receipt
Buprenorphine Receipt Within 30 days Prior to Admission 0.7 (0.48, 1.04) 0.076 0.79 (0.69, 0.91) <0.001
Open in a new tab

Adjusted Odds Ratios are presented for each time interval along with 95% Confidence Intervals and associated p values. Green shading indicates reduced mortality. Red shading indicates increased mortality.

Other factors that were associated with increased all-cause mortality within both 30 days and 1 year of hospitalization include a later year of index hospitalization (30d: aOR 1.05 [95%CI 1.03,1.08]; 1y: aOR 1.02 [95%CI 1.02,1.03] per year), a longer index hospital stay (30d: aOR 6.78 [95%CI 4.85,9.47]; 1y: aOR 3.45 [95%CI 2.96,4.01] per additional week of hospitalization), and having one or more co-morbid substance use disorders in addition to OUD (30d: aOR 1.81 [95%CI 1.44,2.27]; 1y: aOR 1.48 [95%CI 1.36,1.62]). An increased degree of medical complexity, as assessed by the CCI, was associated with increased mortality within 30 days and one year in a ‘dose-dependent’ fashion (i.e., higher CCI scores were associated with increased aOR; all p<.001). A documented posttraumatic stress disorder diagnosis (aOR 1.18 [95%CI 1.11,1.25]) or sedative use disorder (aOR 1.13 [95%CI 1.03,1.24]) were associated with increased mortality within one year after hospitalization but not within 30d.

Factors Associated with Decreased All-Cause Mortality

Several patient-level factors were significantly associated with decreased all-cause mortality within 30 days and one year after medical hospitalization, as summarized in Table 2. These included female sex (30d: aOR 0.62 [95%CI 0.42,0.92]; 1y: aOR 0.76 [95%CI 0.67,0.87]), self-reported Black race (30d: aOR 0.69 [95%CI 0.57,0.85]; 1y: aOR 0.70 [95%CI 0.65,0.75]), and VA service connection >50% (30d: aOR 0.57 [95%CI 0.47,0.68]; 1y: aOR 0.59 [95%CI 0.55,0.64]).

Homelessness was also significantly associated with decreased all-cause mortality across both time intervals (30d: aOR 0.71 [95%CI 0.59,0.85]; 1y: aOR 0.72 [95%CI 0.67,0.77]). Buprenorphine receipt within 30d prior to index hospitalization (aOR 0.79 [95%CI 0.69,0.91]) and certain mental health diagnoses, including depressive disorder (aOR 0.89 [95%CI 0.82,0.94]), bipolar disorder (aOR 0.88 [95%CI 0.82,0.94]), and cannabis use disorder (aOR 0.93 [95%CI 0.87,0.99]) were significantly associated with reduced mortality at one year, but not at 30d after a medical hospitalization. Single marital status was associated with reduced all-cause mortality within 30d (aOR 0.85 [95%CI 0.73, 0.99]) but not one year.

Factors Associated with Increased Opioid-Related Mortality

Associations were different when the analysis was restricted to opioid-related death (Table 3). We observed a reversal in the positive association of age with all-cause mortality. Compared to the 41–64yo age group, the 25–40yo group was significantly associated with increased opioid-related mortality within both time intervals (30d: aOR 3.21 [95%CI 1.81,5.69]; 1y: aOR 2.38 [95%CI 1.97,2.88]), as was the 18–24yo group (30d: aOR 5.24 [95%CI 1.21,22.78]; 1y: aOR 2.71 [95%CI 1.53,4.79]). Homelessness (aOR 1.40 [95%CI 1.20,1.63]), single marital status (aOR 1.31 [95%CI 1.08,1.58]), length of hospital stay (aOR 1.79 [95%CI 1.18,2.72] per week), later year of index hospitalization (aOR 1.07 [95%CI 1.05,1.10] per year), posttraumatic stress disorder (PTSD; aOR 1.23 [95%CI 1.05,1.45]), cocaine use disorder (aOR 1.41 [95%CI 1.14, 1.73]), and having multiple substance use disorders (aOR 1.78 [95%CI 1.33,2.38]) were each associated with increased opioid-related mortality within one year but not within 30 days of hospital discharge. The strong association between increasing CCI and higher odds of all-cause mortality did not manifest for opioid-related causes of death.

Table 3.

Clinical and demographic characteristics associated with opioid-related* mortality within 30 days and 365 days of medical hospitalization.

30-day Mortality 365-day Mortality
Variable Odds Ratio 95% CI P-Value Odds Ratio 95% CI P-Value
Age (Reference: 41–64yo)
 18–24 5.24 (1.21, 22.78) 0.027 2.71 (1.53, 4.79) <0.001
 25–40 3.21 (1.81, 5.69) <0.001 2.38 (1.97, 2.88) <0.001
 65–80 0.81 (0.39, 1.70) 0.58 0.7 (0.55, 0.90) 0.006
 80+ 1.98 (0.26, 14.94) 0.51 0.37 (0.09, 1.51) 0.17
Sex (Ref: Male)
 Female 0.78 (0.31, 1.96) 0.6 0.99 (0.75, 1.30) 0.94
 Unknown 12.67 (1.64, 98.07) 0.015 1.27 (0.17, 9.38) 0.81
Self-Reported Race (Ref: White)
 Black 0.65 (0.36, 1.19) 0.16 0.61 (0.50, 0.74) <0.001
 Other 0.86 (0.21, 3.53) 0.83 0.77 (0.48, 1.24) 0.28
Ethnicity (Ref: Non-Hispanic/Latino)
 Hispanic/Latine 0.29 (0.07, 1.20) 0.088 0.62 (0.44, 0.88) 0.008
 Unknown 0.53 (0.09, 3.00) 0.47 0.83 (0.47, 1.48) 0.53
Urban Status (Ref: Urban)
 Rural 0.46 (0.25, 0.86) 0.015 0.56 (0.46, 0.68) <0.001
 Unknown 0 (0.00, NA) 0.99 0 (0.00, NA) 0.96
Marital Status (Ref: Married)
 Single 0.83 (0.49, 1.40) 0.49 1.31 (1.08, 1.58) 0.006
 Unknown 0 (0.00, NA) 0.98 1.41 (0.34, 5.87) 0.64
Housing Status (Ref: Housed)
 Homeless 1.1 (0.69, 1.76) 0.69 1.4 (1.20, 1.63) <0.001
Year of Intervention
 Starting Year (Ref: 2011) 1.04 (0.96, 1.13) 0.3 1.07 (1.05, 1.10) <0.001
Length of Hospital Stay (Weeks)
 Length of Stay 2.48 (0.76, 8.03) 0.13 1.79 (1.18, 2.72) 0.007
Service Connection (Ref: 1–49%)
  No Service-Related Disability 1.34 (0.72, 2.51) 0.36 1.06 (0.86, 1.30) 0.58
 50–100% 0.41 (0.21, 0.81) 0.01 0.53 (0.43, 0.66) <0.001
Charlson Comorbidity Index (Ref: 0)
 Score 1–5 1.14 (0.70, 1.85) 0.6 0.97 (0.83, 1.14) 0.75
 Score 5–10 1.01 (0.39, 2.60) 0.99 0.79 (0.56, 1.09) 0.15
 Score 10+ 6.34 (1.46, 27.43) 0.014 1.8 (0.79, 4.10) 0.16
Mental Health Diagnosis
 Depressive Disorders 1.17 (0.60, 2.29) 0.64 1.03 (0.84, 1.27) 0.79
 Anxiety Disorders 1.46 (0.87, 2.43) 0.15 1.03 (0.88, 1.21) 0.72
 Post-Traumatic Stress Disorder 1.3 (0.80, 2.10) 0.29 1.23 (1.05, 1.45) 0.011
 Bipolar Disorder 1.2 (0.74, 1.96) 0.46 1.15 (0.98, 1.36) 0.091
 Schizophrenia 0.83 (0.41, 1.68) 0.6 0.93 (0.74, 1.18) 0.57
Co-Occurring Substance Use Disorder
  Sedative Use Disorder 1.16 (0.59, 2.30) 0.67 1.12 (0.89, 1.40) 0.34
 Stimulant Use Disorder 1.16 (0.57, 2.36) 0.69 0.94 (0.74, 1.18) 0.58
 Multiple Use Disorders 2.25 (0.87, 5.88) 0.096 1.78 (1.33, 2.38) <0.001
  Alcohol Use Disorder 1.35 (0.74, 2.45) 0.33 1.02 (0.85, 1.23) 0.8
 Cannabis Use Disorder 0.75 (0.46, 1.22) 0.25 0.89 (0.76, 1.05) 0.15
  Cocaine Use Disorder 1.35 (0.71, 2.57) 0.36 1.41 (1.14, 1.73) 0.001
Medication Receipt
 Buprenorphine Receipt Within 30 days prior to Admission 0.33 (0.08, 1.37) 0.13 0.53 (0.37, 0.78) 0.001
Open in a new tab

Adjusted Odds Ratios are presented for each time interval along with p values. Green shading indicates reduced mortality. Red shading indicates increased mortality.

*

Opioid-related deaths are defined as underlying cause of death codes for drug poisoning (X40–X44, X60–X64, X85, Y10–Y14) and/or contributing cause of death code indicating opioid involvement (T40.0, T40.1, T40.2, T40.3, T40.4).

Factors Associated with Decreased Opioid-Related Mortality

Table 3 displays factors associated with decreased opioid-related mortality. Older age (65–80yo) was associated with reduced opioid-related mortality at one year (aOR 0.70 [95%CI 0.55,0.90]) but not at 30 days compared to a reference of 41–64yo. Self-reported Black race (aOR 0.61 [95%CI 0.50,0.74]), Latinx ethnicity (aOR 0.62 [95%CI 0.44,0.88]), and buprenorphine receipt (aOR 0.53 [95%CI 0.37, 0.78]) were associated with reduced opioid-related mortality within one year but not within 30 days. Higher level of service connection (30d: aOR 0.41 [95%CI 0.21, 0.81]; 1y: aOR 0.53 [95%CI 0.43, 0.66]) and rural designation (30d: aOR 0.46 [95%CI 0.25,0.86]; 1y: (aOR 0.56 [95%CI 0.46,0.68]) were associated with reduced opioid-related mortality within both time intervals. Certain factors that were associated with reduced all-cause mortality, such as homelessness and select mental health diagnoses, were not associated with reduced opioid-related mortality.

Discussion

Post-hospitalization care transitions are times of vulnerability and opportunity in caring for patients with OUD. Understanding which patient-level factors predict mortality risk following hospitalization may help to tailor care and improve health outcomes during this period. Using a national sample of Veterans with OUD who experienced at least one medical hospitalization between January 2011-December 2021, we identified several factors associated with both increased and decreased mortality in the post-discharge period. Among our total cohort, nearly 8% died within one year of their index hospitalization. This number is relatively low compared to regional studies of post-hospitalization mortality among patients with OUD (range 8%–31%).(Hochstatter et al., 2023; King et al., 2022) While little is known about post-hospitalization mortality among Veterans in general, mortality was lower in our sample than for other select groups within the VA system such as older adults and patients admitted with COVID-19 infection (range: 18%–21%),(Ohl et al., 2023; Rhodes et al., 2023) as well as hospitalized adults in non-VA community settings (range 7–12%).(van Walraven, 2013; van Walraven et al., 2015; Walter et al., 2001)

Several factors were associated with increased short-term (0–30 days) and long-term (0–365 days) all-cause mortality during the post-hospitalization period within our cohort. Positive associations for older age, higher medical complexity, longer length of hospital stay, and having multiple SUDs are unsurprising, though our results highlight the need to strongly weigh these factors into discharge planning for people with OUD. Factors associated with increased all-cause mortality within 30 days of discharge, such as age, medical complexity, having more than one substance use disorder, and length of hospital stay identify patient characteristics where intensive supports to facilitate rapid linkage to follow-up care are especially needed. Factors which were only associated with increased mortality at one year, such as PTSD, highlight instances where preserving continuity of overarching medical and mental healthcare is key to saving lives. Programs such as medical respite,(Bring et al., 2020; Doran et al., 2013) residential substance use treatment centers that are equipped to provide full-spectrum medical care,(Englander et al., 2018; Gelman et al., 2022) and supportive housing can help to provide additional substance use, medical, and social support for individuals with OUD during care transitions and should be prioritized for patients at highest risk of early death.

Patient factors that were associated with reduced all-cause mortality after hospitalization were more complex. Some of these factors, such as receipt of buprenorphine, female sex, and Black race align with broader national mortality statistics related to the overdose public health crisis during our study period,(Han et al., 2021; Kline et al., 2021; Prevention, 2022) although more recent trends show overdose deaths disproportionately increasing among Black Americans.(Furr-Holden et al., 2021; Larochelle et al., 2021) The association of homelessness and certain mental health conditions with reduced all-cause mortality after discharge was unexpected. We hypothesize that this may be due in part to the VA’s robust wraparound housing, mental health, and substance use treatment services.(Bennett et al., 2022; O’Toole et al., 2016) Hospitalization may be an entry point for Veterans to connect with these services, improving treatment access and psychosocial support in the immediate post-hospitalization period. Moreover, VA programs targeting homelessness might streamline and/or prioritize access to post-hospitalization psychosocial support for this group.(Chapman et al., 2024) These hypotheses are supported by the positive association of service connection - which is a marker of both eligibility to receive VA benefits and prioritized access to care - with reduced all-cause mortality. However, we recognize that administrative coding represents an imprecise way to characterize housing status that may disproportionately select for Veterans who were accessing housing-related and other psychosocial services. Additionally, Veterans with a high degree of service connection may represent a group that has greater health literacy and self-advocacy skills, which in themselves could be individual-level attributes that influence post-discharge survival. Further research using more rigorous assessment methods within both VA and non-VA health systems can serve to confirm our findings and evaluate the generalizability of these results.

We observed an interesting shift in factors associated with post-hospitalization mortality when we examined opioid-related death specifically. The association between age and increased mortality reversed, such that younger age was associated with a higher risk of opioid-related death. The protective associations of homelessness and certain mental health conditions were no longer present, with homelessness associated with increased mortality at one year. Medical complexity and hospital length of stay had far less influence than for all-cause mortality. These shifts illustrate a different patient phenotype that may be at highest risk of opioid-related death following hospitalization – younger, with fewer chronic health conditions, and with features such as PTSD, homelessness, and multiple SUD making outsized contributions to mortality risk.

The percentage of individuals who received prescribed naloxone from the VA was low, which could mean that those who received it had higher baseline overdose risk (i.e., selection bias). We also could not account for naloxone that was obtained either outside the VA system or that was dispensed without a prescription (e.g., through community emergency medical services or emergency departments), which likely underestimated naloxone receipt. These are consequential gaps in our data, as legislative action and increasing public awareness led to expanded access to naloxone through community programs during our study period.(McDonald & Strang, 2016; Puzantian & Gasper, 2018) The VA has also increasingly invested in overdose education and naloxone distribution initiatives.(Bounthavong et al., 2017; Oliva et al., 2017; Oliva et al., 2021) Future research using more recent data and accounting for receipt of naloxone outside the VA should investigate the impact of naloxone receipt on post-hospitalization mortality.

Two other findings that require further study are the association of rural designation with reduced opioid-related mortality and the lack of association between buprenorphine receipt and 30-day opioid-related mortality. For the former, it is possible that regional differences in drug composition (e.g., illicit fentanyl, xylazine) and efforts by the VA to expand access to rural healthcare access contributed to a lower overdose risk in rural areas. Conversely, it is possible that individuals in rural areas were more likely to seek care outside of the VA system, leading to unobserved factors that may have been protective such as accessing community-based harm reduction services and OUD treatment. Regarding buprenorphine receipt, the relatively small number of individuals who received buprenorphine may have limited our power to detect significant associations.

While not conclusive, our work has important implications for clinical care, program innovation, policymaking, and future research. First, further investigating the associations between homelessness and common mental health conditions with reduced all-cause mortality may add clarity to those findings and potentially help to highlight VA-specific interventions that are positively impacting health among at-risk Veterans. If effective, these interventions could be adapted to non-VA health settings and supported through policies that adequately reimburse their core components (e.g., supportive housing, low-barrier OUD treatment, medical respite, mental health services). Second, our research highlights factors associated with increased mortality following hospitalization that may lead to improved risk stratification, outpatient resource allocation, and care transition intervention development to support patients who are at highest risk of early death. Third, our study raises questions about how other factors, such as admitting hospital diagnosis and other features of the hospital course, may influence post-discharge outcomes, including outcomes such as non-fatal overdose and post-hospitalization care linkage. Finally, our study highlights the importance of providing medications to treat OUD like buprenorphine as a crucial intervention to save lives.

Limitations

Our study has several limitations. Our cohort was predominantly male and White, thus limiting generalizability. While the VA is the nation’s largest integrated health system, its service model may not be generalizable to health systems which rely on fee-for-service billing. Limiting our analysis to medical hospitalizations affects generalizability to other inpatient settings such as residential substance use treatment and psychiatric hospitalizations. Our data did not include methadone or extended-release naltrexone provision, thus providing an incomplete picture of pre-hospitalization OUD treatment. Our data also did not include care received outside of the VA system, which may have affected estimates of naloxone and buprenorphine receipt. Excluding individuals with ICD-10 codes for OUD in remission may have missed some individuals who were hospitalized in the context of a recent return to use. Moreover, relying on heath record data such as ICD codes to ascertain variables such as OUD diagnosis and homelessness is imprecise.(Hurley et al., 2025; Lagisetty et al., 2021; Osterhage et al., 2024) Finally, reliance on death certificate data may have led to inaccurate characterization of cause of death in some cases. Despite these limitations, we believe that our study represents a unique and important contribution to the care transition literature with direct implications for future research and innovation.

Conclusions

Among a national sample of hospitalized Veterans with OUD, several patient demographic (e.g., age, medical complexity) and clinical (e.g., homelessness, multiple substance use disorders) factors were associated with all-cause and opioid-related mortality after medical hospitalization. Factors that were associated with reduced all-cause mortality, such as homelessness, may represent key areas where hospitalization can serve as an impetus for treatment initiation and connection to community-based resources, though these findings require confirmation. Our study can inform innovation and future research exploring tailored interventions to improve care transitions for people with OUD.

Supplementary Material

1

Highlights.

  • In this retrospective cohort study of 90,920 Veterans with OUD hospitalized between Jan 2011-Dec 2021, older age, medical complexity, multiple substance use disorders, and length of hospitalization were associated with increased all-cause mortality following medical hospitalization.

  • Buprenorphine receipt and service connection were associated with reduced all-cause and opioid-related mortality.

  • Different patient phenotypes (e.g., older and more medically complex vs younger with more social instability) were at elevated risk for mortality from different causes immediately following hospitalization. This information can be used to tailor care transition supports in the post-hospitalization period.

  • Identifying homelessness and certain mental health conditions during a hospitalization may represent an important impetus for prioritized connection to services like supportive housing and mental healthcare, which could improve post-hospitalization outcomes. These results need verification in a more diverse sample.

ACKOWLEDGEMENTS:

The authors would like to acknowledge Hongwei Zhao for her guidance on statistical analysis. The Staff of the Center for the Clinical Trials Network in the National Institute on Drug Abuse National Drug Abuse Treatment Clinical Trials Network had an advisory role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. The Center for the Clinical Trials Network in the National Institute on Drug Abuse appointed members and coordinated meetings of the data safety monitoring board. This manuscript was reviewed and approved by the Publications Committee of the National Drug Abuse Treatment Clinical Trials Network. The views and opinions presented in this manuscript to do represent those of any funding organizations, including the NIDA CTN, the Veterans’ Health Administration, and Kaiser Permanente.

FUNDING SOURCES:

Funding for this study and analysis was provided by the National Institute on Drug Abuse under the following award 3UG1DA040316-04S3. MI is supported by a Career Development Award from the National Institute on Drug Abuse (K23DA062174-01). ATK is supported by a VA Health Services Research and Development (HSR&D) Career Development Award (CDA 21-209, 1 IK2 HX003531-01A2) and funding from the National Institute on Drug Abuse (1OT2DA061144-01). Infrastructure support for author IAB was provided, in part, by the Health Systems Node (HSN; NIH/NIDA 2UG1DA040314-11) of the National Institute on Drug Abuse Clinical Trials Network. Infrastructure support for author AJG was provided, in part, by the Greater Intermountain Node (GIN; NIH/NIDA 1UG1DA049444) of the National Institute on Drug Abuse Clinical Trials Network and the Department of Veterans Affairs Health Services Research and Development Service Informatics, Decision-Enhancement, and Analytic Sciences (IDEAS; CIN 13-414) Center of Innovation.

Footnotes

DECLARATION OF INTERESTS: MI is a section editor at JAMA Internal Medicine. Author AJG receives an honorarium for an online chapter on alcohol management in the perioperative period from the UpToDate online reference. In the last three years, AJG has been on the board of directors for the American Society of Addiction Medicine (ASAM), the Association for Multidisciplinary Education and Research in Substance use and Addiction (AMERSA), and the International Society of Addiction Journal Editors (ISAJE), all non-for-profit organizations; he is and was not remunerated for these activities. AJG receives remuneration from AMERSA, Inc. for being the Editor in Chief of their peer reviewed journal. AJS receives royalties as a section editor for UpToDate, Inc.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. (2023). In Calonge BN & Styka AN (Eds.), Review of the Department of Veterans Affairs Presumption Decision Process. 10.17226/27166 [DOI] [Google Scholar]
  2. Bennett AS, Guarino H, Britton PC, O’Brien-Mazza D, Cook SH, Taveras F, Cortez J, & Elliott L (2022). U.S. Military veterans and the opioid overdose crisis: a review of risk factors and prevention efforts. Ann Med, 54(1), 1826–1838. 10.1080/07853890.2022.2092896 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bounthavong M, Harvey MA, Wells DL, Popish SJ, Himstreet J, Oliva EM, Kay CL, Lau MK, Randeria-Noor PP, Phillips AG, & Christopher MLD (2017). Trends in naloxone prescriptions prescribed after implementation of a National Academic Detailing Service in the Veterans Health Administration: A preliminary analysis. J Am Pharm Assoc (2003), 57(2S), S68–S72. 10.1016/j.japh.2016.11.003 [DOI] [PubMed] [Google Scholar]
  4. Bring C, Kruse M, Ankarfeldt MZ, Brunes N, Pedersen M, Petersen J, & Andersen O (2020). Post-hospital medical respite care for homeless people in Denmark: a randomized controlled trial and cost-utility analysis. BMC Health Services Research, 20(1), 508. 10.1186/s12913-020-05358-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Callister C, Lockhart S, Holtrop JS, Hoover K, & Calcaterra SL (2022). Experiences with an addiction consultation service on care provided to hospitalized patients with opioid use disorder: a qualitative study of hospitalists, nurses, pharmacists, and social workers. Subst Abus, 43(1), 615–622. 10.1080/08897077.2021.1975873 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Chapman AB, Scharfstein D, Byrne T, Montgomery AE, Suo Y, Effiong A, Shorter C, Huebler S, Greene T, Tsai J, Gelberg L, Kertesz SG, & Nelson RE (2024). The effect of a Veterans Affairs rapid rehousing and homelessness prevention program on long-term housing instability. Health Serv Res, e14428. 10.1111/1475-6773.14428 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Charlson ME, Pompei P, Ales KL, & MacKenzie CR (1987). A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis, 40(5), 373–383. 10.1016/0021-9681(87)90171-8 [DOI] [PubMed] [Google Scholar]
  8. Deng H, Raheemullah A, Fenno LE, & Lembke A (2022). A telehealth inpatient addiction consult service is both feasible and effective in reducing readmission rates. Journal of addictive diseases, 1–8. 10.1080/10550887.2022.2090822 [DOI] [Google Scholar]
  9. Doran KM, Ragins KT, Gross CP, & Zerger S (2013). Medical respite programs for homeless patients: a systematic review. J Health Care Poor Underserved, 24(2), 499–524. 10.1353/hpu.2013.0053 [DOI] [PubMed] [Google Scholar]
  10. Englander H, Dobbertin K, Lind B, Nicolaidis C, Graven P, Dorfman C, & Korthuis P (2019). Inpatient Addiction Medicine Consultation and Post-Hospital Substance Use Disorder Treatment Engagement: a Propensity-Matched Analysis. Journal of general internal medicine. 10.1007/s11606-019-05251-9 [DOI] [Google Scholar]
  11. Englander H, Jones A, Krawczyk N, Patten A, Roberts T, Korthuis PT, & McNeely J (2022). A Taxonomy of Hospital-Based Addiction Care Models: a Scoping Review and Key Informant Interviews. J Gen Intern Med, 37(11), 2821–2833. 10.1007/s11606-022-07618-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Englander H, Priest KC, Snyder H, Martin M, Calcaterra S, & Gregg J (2019). A Call to Action: Hospitalists’ Role in Addressing Substance Use Disorder. J Hosp Med, 14, E1–E4. 10.12788/jhm.3311 [DOI] [Google Scholar]
  13. Englander H, Thakrar AP, Bagley SM, Rolley T, Dong K, & Hyshka E (2024). Caring for Hospitalized Adults With Opioid Use Disorder in the Era of Fentanyl: A Review. JAMA Intern Med, 184(6), 691–701. 10.1001/jamainternmed.2023.7282 [DOI] [PubMed] [Google Scholar]
  14. Englander H, Wilson T, Collins D, Phoutrides E, Weimer M, Korthuis PT, Calcagni J, & Nicolaidis C (2018). Lessons learned from the implementation of a medically enhanced residential treatment (MERT) model integrating intravenous antibiotics and residential addiction treatment. Subst Abus, 39(2), 225–232. 10.1080/08897077.2018.1452326 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Fine DR, Yu L, Triant VA, Baggett TP, & Metlay JP (2020). Baseline Factors Associated with Mortality in Patients Who Engaged in Buprenorphine Treatment for Opioid Use Disorder: a Cohort Study. J Gen Intern Med, 35(8), 2375–2382. 10.1007/s11606-020-05779-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Furr-Holden D, Milam AJ, Wang L, & Sadler R (2021). African Americans now outpace whites in opioid-involved overdose deaths: a comparison of temporal trends from 1999 to 2018. Addiction, 116(3), 677–683. 10.1111/add.15233 [DOI] [PubMed] [Google Scholar]
  17. Garnett MF MA (2024). Drug overdose deaths in the United States, 2003–2023. (NCHS Data Brief, no 522). [Google Scholar]
  18. Gelman SS, Stenehjem E, Foster RA, Tinker N, Grisel N, & Webb BJ (2022). A Novel Program to Provide Drug Recovery Assistance and Outpatient Parenteral Antibiotic Therapy in People Who Inject Drugs. Open Forum Infect Dis, 9(2), ofab629. 10.1093/ofid/ofab629 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Han B, Cotto J, Etz K, Einstein EB, Compton WM, & Volkow ND (2021). Methamphetamine Overdose Deaths in the US by Sex and Race and Ethnicity. JAMA Psychiatry, 78(5), 564–567. 10.1001/jamapsychiatry.2020.4321 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Harris MTH, Peterkin A, Bach P, Englander H, Lapidus E, Rolley T, Weimer MB, & Weinstein ZM (2021). Adapting inpatient addiction medicine consult services during the COVID-19 pandemic. Addict Sci Clin Pract, 16(1), 13. 10.1186/s13722-021-00221-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Hebert AH, & Hill AL (2024). Impact of opioid overdoses on US life expectancy and years of life lost, by demographic group and stimulant co-involvement: a mortality data analysis from 2019 to 2022. Lancet Reg Health Am, 36, 100813. 10.1016/j.lana.2024.100813 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Hochstatter KR, Nordeck C, Mitchell SG, Schwartz RP, Welsh C, & Gryczynski J (2023). Polysubstance use and post-discharge mortality risk among hospitalized patients with opioid use disorder. Prev Med Rep, 36, 102494. 10.1016/j.pmedr.2023.102494 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Hurley RW, Bland KT, Chaskes MD, Hill EL, & Adams MCB (2025). Diagnosis and coding of opioid misuse: a systematic scoping review and implementation framework. Pain Med, 26(7), 372–396. 10.1093/pm/pnaf019 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Incze MA, Huebler S, Baylis JD, Stofko A, Kelley AT, Binswanger IA, & Gordon AJ (2025). Mortality Among Veterans With Opioid Use Disorder After Medical Hospitalization. JAMA Intern Med. 10.1001/jamainternmed.2025.0356 [DOI] [Google Scholar]
  25. Incze MA, Huebler S, Chen D, Stofko A, Wu C, Baylis JD, Wells C, & Babbel D (2024). Hospitalists’ Attitudes and Experiences with Post-Hospitalization Care Transitions for Patients with Substance Use Disorders: A Cross Sectional Analysis. Journal of Addiction Medicine, In Press. [Google Scholar]
  26. Incze MA, Huebler S, Szczotka K, Grant S, Kertesz SG, & Gordon AJ (2024). Expert Panel Consensus on the Effectiveness and Implementation of Models to Support Posthospitalization Care Transitions for People With Substance Use Disorders. J Addict Med, 18(6), 696–704. 10.1097/ADM.0000000000001369 [DOI] [PubMed] [Google Scholar]
  27. Incze MA, Kelley AT, James H, Nolan S, Stofko A, Fordham C, & Gordon AG (2024). Post-Hospitalization Care Transition Strategies for Patients with Substance Use Disorders: A Narrative Review and Taxonomy. Journal of general internal medicine, In Press. [Google Scholar]
  28. Incze MA, Sehgal SL, Hansen A, Garcia L, & Stolebarger L (2023). Evaluation of a Primary Care-Based Multidisciplinary Transition Clinic for Patients Newly Initiated on Buprenorphine in the Emergency Department. Subst Abus, 44(3), 220–225. 10.1177/08897077231188592 [DOI] [PubMed] [Google Scholar]
  29. James H, Morgan J, Ti L, & Nolan S (2023). Transitions in care between hospital and community settings for individuals with a substance use disorder: A systematic review. Drug Alcohol Depend, 243, 109763. 10.1016/j.drugalcdep.2023.109763 [DOI] [PubMed] [Google Scholar]
  30. King C, Cook R, Korthuis PT, Morris CD, & Englander H (2022). Causes of Death in the 12 Months After Hospital Discharge Among Patients With Opioid Use Disorder. J Addict Med, 16(4), 466–469. 10.1097/ADM.0000000000000915 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Kline A, Mattern D, Cooperman N, Williams JM, Dooley-Budsock P, Foglia R, & Borys S (2021). Opioid overdose in the age of fentanyl: Risk factor differences among subpopulations of overdose survivors. Int J Drug Policy, 90, 103051. 10.1016/j.drugpo.2020.103051 [DOI] [PubMed] [Google Scholar]
  32. Krawczyk N, Miller M, Englander H, Rivera BD, Schatz D, Chang J, Cerda M, Berry C, & McNeely J (2024). Toward a Consensus on Strategies to Support Opioid Use Disorder Care Transitions Following Hospitalization: A Modified Delphi Process. J Gen Intern Med. 10.1007/s11606-024-09108-8 [DOI] [Google Scholar]
  33. Krawczyk N, Rivera BD, Jent V, Keyes KM, Jones CM, & Cerda M (2022). Has the treatment gap for opioid use disorder narrowed in the U.S.?: A yearly assessment from 2010 to 2019”. Int J Drug Policy, 110, 103786. 10.1016/j.drugpo.2022.103786 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Lagisetty P, Garpestad C, Larkin A, Macleod C, Antoku D, Slat S, Thomas J, Powell V, Bohnert ASB, & Lin LA (2021). Identifying individuals with opioid use disorder: Validity of International Classification of Diseases diagnostic codes for opioid use, dependence and abuse. Drug Alcohol Depend, 221, 108583. 10.1016/j.drugalcdep.2021.108583 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Larochelle MR, Bernson D, Land T, Stopka TJ, Wang N, Xuan Z, Bagley SM, Liebschutz JM, & Walley AY (2018). Medication for Opioid Use Disorder After Nonfatal Opioid Overdose and Association With Mortality: A Cohort Study. Ann Intern Med, 169(3), 137–145. 10.7326/M17-3107 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Larochelle MR, Slavova S, Root ED, Feaster DJ, Ward PJ, Selk SC, Knott C, Villani J, & Samet JH (2021). Disparities in Opioid Overdose Death Trends by Race/Ethnicity, 2018–2019, From the HEALing Communities Study. Am J Public Health, 111(10), 1851–1854. 10.2105/AJPH.2021.306431 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Marks LR, Munigala S, Warren DK, Liang SY, Schwarz ES, & Durkin MJ (2019). Addiction Medicine Consultations Reduce Readmission Rates for Patients With Serious Infections From Opioid Use Disorder. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America, 68(11), 1935–1937. 10.1093/cid/ciy924 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Marks LR, Munigala S, Warren DK, Liang SY, Schwarz ES, & Durkin MJ (2019). Addiction Medicine Consultations Reduce Readmission Rates for Patients With Serious Infections From Opioid Use Disorder. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America, 68(11), 1935–1937. 10.1093/cid/ciy924 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. McDonald R, & Strang J (2016). Are take-home naloxone programmes effective? Systematic review utilizing application of the Bradford Hill criteria. Addiction, 111(7), 1177–1187. 10.1111/add.13326 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. McNeely J, Wang SS, Rostam Abadi Y, Barron C, Billings J, Tarpey T, Fernando J, Appleton N, Fawole A, Mazumdar M, Weinstein ZM, Kalyanaraman Marcello R, Dolle J, Cooke C, Siddiqui S, & King C (2024). Addiction Consultation Services for Opioid Use Disorder Treatment Initiation and Engagement: A Randomized Clinical Trial. JAMA Intern Med, 184(9), 1106–1115. 10.1001/jamainternmed.2024.3422 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Moreno JL, Wakeman SE, Duprey MS, Roberts RJ, Jacobson JS, & Devlin JW (2019). Predictors for 30-Day and 90-Day Hospital Readmission Among Patients With Opioid Use Disorder. J Addict Med, 13(4), 306–313. 10.1097/ADM.0000000000000499 [DOI] [PubMed] [Google Scholar]
  42. O’Toole TP, Johnson EE, Aiello R, Kane V, & Pape L (2016). Tailoring Care to Vulnerable Populations by Incorporating Social Determinants of Health: the Veterans Health Administration’s “Homeless Patient Aligned Care Team” Program. Prev Chronic Dis, 13, E44. 10.5888/pcd13.150567 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Ober AJ, Murray-Krezan C, Page K, Friedmann PD, Anderson J, Osilla KC, Ryzewicz S, Huerta S, Mazer MW, Hoskinson RA, Garvey R, Peltz A, Watkins KE, Nuckols T, IsHak WW, Mariano LT, & Danovitch I (2025). Hospital Addiction Consultation Service and Opioid Use Disorder Treatment: The START Randomized Clinical Trial. JAMA Intern Med. 10.1001/jamainternmed.2024.8586 [DOI] [Google Scholar]
  44. Ober AJ, Osilla KC, Klein DJ, Burgette LF, Leamon I, Mazer MW, Messineo G, Collier S, Korouri S, Watkins KE, Ishak W, Nuckols T, & Danovitch I (2023). Pilot randomized controlled trial of a hospital-based substance use treatment and recovery team (START) to improve initiation of medication for alcohol or opioid use disorder and linkage to follow-up care. J Subst Use Addict Treat, 150, 209063. 10.1016/j.josat.2023.209063 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Ohl ME, Richardson Miell K, Beck BF, Mecham B, Bailey G, Mengeling M, & Vaughan-Sarrazin M (2023). Mortality Among US Veterans Admitted to Community vs Veterans Health Administration Hospitals for COVID-19. JAMA Netw Open, 6(5), e2315902. 10.1001/jamanetworkopen.2023.15902 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Oliva EM, Christopher MLD, Wells D, Bounthavong M, Harvey M, Himstreet J, Emmendorfer T, Valentino M, Franchi M, Goodman F, Trafton JA, Veterans Health Administration Opioid Overdose, E., Naloxone Distribution National, S., & Development, W. (2017). Opioid overdose education and naloxone distribution: Development of the Veterans Health Administration’s national program. J Am Pharm Assoc (2003), 57(2S), S168–S179 e164. 10.1016/j.japh.2017.01.022 [DOI] [PubMed] [Google Scholar]
  47. Oliva EM, Richardson J, Harvey MA, & Bellino P (2021). Saving Lives: The Veterans Health Administration (VHA) Rapid Naloxone Initiative. Jt Comm J Qual Patient Saf, 47(8), 469–480. 10.1016/j.jcjq.2021.06.004 [DOI] [PubMed] [Google Scholar]
  48. Osterhage KP, Hser YI, Mooney LJ, Sherman S, Saxon AJ, Ledgerwood M, Holtzer CC, Gehring MA, Clingan SE, Curtis ME, & Baldwin LM (2024). Identifying patients with opioid use disorder using International Classification of Diseases (ICD) codes: Challenges and opportunities. Addiction, 119(1), 160–168. 10.1111/add.16338 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Prevention, C. f. D. C. a. (2022). vital statistics rapid release - provisional drug overdose data. https://www.cdc.gov/nchs/nvss/vsrr/drug-overdose-data.htm [Google Scholar]
  50. Puzantian T, & Gasper JJ (2018). Provision of Naloxone Without a Prescription by California Pharmacists 2 Years After Legislation Implementation. JAMA, 320(18), 1933–1934. 10.1001/jama.2018.12291 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Rhodes RL, Keller HE, Lensing S, Padala KP, Padala PR, Brown LM, Roberson PK, & Sullivan DH (2023). The influence of social determinants of health on post-discharge mortality of veterans who received care in a transitional care unit. Journal of the American Geriatrics Society, 71(10), 3299–3303. 10.1111/jgs.18480 [DOI] [PubMed] [Google Scholar]
  52. Salvalaggio G, Dong KA, Hyshka E, McCabe C, Nixon L, Rosychuk RJ, Dmitrienko K, Krajnak J, Mrklas K, & Wild TC (2022). Impact of an addiction medicine consult team intervention in a Canadian inner city hospital on acute care utilization: a pragmatic quasi-experimental study. Substance abuse treatment, prevention, and policy, 17(1), 20. 10.1186/s13011-022-00445-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Smith A, Hansen J, & Colvard M (2021). Impact of a pharmacist-led substance use disorder transitions of care clinic on postdischarge medication treatment retention. Journal of Substance Abuse Treatment, 130, 108440. 10.1016/j.jsat.2021.108440 [DOI] [PubMed] [Google Scholar]
  54. Snow RL, Simon RE, Jack HE, Oller D, Kehoe L, & Wakeman SE (2019). Patient experiences with a transitional, low-threshold clinic for the treatment of substance use disorder: A qualitative study of a bridge clinic. J Subst Abuse Treat, 107, 1–7. 10.1016/j.jsat.2019.09.003 [DOI] [PubMed] [Google Scholar]
  55. Suen LW, Makam AN, Snyder HR, Repplinger D, Kushel MB, Martin M, & Nguyen OK (2022). National Prevalence of Alcohol and Other Substance Use Disorders Among Emergency Department Visits and Hospitalizations: NHAMCS 2014–2018. J Gen Intern Med, 37(10), 2420–2428. 10.1007/s11606-021-07069-w [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Taylor JL, Wakeman SE, Walley AY, & Kehoe LG (2023). Substance use disorder bridge clinics: models, evidence, and future directions. Addict Sci Clin Pract, 18(1), 23. 10.1186/s13722-023-00365-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Trowbridge P, Weinstein ZM, Kerensky T, Roy P, Regan D, Samet JH, & Walley AY (2017). Addiction consultation services – Linking hospitalized patients to outpatient addiction treatment. Journal of Substance Abuse Treatment, 79, 1–5. 10.1016/j.jsat.2017.05.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Tsai J, Szymkowiak D, & Jutkowitz E (2022). Developing an operational definition of housing instability and homelessness in Veterans Health Administration’s medical records. PLoS One, 17(12), e0279973. 10.1371/journal.pone.0279973 [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. van Walraven C (2013). Trends in 1-year survival of people admitted to hospital in Ontario, 1994–2009. CMAJ, 185(16), E755–762. 10.1503/cmaj.130875 [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. van Walraven C, McAlister FA, Bakal JA, Hawken S, & Donze J (2015). External validation of the Hospital-patient One-year Mortality Risk (HOMR) model for predicting death within 1 year after hospital admission. CMAJ, 187(10), 725–733. 10.1503/cmaj.150209 [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Vandenbroucke JP, von Elm E, Altman DG, Gotzsche PC, Mulrow CD, Pocock SJ, Poole C, Schlesselman JJ, Egger M, & initiative S (2007). Strengthening the Reporting of Observational Studies in Epidemiology (STROBE): explanation and elaboration. Ann Intern Med, 147(8), W163–194. 10.7326/0003-4819-147-8-200710160-00010-w1 [DOI] [PubMed] [Google Scholar]
  62. Wakeman SE, Metlay JP, Chang Y, Herman GE, & Rigotti NA (2017). Inpatient Addiction Consultation for Hospitalized Patients Increases Post-Discharge Abstinence and Reduces Addiction Severity. Journal of general internal medicine, 32(8), 909–916. 10.1007/s11606-017-4077-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Walter LC, Brand RJ, Counsell SR, Palmer RM, Landefeld CS, Fortinsky RH, & Covinsky KE (2001). Development and validation of a prognostic index for 1-year mortality in older adults after hospitalization. JAMA, 285(23), 2987–2994. 10.1001/jama.285.23.2987 [DOI] [PubMed] [Google Scholar]
  64. Weinstein ZM, Cheng DM, D’Amico MJ, Forman LS, Regan D, Yurkovic A, Samet JH, & Walley AY (2020). Inpatient addiction consultation and post-discharge 30-day acute care utilization. Drug Alcohol Depend, 213, 108081. 10.1016/j.drugalcdep.2020.108081 [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Weinstein ZM, Wakeman SE, & Nolan S (2018). Inpatient Addiction Consult Service: Expertise for Hospitalized Patients with Complex Addiction Problems. The Medical Clinics of North America, 102(4), 587–601. 10.1016/j.mcna.2018.03.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Wilson JD, Altieri Dunn SC, Roy P, Joseph E, Klipp S, & Liebschutz J (2022). Inpatient Addiction Medicine Consultation Service Impact on Post-discharge Patient Mortality: a Propensity-Matched Analysis. Journal of general internal medicine. 10.1007/s11606-021-07362-8 [DOI] [Google Scholar]

Associated Data

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

Supplementary Materials

1
NIHMS2109208-supplement-1.docx (24.5KB, docx)

RESOURCES