Psychology of consulting and coaching

Justice Involvement and Treatment Use Among Rural Veterans

Andrea K. Finlay Veterans Affairs Palo Alto Health Care System,

Menlo Park, California, and Department of Veterans Affairs, Menlo Park, California

Alex H. S. Harris Veterans Affairs Palo Alto Health Care System, Menlo Park, California, and Stanford University

School of Medicine

Joel Rosenthal Department of Veterans Affairs,

Menlo Park, California

Jessica Blue-Howells Department of Veterans Affairs,

Los Angeles, California

Sean Clark Department of Veterans Affairs,

Lexington, Kentucky

Bessie Flatley Department of Veterans Affairs,

Philadelphia, Pennsylvania

Christine Timko Veterans Affairs Palo Alto Health Care System, Menlo Park, California, and Stanford University School

of Medicine

Veterans in rural areas and veterans involved in the criminal justice system have experienced less access to, and use of, health care. However, there is limited informa- tion on the treatment needs and health care access of justice-involved veterans who live in rural areas. This study used national Veterans Health Administration data from Fiscal Year 2014 to examine the interactive effect of rural/urban residence and justice involvement on mental health and substance use disorder treatment entry among veterans diagnosed with mental health or substance use disorders. Of veterans residing in rural areas, 15,328 (2.5%) were justice-involved and 607,819 (97.5%) veterans had no known justice involvement. Among veterans in urban areas, 54,525 (4.3%) were justice-involved and 1,213,766 (95.7%) had no known justice involvement. Rural

Andrea K. Finlay, Center for Innovation to Implementa- tion, Veterans Affairs Palo Alto Health Care System, Menlo Park, California, and National Center on Homelessness Among Veterans, Department of Veterans Affairs, Menlo Park, California; Alex H. S. Harris, Center for Innovation to Implementation, Veterans Affairs Palo Alto Health Care Sys- tem, and Department of Surgery, Stanford University School of Medicine; Joel Rosenthal, Veterans Justice Programs, De- partment of Veterans Affairs, Menlo Park, California; Jessica Blue-Howells, Veterans Justice Programs, Department of Veterans Affairs, Los Angeles, California; Sean Clark, Vet- erans Justice Programs, Department of Veterans Affairs, Lex- ington, Kentucky; Bessie Flatley, National Center on Home- lessness Among Veterans, Department of Veterans Affairs, Philadelphia, Pennsylvania; Christine Timko, Center for In- novation to Implementation, Veterans Affairs Palo Alto Health Care System, and Department of Psychiatry and Be- havioral Sciences, Stanford University School of Medicine.

Joel Rosenthal is now retired. Andrea Finlay was funded by a Veterans Affairs

Health Services Research & Development (VA HSR&D) Career Development Award (CDA 13-279). Alex Harris and Christine Timko were funded by VA HSR&D Research Career Scientist Awards (RCS 14- 232 and Senior RCS 00-001, respectively). All authors were employed and funded by the Department of Vet- erans Affairs when this study was conducted. The views expressed are those of the authors and do not represent the position or policy of the Department of Veterans Affairs or the United States Government.

Correspondence concerning this article should be ad- dressed to Andrea K. Finlay, Center for Innovation to Implementation, Veterans Affairs Palo Alto Health Care System, 795 Willow Road (MPD-152), Menlo Park, CA 94025. E-mail: [email protected]

Journal of Rural Mental Health In the public domain 2018, Vol. 42, No. 1, 46 –59

http://dx.doi.org/10.1037/rmh0000092

46

residence was associated with lower access to mental health or substance use disorder treatment in the non-justice-involved sample but not the justice-involved sample. Justice-involvement was associated with higher odds of entering mental health or substance use disorder treatment for veterans in rural and in urban areas. Substance use disorder treatment entry could be improved for all veterans, particularly rural veterans who are not justice-involved. Telehealth, outreach services, and integrated treatment may help address challenges to treatment entry experienced by rural veterans and justice-involved veterans.

Keywords: rural health, criminal justice, mental disorders, substance use disorders, mental health services

Supplemental materials: http://dx.doi.org/10.1037/rmh0000092.supp

Among the veteran population in the United States, 5.2 million—roughly 24%—live in rural areas, with 2.9 million rural veterans (56%) enrolled in Veterans Health Administration (VHA) care (Department of Veterans Affairs, 2017). Although prior literature has focused on health disparities among rural veterans com- pared with their urban counterparts (Kondo et al., 2017), rural veterans involved in the crimi- nal justice system are a largely understudied population that may have heightened risk for poor access and use of treatment services. Jus- tice-involved veterans—military veterans de- tained by or under the supervision of the crim- inal justice system—who live in rural areas have lower odds of engaging in mental health treatment, and of entering and engaging in ad- diction treatment, compared with their urban counterparts (Finlay, Smelson, et al., 2016; Fin- lay et al., 2017). However, these veterans were not compared with other rural veterans with similar conditions who were not involved in the criminal justice system. Thus, our understand- ing of the treatment needs of justice-involved veterans in the context of the broader population of veterans served at VHA is limited. The cur- rent study aims to examine the intersection of rural residence status with justice status among veterans diagnosed with mental health or sub- stance use disorders who sought treatment in VHA facilities.

Broader trends within VHA may explain less use of mental health and substance use disorder treatment observed among justice-involved vet- erans in rural areas compared with their urban counterparts; however, research findings on dif- ferences in treatment use between rural and urban veterans are mixed. Rural veterans have lower odds of receiving mental health services

compared with urban veterans (Mott, Grubbs, Sansgiry, Fortney, & Cully, 2015; Teich, Ali, Lynch, & Mutter, 2017). Among veterans who were part of VHA intensive mental health care services, veterans who lived in isolated rural areas received less rehabilitation, psychother- apy, and crisis intervention services than urban veterans in the same program (Mohamed, Neale, & Rosenheck, 2009). Receipt of pharma- cotherapy for alcohol or opioid use disorders was lower among veterans in rural areas com- pared with urban areas, regardless of justice involvement (Finlay, Binswanger, et al., 2016; Finlay, Harris, et al., 2016). Other research has found that, compared with urban veterans, more rural veterans had at least one mental health treatment visit (Johnson et al., 2015). Among veterans who served in recent conflicts in Iraq or Afghanistan, rural veterans had higher odds of receiving pharmacological or psychotherapy treatments for posttraumatic stress disorder (PTSD) or depression than urban veterans (Hudson et al., 2014). Although not directly compared with urban residents, among rural residents who used stimulants, less than 20% received substance abuse treatment in the 2-year period after study baseline (Carlson et al., 2010). Another study found similarly low rates of addiction treatment use among rural residents who used stimulants (Curran, Ounpraseuth, Allee, Small, & Booth, 2011). These studies suggest that rural residents face barriers to ad- diction treatment.

Studies of rural veterans and other rural res- idents have identified barriers to treatment. For example, bureaucratic difficulties when enroll- ing in the VHA system and a lack of under- standing of eligibility requirements were cited by women veterans in rural areas as barriers to

47RURAL AND JUSTICE-INVOLVED VETERANS

treatment in general (Brooks, Dailey, Bair, & Shore, 2016). Travel distances (Brooks et al., 2016), stigma or a lack of privacy to seek treat- ment (Browne et al., 2016), a mistrust of health care providers (Goins, Williams, Carter, Spen- cer, & Solovieva, 2005), and a lack of perceived need for substance use disorder treatment (Bor- ders, Booth, Stewart, Cheney, & Curran, 2015) have also been identified as barriers for rural veterans. Finally, health care system barriers such as a lack of treatment options or integrated services that address the array of patients’ needs (Browne et al., 2016; Sigmon, 2014), and a dearth of primary care physicians to provide office-based pharmacotherapy for addiction treatment (Quest, Merrill, Roll, Saxon, & Rosenblatt, 2012; Rosenblatt, Andrilla, Catlin, & Larson, 2015), may explain lower rates of treatment access and use in rural areas. Among women veterans in rural areas, mental health treatment was the second most frequently re- quested local service (Brooks et al., 2016).

Many of the challenges faced by rural and urban residents in accessing health care may be further exacerbated by justice involvement. For example, public transportation or transportation services for medical care are limited in rural areas (Buzza et al., 2011; Goins et al., 2005). Therefore, veterans who have lost their driver’s license because of criminal-justice-related is- sues may struggle to keep appointments. In ur- ban areas, adults who were referred by the crim- inal justice system to outpatient methadone treatment had higher odds of delayed admission to treatment than adults who self-referred (Gry- czynski, Schwartz, Salkever, Mitchell, & Jaffe, 2011), and in rural areas where substance use disorder treatment options are more limited than in urban areas (Sigmon, 2014), waitlists may be even longer. Mistrust of the health care system or health care professionals has been cited by both justice-involved populations (Howerton et al., 2007; Wainwright, McDonnell, Lennox, Shaw, & Senior, 2017) and rural populations (Goins et al., 2005) as a barrier to treatment seeking. Whether these negative attitudes are especially prominent among rural residents with justice involvement is unknown. Criminal- justice-related barriers to treatment cited by those on parole include previous incarceration having interrupted treatment or administrative/ bureaucratic processes, stigma, and poor treat-

ment by health care providers (Marlow, White, & Chesla, 2010).

Criminal justice mandates or legal supervi- sion can also affect treatment entry, though re- sults are mixed as to how. Rural residents who used drugs and had a court mandate to attend treatment had 3 times higher odds of entering substance use disorder treatment than residents without a mandate (Oser et al., 2011). In a national sample, patients referred from the criminal justice system to substance use disor- der treatment had higher odds of waiting more than 1 month for treatment compared with those self-referred (Andrews, Shin, Marsh, & Cao, 2013). Receipt of mental health or addiction treatment services is relatively low among crim- inally justice-involved samples, but there are noteworthy limitations to the study methodolo- gies. For example, in a nationally representative sample, people on probation or parole were more likely to have received substance use dis- order treatment in the past 12 months compared with the general population; however, only 28% of people on probation or parole had received alcohol or drug treatment in their lifetime (Vaughn, DeLisi, Beaver, Perron, & Abdon, 2012). Furthermore, it is unclear whether the results were limited to participants with an ac- tive mental health or substance use disorder diagnosis, so appropriate use of treatment for this sample is unclear. Similarly, a longitudinal study of rural residents on probation or parole found that 69% of the sample did not use mental health services and 61% did not use addiction treatment services over 1.5 years (Mowbray, McBeath, Bank, & Newell, 2016), but is it unclear whether the analyses were limited to participants with a mental health or substance use disorder diagnosis.

Among rural residents with alcohol use dis- order, some, but not all, types of justice involve- ment were associated with treatment entry. Be- ing on probation and being arrested and booked for drug-related charges or driving under the influence (DUI) were significantly associated with any alcohol or drug treatment received, but being on parole or being arrested and booked for other crimes (e.g., property, crimes against people, drunkenness) were not significantly linked (Booth, Curran, Han, & Edlund, 2013). Of note, 62% of the sample with justice in- volvement did not enter, or perceive a need for, treatment, suggesting that many justice-

48 FINLAY ET AL.

involved adults with alcohol use disorder are not getting needed treatment. Previous research indicated that justice-involved veterans were less likely to enter mental health and substance use disorder treatment at VHA facilities than homeless veterans (McGuire, Rosenheck, & Kasprow, 2003), and veterans with a prison history had lower odds of receiving pharmaco- therapy for opioid use disorder than veterans not justice-involved (Finlay, Harris, et al., 2016).

Rural residence and justice involvement have been examined separately in relation to dispar- ities in mental health or addiction treatment use. The literature is mixed for rural versus urban veterans and justice-involved versus non- justice-involved veterans. Furthermore, there are numerous barriers to treatment for both rural and justice-involved veterans, but it is unknown whether these factors interact and whether vet- erans who live in rural areas and are justice- involved have the most difficulty accessing mental health or substance use disorder treat- ment.

Current Study

To determine their potential multiplicative effects, we examined whether residence status and justice-involved status were associated with entry into VHA mental health and substance use disorder treatment among all veterans at VHA with mental health or substance use disorder diagnoses. Based on previous literature indicat- ing disparities in treatment use (Kondo et al., 2017), we hypothesized that veterans in urban areas would have higher odds of entering treat- ment compared with veterans in rural areas. We hypothesized that justice-involved veterans would have lower odds of entering treatment than other veterans, because most veterans (83%) are incarcerated for crimes other than drug- or DUI-related charges (Bronson, Carson, Noonan, & Berzofsky, 2015), and previous re- search indicated that justice-involved veterans are less likely to enter VHA care than other veterans (Finlay, Harris, et al., 2016; McGuire et al., 2003). Finally, we also hypothesized an interaction effect such that the negative effect of justice involvement would be even more pro- nounced in rural areas compared with urban areas.

Method

Participants and Data Source

This is a retrospective observational study of veterans who used VHA health care services. The VHA Corporate Data Warehouse, which is a national database of electronic health records, was used to identify patients who received a mental health or substance use disorder diagno- sis (excluding those in remission) during an outpatient or inpatient visit at a VHA facility in Fiscal Year 2014 (October 1, 2013 through Sep- tember 30, 2014); these patients were included in the study. Justice involvement, rural or urban residence, mental health and substance use dis- order diagnoses, and treatment use were identi- fied from the Corporate Data Warehouse data- base and from the Homeless Outreach Management and Evaluation System database, which includes information collected by VHA staff during homeless and justice outreach ser- vices. This study was approved by the Stanford University Institutional Review Board and the Veterans Affairs Palo Alto Research and Devel- opment committee.

Measures

Justice involvement. The VHA has two programs for justice-involved veterans. The Health Care for Reentry Veterans (HCRV) pro- gram conducts outreach with veterans in pris- ons, and the Veterans Justice Outreach (VJO) program conducts outreach in other criminal justice settings, primarily jails and courts (Blue- Howells, Clark, van den Berk-Clark, & McGuire, 2013). VHA staff members who pro- vide this outreach serve a variety of functions, including helping to facilitate VHA benefits pa- perwork, identifying and scheduling treatment at VHA facilities, and assistance in navigating the complex VHA care delivery and benefits system. In the VHA, clinic stop codes are three- digit codes that indicate the type of outpatient care and record workload that occurred during a VHA outpatient visit. Veterans were considered justice-involved if they received a 591 clinic code, indicating contact with the HCRV pro- gram, or a 592 clinic code, indicating contact with the VJO program, prior to or during Fiscal Year 2014. Clinic codes were drawn from the Corporate Data Warehouse database. In addi-

49RURAL AND JUSTICE-INVOLVED VETERANS

tion, veterans who had a Homeless Outreach Management and Evaluation System record in- dicating contact with the HCRV program or the VJO program also were considered justice- involved. Veterans with no known contact with HCRV or VJO are referred to in this article as “other veterans.”

Mental health or substance use disorder diagnoses. Patients who received Interna- tional Classification of Diseases, Ninth Revi- sion, Clinical Modification (ICD-9-CM; Medi- code (Firm), 1996) diagnosis codes indicating mental health or substance use disorders were included in the study. Mental health diagnoses were grouped into the following categories: de- pressive disorders (296.2, 296.3, 301.12, 300.4, 293.83, 298.0, 311.0, 311.9, 296.9, 309.0, 309.1), PTSD (309.81), anxiety disorders (300. 02, 300.0), bipolar disorder (296.x, except 296.2, 296.3, and 296.9), schizophrenia (295.x, except 295.5), other mental health disorders (297.0, 297.1, 297.2, 297.3, 297.8, 297.9, 298.1, 298.2, 298.3, 298.4, 298.8, 298.9), or personal- ity disorders (301.x, except 301.12). Substance use disorder diagnoses were grouped into the following categories: alcohol use disorder (291.x, 303.0, 303.9, 303.91, 303.92, 305.0, 305.01, 305.02), opioid use disorder (304.0, 304.7, 305.5), cocaine use disorder (304.2, 305.6), amphetamine use disorder (304.4, 305.7), cannabis use disorder (304.3, 305.2), sedative use disorder (304.1, 305.4), or other drug use disorders (292.0, 292.1, 292.11, 292. 12, 292.2, 292.81, 292.82, 282.83, 292.84, 292. 85, 292.89, 292.90, 304.5, 304.6, 304.8, 304.9, 305.3, 305.8, 305.9).

Residence. Rural or urban residence was drawn from the Corporate Data Warehouse files based on the patient’s last known address. Type of residence was defined using the U.S. Census Bureau definition (Ratcliffe, Burd, Holder, & Fields, 2016). Urban areas were defined as hav- ing an urban core of at least 1,000 residents per square mile or 50,000 or more people in the urban nucleus. Rural areas were defined as non- urban areas.

Treatment use. All treatment use oc- curred within the VHA health care system. For veterans diagnosed with a mental health disorder, we counted the number of mental health outpatient visits, mental health inpa- tient days, and mental health residential days recorded in their electronic health record in

the 1-year period following their first mental health diagnosis. For veterans diagnosed with depressive disorders, PTSD, bipolar disor- ders, and schizophrenia, we examined phar- macy records to determine whether veterans received pharmacotherapy for their diagnosed conditions. Based on the 2010 VHA clinical guidelines, we included only medications that were recommended as first line (likely to be beneficial) or second line (some benefit or trade-off between benefit and harm). There- fore, we did not calculate receipt of pharma- cotherapy for other mental health disorders or personality disorders because there were no medications recommended that met our crite- ria. Entry to mental health treatment was coded as any use of mental health outpatient, inpatient, or residential care or any receipt of pharmacotherapy for depressive disorders, PTSD, bipolar disorder, or schizophrenia in the 1-year period following first mental health diagnosis.

For veterans diagnosed with a substance use disorder, we counted the number of sub- stance use disorder outpatient visits and sub- stance use disorder residential days recorded in their electronic health record in the 1-year period following their first substance use dis- order diagnosis. For veterans diagnosed with alcohol or opioid use disorder, we examined pharmacy records to determine whether vet- erans received pharmacotherapy for their di- agnosed conditions. As with mental health disorders, medications for alcohol or opioid use disorder were selected using VHA clinical guidelines and were limited to first or second line medications. We did not calculate receipt of pharmacotherapy for any of the other sub- stance use disorder diagnoses because there were no medications that met our criteria for inclusion. Entry to substance use disorder treatment was coded as any use of substance use disorder outpatient or residential care or any receipt of pharmacotherapy for alcohol or opioid use disorders in the 1-year period fol- lowing first substance use disorder diagnosis.

Patient characteristics. Demographic variables included gender, age, race/ethnicity (Hispanic, non-Hispanic: American Indian/ Alaskan Native, Asian, Black, White; based on the U.S. Census Bureau categories), marital sta- tus (single, married, separated/divorced, wid- owed), and homeless status (drawn from a

50 FINLAY ET AL.

homeless indicator variable, receipt of housing services, and ICD-9 codes for unstable housing and homelessness), which indicates a person is at risk for or experiencing homelessness or re- ceiving homeless services. Military-related characteristics included service in Iraq or Af- ghanistan (yes/no) and service-connected dis- ability rating (none, �50%, �50%), which re- flects a VHA-determined disability occurring during or aggravated by military service. De- mographic and military characteristics were coded from records collected the same day the veteran received her or his first mental health or substance use disorder diagnosis during Fiscal Year 2014, or if missing, then was drawn from the next health care visit for which that infor- mation was available. Service in Iraq or Af- ghanistan was drawn from the Iraq/Afghanistan Roster, which is a database of veterans who served in conflicts in Iraq or Afghanistan and enrolled in VHA care. The Deyo Comorbidity Index, a sum of up to 17 comorbid medical diagnoses, such as HIV, liver disease, diabetes, and congestive heart failure (Deyo, Cherkin, & Ciol, 1992), was coded from records from the 1-year period after the first mental health or substance use disorder diagnosis and was a proxy for overall physical health.

Data Analysis

We examined descriptive statistics of patient characteristics and the proportion of patients diagnosed with mental health or substance use disorders, stratified by rural or urban residence and justice status. We then conducted two mixed-effects logistic regression models, with main effects for residence and justice status and the interaction of the two variables in each model, adjusting for all other patient character- istics and including a random effect for facility (N � 141 VHA medical centers). To examine entry into mental health treatment as the depen- dent variable, we limited the sample to veterans who were diagnosed with a mental health dis- order and examined justice-involved veterans compared with other veterans (Model 1). For entry into substance use disorder treatment as the dependent variable, we limited the sample to veterans who were diagnosed with a substance use disorder and compared justice-involved vet- erans with other veterans (Model 2). Because the interaction effects were significant for most

models, to facilitate interpretation, we con- ducted two additional mixed-effects logistic re- gression models focused only on rural veterans. We examined the association between justice involvement and entry into mental health or substance use disorder treatment, adjusting for all other patient characteristics and included a random effect for facility.

Results

Patient Characteristics

Of 623,147 rural VHA patients in Fiscal Year 2014 who were diagnosed with a mental health or substance use disorder, 15,328 (2.5%) were justice-involved and 607,819 (97.5%) veterans had no known justice-related contact. Of 1,268,291 urban VHA patients in Fiscal Year 2014 who were diagnosed with a mental health or substance use disorder, 54,525 (4.3%) were justice-involved and 1,213,766 (95.7%) veter- ans had no known justice-related contact. In general, justice-involved veterans were younger, had higher rates of homelessness, and had fewer comorbid medical conditions com- pared with other veterans (see Table 1). More justice-involved veterans served in Iraq or Af- ghanistan compared with other veterans.

Mental health and substance use disorder di- agnoses for the sample are reported in Table 2. Among veterans with mental health or sub- stance use disorders, depression was the most common mental health disorder, with 66% of rural justice-involved veterans, 62% of rural other veterans, 63% of urban justice-involved veterans, and 61% of urban other veterans with this diagnosis. PTSD and anxiety were the sec- ond and third most common mental health dis- orders. Less than 20% of veterans in the sample were diagnosed with bipolar disorder, schizo- phrenia, other mental health disorders, or per- sonality disorders. Among veterans with sub- stance use disorders, alcohol use disorder was the most common condition with 49% of rural justice-involved veterans, 17% of rural other veterans, 52% of urban justice-involved veter- ans, and 19% of urban other veterans with this diagnosis. Other drug use disorders were the second most common group of substance use disorders for all veterans. Less than 20% of justice-involved veterans (both rural and urban) and less than 6% of other veterans (rural and

51RURAL AND JUSTICE-INVOLVED VETERANS

urban) were diagnosed with opioid, cocaine, amphetamine, cannabis, or sedative use disor- ders.

Treatment Entry

Among veterans in rural areas who were diagnosed with a mental health disorder, 93% of justice-involved veterans and 80% of other veterans entered mental health treatment. Among veterans in urban areas who were diagnosed with a mental health disorder, 94%

of justice-involved veterans and 83% of other veterans entered mental health treatment. Among veterans in rural areas who were di- agnosed with a substance use disorder, 55% of justice-involved veterans and 20% of other veterans entered substance use disorder treat- ment. Among veterans in urban areas who were diagnosed with a substance use disorder, 59% of justice-involved veterans and 28% of other veterans entered substance use disorder treatment.

Table 1 Patient Characteristics of Veterans With Mental Health or Substance Use Disorders by Residence and Justice Status in Fiscal Year 2014

Characteristics

Rural Urban

Justice-involved (n � 15,328)

%

Other (n � 607,819)

%

Justice-involved (n � 54,525)

%

Other (n � 1,213,766)

%

Gender Female 6 8 6 11 Male 94 92 94 89

Age �35 26 10 21 14 35–44 16 9 13 11 45–54 22 14 27 16 55–64 26 28 31 27 65� 9 39 8 33

Race/ethnicity Hispanic 4 3 8 9 Non-Hispanic

American Indian/Alaskan Native 3 2 1 1 Asian 1 1 1 2 Black/African American 15 10 37 25 White 77 84 52 63

Marital status Single 31 15 39 23 Married 27 56 19 46 Divorced/separated 40 25 40 27 Widowed 2 4 3 4

Homeless No 92 99 87 98 Yes 8 �1 13 2

Service in Iraq or Afghanistan No 71 85 79 82 Yes 29 15 21 18

Service-connected disability rating None 35 33 44 34 0–49% 28 24 28 26 50–100% 37 42 29 40

Deyo Comorbidity Index 0 conditions 64 48 63 52 1 condition 19 22 19 21 2 conditions 7 11 7 10 3� conditions 10 18 11 17

52 FINLAY ET AL.

Rural Residence and Justice Involvement

Among veterans with a mental health disorder, the first logistic regression model examined the interaction of residence and justice status associ- ated with entry into mental health treatment. The interaction of residence by justice status was sig- nificant (� � 0.09, standard error [SE] � 0.04, p � .001), as were the main effects for residence (� � �0.08, SE � 0.01, p � .001) and justice status (� � �0.84, SE � 0.02, p � .001), adjust- ing for other covariates. Justice-involved veterans, whether rural or urban residents, had higher odds of entering mental health treatment than other veterans (see Figure 1). Other veterans who lived in rural areas had the lowest odds of entering treatment. The interaction effect indicated that jus- tice-involved veterans had the same odds of en- tering mental health treatment regardless of resi- dence, whereas among other veterans, living in a rural area was associated with lower mental health treatment entry.

The second logistic regression model exam- ined residence and justice-involved status among veterans diagnosed with a substance use disorder. The interaction term of residence by justice status was significant (� � 0.28, SE �

0.03, p � .001), and there were significant main effects for residence (� � �0.33, SE � 0.01, p � .001) and justice status (� � 1.07, SE � 0.01, p � .001), adjusting for other covariates. Justice-involved veterans, whether they lived in urban or rural areas, had higher odds of entering substance use disorder treatment than other vet- erans (see Figure 2). Other veterans who lived in rural areas had the lowest odds of entering treatment. The interaction term indicated that justice-involved veterans who lived in rural ar- eas had lower odds of entering substance use disorder treatment compared with justice- involved veterans who lived in urban areas. Among other veterans a similar pattern was observed: Veterans in rural areas had lower odds of entering substance use disorder treat- ment. Because of these differences by geo- graphic location, we analyzed veterans in rural areas separately in the remaining models (re- sults for veterans in urban areas are reported in the online supplemental Appendix).

Treatment Entry Among Rural Veterans

Among rural veterans with a mental health disorder, justice-involved veterans had 2.59

Table 2 Mental Health and Substance Use Disorder Diagnoses Among Veterans by Residence and Justice Status in Fiscal Year 2014

Diagnoses

Rural Urban

Justice-involved (n � 15,328)

%

Other (n � 607,819)

%

Justice-involved (n � 54,525)

%

Other (n � 1,213,766)

%

Mental health diagnoses Depression 66 62 63 61 Posttraumatic stress disorder 51 42 42 39 Anxiety 35 31 29 31 Bipolar disorder 16 6 16 8 Schizophrenia 7 3 10 5 Other mental health disorders 6 3 8 3 Personality disorders 12 3 11 3

Substance use disorder diagnoses Alcohol use disorder 49 17 52 19 Opioid use disorder 13 2 13 3 Cocaine use disorder 13 1 23 4 Amphetamine use disorder 7 1 7 1 Cannabis use disorder 19 4 20 5 Sedative use disorder 3 1 3 1 Other drug use disorders 29 4 33 6 Co-occurring mental health and substance use

disorder diagnoses 56 16 58 19

53RURAL AND JUSTICE-INVOLVED VETERANS

times higher odds (95% confidence interval [CI] [2.42, 2.77]) of entering mental health treatment compared with other veterans (see Table 3). Other patient factors associated with entry into mental health treatment included being female, being from an ethnic or racial minority group, using homeless services or being at risk for homelessness, serving in Iraq or Afghanistan, having a service-connected disability rating, or having comorbid medical conditions.

Among rural veterans diagnosed with a sub- stance use disorder, justice-involved veterans had 3.61 times higher odds (95% CI [3.45, 3.77]) of entering substance use disorder treat-

ment than other veterans. Other patient factors associated with entry into substance use disor- der treatment included being female, being un- der Age 35, being Black/African American, us- ing homeless services or being at risk for homelessness, having a service-connected dis- ability rating, or having comorbid medical con- ditions.

Discussion

The current study examined the associations of rurality of residence and justice status with treatment access among veterans at VHA with

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

1

Justice-involved veterans Other veterans

seitilibaborp detciderP

Rural Urban

Figure 1. Interaction effect of justice status and residence and predicted probabilities of entry into mental health treatment among justice-involved veterans and other veterans who have mental health conditions.

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

1

Justice-involved veterans Other veterans

seitilibaborp detciderP

Rural Urban

Figure 2. Interaction effect of justice status and residence and predicted probabilities of entry into substance use disorder treatment among justice-involved veterans and other veterans who have substance use disorder conditions.

54 FINLAY ET AL.

mental health or substance use disorder diagno- ses. Although main and interaction effects were observed, results largely contradicted hypotheses. Justice-involved veterans had higher odds of en- tering mental health and substance use disorder treatment than other veterans. The only difference by residence observed for justice-involved veter- ans was that urban veterans had higher odds of entering substance use disorder treatment com- pared with rural veterans. For other veterans, ur- ban veterans had higher odds of entering mental health and substance use disorder treatment com- pared with rural veterans, consistent with prior literature (Mohamed et al., 2009; Mott et al., 2015; Teich et al., 2017).

Justice-involved veterans had higher odds of entering mental health and substance use disorder treatment than other veterans, which may be ex- plained by criminal justice factors, such as court mandates, or dedicated support services provided to justice-involved veterans through VHA out- reach. Veterans who were justice-involved may have had court mandates or legal supervision through probation or parole, which have been previously linked with higher odds of treatment

entry (Oser et al., 2011; Vaughn et al., 2012). However, even among adults with alcohol use disorder, parole was not associated with treatment entry (Booth et al., 2013), suggesting that a clear treatment need and legal supervision may not be sufficient to compel treatment attendance. Barriers to treatment, such as bureaucratic processes inter- rupted by incarceration, stigma, and lack of trust for health care providers (Brooks et al., 2016; Goins et al., 2005; Howerton et al., 2007; Marlow et al., 2010; Wainwright et al., 2017), may have been addressed by outreach workers from the VJO and HCRV programs who help to manage paper- work and other administrative processes and can provide an initial positive experience entering VHA treatment. For example, motivational inter- viewing, one technique used by outreach staff, has been shown to improve treatment entry for veter- ans exiting jail (Davis, Baer, Saxon, & Kivlahan, 2003).

Mental Health Treatment

Entry into mental health treatment was rela- tively high for all veterans in the study, with an

Table 3 Patient Characteristics Associated With Mental Health or Substance Use Disorder Treatment Entry Among Rural Veterans in Fiscal Year 2014

Characteristics

Mental health treatment entry

Substance use disorder treatment entry

OR 95% CI OR 95% CI

Justice-involved (reference: no) 2.59��� [2.42, 2.77] 3.61��� [3.45, 3.77] Gender (reference: male) 1.19��� [1.15, 1.22] 1.24��� [1.17, 1.31] Age, years (reference: �35)

35–44 .95�� [.92, .99] .87��� [.82, .92] 45–54 .88��� [.85, .91] .68��� [.65, .72] 55–64 .70��� [.67, .73] .40��� [.38, .43] 65� .41��� [.40, .42] .21��� [.20, .22]

Race/ethnicity (reference: non-Hispanic White) American Indian/Alaskan Native 1.06 [1.00, 1.12] 1.07 [.96, 1.18] Asian 1.15��� [1.08, 1.24] .92 [.80, 1.06] Black/African American 1.20��� [1.17, 1.23] 1.40��� [1.35, 1.47] Hispanic 1.05� [1.01, 1.10] 1.06 [.98, 1.14]

Not married (reference: married) .97��� [.96, .99] .88��� [.85, .90] Homeless (reference: no) 2.80��� [2.47, 3.17] 2.64��� [2.45, 2.86] Service in Iraq/Afghanistan (reference: no) 1.07��� [1.04, 1.10] .97 [.92, 1.01] Service-connected disability rating (reference: no)

�50% 1.12��� [1.10, 1.14] 1.04� [1.00, 1.08] �50% 1.83��� [1.80, 1.86] 1.09��� [1.06, 1.13]

Comorbid medical conditions 1.10��� [1.09, 1.10] 1.06��� [1.04, 1.07]

Note. OR � odds ratio; CI � confidence interval. � p � .05. �� p � .01. ��� p � .001.

55RURAL AND JUSTICE-INVOLVED VETERANS

80% or higher rate of treatment entry observed for all groups. The VHA has a variety of pro- grams for special populations to support their use of VHA, including the Office of Rural Health (Department of Veterans Affairs, 2017), the VJO and HCRV programs (Blue-Howells et al., 2013), and Women’s Health programs (De- partment of Veterans Affairs, 2010). Although veterans in rural areas who were not justice- involved had the lowest rate of mental health treatment entry, the majority entered treatment. Focusing on other aspects of their treatment beyond access, such as resolution of symptoms or receipt of evidence-based practices, may yield results that provide more guidance on how to improve the mental health of veterans at VHA.

Substance Use Disorder Treatment

Compared with mental health treatment, en- try into substance use disorder treatment is low. Just over half of justice-involved veterans in the current study entered substance use disorder treatment. Among justice-involved populations, receipt of pharmacotherapy or psychosocial treatment for substance use disorders has been linked with reduced drug and alcohol use, re- tention in treatment, and reduced criminal ac- tivity and recidivism (Amato et al., 2005; Dolan et al., 2005; Gryczynski et al., 2012; Jonas et al., 2014; Mattick, Breen, Kimber, & Davoli, 2009; Rawson et al., 2004). Improved entry to sub- stance use disorder treatment for justice- involved veterans will likely positively impact their health as well as have a societal impact of crime reduction.

For other veterans, only 20% in rural areas and 28% in urban areas entered substance use disorder treatment, similar to studies of rural residents from the general population (Carlson et al., 2010; Curran et al., 2011). Treatment options are less available in rural areas (Sig- mon, 2014), such as a dearth of providers with buprenorphine waivers (Quest et al., 2012; Rosenblatt et al., 2015), but clearly, substance use disorder treatment options need to be im- proved across the VHA system. Telehealth, which patients indicate is an acceptable treat- ment modality (Benavides-Vaello, Strode, & Sheeran, 2013), may be one option to increase availability of substance use disorder treatment in VHA facilities and communities in which

options are limited. Transportation difficulties and lack of privacy that have been mentioned as barriers to treatment (Brooks et al., 2016; Browne et al., 2016; Buzza et al., 2011; Goins et al., 2005) also may be addressed by telehealth. Substance use disorder treatment has not been successfully integrated across VHA primary care or mental health services (Tracy, Trafton, Weingardt, Aton, & Humphreys, 2007), which may partially explain lower rates of addiction treatment entry in the current study. Rural res- idents have cited a lack of integrated treatment services that address their array of needs as a barrier to treatment (Browne et al., 2016). Vet- erans in the current study may be prioritizing medical or mental health care over addiction treatment. Finally, among people who use co- caine, rural residents had a significantly lower perceived need for treatment than rural residents (Borders et al., 2015). Perhaps many of these veterans in this study, particularly in rural areas, did not perceive a need for substance use dis- order treatment.

Limitations

There are limitations to our study. First, we were unable to assess treatment that veterans were receiving at VA Centers or in the commu- nity. Among others, veterans who lived espe- cially far from VHA facilities may have sought community treatment options. Future data link- age to examine non-VHA treatment use in con- junction with VHA health care use will help inform these results. Second, our methods of identifying residence status or justice status are limited. For example, last known residence may not be accurate for homeless veterans, and there may have been veterans with criminal justice involvement who did not have contact with the VJO or HCRV programs and were coded as not justice-involved. Finally, justice-involved vet- erans may have entered treatment because of criminal justice factors such as court mandates or probation requirements to attend mental health or substance use disorder treatment; we did not have specific information on mandated treatment requirements or veterans who self- or attorney-referred to treatment to help improve criminal justice outcomes (e.g., court sentenc- ing). We could identify veterans who partici- pated in treatment courts (e.g., Veterans Treat- ment Court, drug court) and likely had court

56 FINLAY ET AL.

mandates to attend mental health or substance use disorder treatment as part of their court experience, but they represented only 5% of justice-involved veterans, and there was no impact on the pattern of results when they were removed from the sample. Understand- ing the ways in which the justice system may shape veterans’ engagement with VHA health care will be important for future research to examine more thoroughly in order to fully understand why justice-involved veterans were more likely to access care than veterans without justice involvement.

Conclusions

The majority of justice-involved and other veterans accessed mental health treatment, re- gardless of residence. Entry into substance use disorder treatment, however, was low, espe- cially for rural veterans without justice involve- ment. Lessons may be learned from mental health settings about how to improve access to substance use disorder care for veterans. Prom- ising strategies to improve addiction treatment access include telehealth, which may address a lack of treatment options or transportation dif- ficulties, outreach services to help veterans nav- igate the VHA bureaucracy, and integration of addiction treatment into primary care and men- tal health care settings. Interagency coordina- tion among the VA, community treatment pro- viders, and other entities such as criminal justice partners may be needed to overcome the general lack of treatment resources in rural ar- eas. Importantly, motivating veterans to initiate and actively engage in treatments (U.S. Depart- ment of Health and Human Services, 2016), along with improving access to care, is neces- sary to help people with mental health and sub- stance-related conditions achieve recovery.

References

Amato, L., Davoli, M., Perucci, C. A., Ferri, M., Faggiano, F., & Mattick, R. P. (2005). An over- view of systematic reviews of the effectiveness of opiate maintenance therapies: Available evidence to inform clinical practice and research. Journal of Substance Abuse Treatment, 28, 321–329. http:// dx.doi.org/10.1016/j.jsat.2005.02.007

Andrews, C. M., Shin, H. C., Marsh, J. C., & Cao, D. (2013). Client and program characteristics associ-

ated with wait time to substance abuse treatment entry. The American Journal of Drug and Alcohol Abuse, 39, 61– 68. http://dx.doi.org/10.3109/ 00952990.2012.694515

Benavides-Vaello, S., Strode, A., & Sheeran, B. C. (2013). Using technology in the delivery of mental health and substance abuse treatment in rural com- munities: A review. The Journal of Behavioral Health Services & Research, 40, 111–120. http:// dx.doi.org/10.1007/s11414-012-9299-6

Blue-Howells, J. H., Clark, S. C., van den Berk- Clark, C., & McGuire, J. F. (2013). The U.S. Department of Veterans Affairs Veterans Justice programs and the sequential intercept model: Case examples in national dissemination of intervention for justice-involved veterans. Psychological Ser- vices, 10, 48 –53. http://dx.doi.org/10.1037/ a0029652

Booth, B. M., Curran, G. M., Han, X., & Edlund, M. J. (2013). Criminal justice and alcohol treat- ment: Results from a national sample. Journal of Substance Abuse Treatment, 44, 249 –255. http:// dx.doi.org/10.1016/j.jsat.2012.07.008

Borders, T. F., Booth, B. M., Stewart, K. E., Cheney, A. M., & Curran, G. M. (2015). Rural/urban resi- dence, access, and perceived need for treatment among African American cocaine users. The Jour- nal of Rural Health, 31, 98 –107. http://dx.doi.org/ 10.1111/jrh.12092

Bronson, J., Carson, A., Noonan, M., & Berzofsky, M. (2015). Veterans in prison and jail, 2011–12 (Report No. NCJ249144). Retrieved from https:// www.bjs.gov/content/pub/pdf/vpj1112.pdf

Brooks, E., Dailey, N. K., Bair, B. D., & Shore, J. H. (2016). Listening to the patient: Women veterans’ insights about health care needs, access, and qual- ity in rural areas. Military Medicine, 181, 976 – 981. http://dx.doi.org/10.7205/MILMED-D-15- 00367

Browne, T., Priester, M. A., Clone, S., Iachini, A., DeHart, D., & Hock, R. (2016). Barriers and fa- cilitators to substance use treatment in the rural south: A qualitative study. The Journal of Rural Health, 32, 92–101. http://dx.doi.org/10.1111/jrh .12129

Buzza, C., Ono, S. S., Turvey, C., Wittrock, S., Noble, M., Reddy, G., . . . Reisinger, H. S. (2011). Distance is relative: Unpacking a principal barrier in rural healthcare. Journal of General Internal Medicine, 26(Suppl. 2), 648 – 654. http://dx.doi .org/10.1007/s11606-011-1762-1

Carlson, R. G., Sexton, R., Wang, J., Falck, R., Leukefeld, C. G., & Booth, B. M. (2010). Predic- tors of substance abuse treatment entry among rural illicit stimulant users in Ohio, AR, and Ken- tucky. Substance Abuse, 31, 1–7. http://dx.doi.org/ 10.1080/08897070903442459

57RURAL AND JUSTICE-INVOLVED VETERANS

Curran, G. M., Ounpraseuth, S. T., Allee, E., Small, J., & Booth, B. M. (2011). Trajectories in use of substance abuse and mental health services among stimulant users in rural areas. Psychiatric Services, 62, 1230 –1232. http://dx.doi.org/10.1176/ps.62.10 .pss6210_1230

Davis, T. M., Baer, J. S., Saxon, A. J., & Kivlahan, D. R. (2003). Brief motivational feedback im- proves post-incarceration treatment contact among veterans with substance use disorders. Drug and Alcohol Dependence, 69, 197–203. http://dx.doi .org/10.1016/S0376-8716(02)00317-4

Department of Veterans Affairs. (2010). VHA Direc- tive 1330.01(1): Health care services for women veterans. Retrieved from https://www.va.gov/ vhapublications/ViewPublication.asp?pub_ID� 5332

Department of Veterans Affairs. (2017). Office of Rural Health 2015–2019 strategic plan. Retrieved from https://www.ruralhealth.va.gov/

Deyo, R. A., Cherkin, D. C., & Ciol, M. A. (1992). Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. Journal of Clinical Epidemiology, 45, 613– 619. http://dx.doi .org/10.1016/0895-4356(92)90133-8

Dolan, K. A., Shearer, J., White, B., Zhou, J., Kaldor, J., & Wodak, A. D. (2005). Four-year follow-up of imprisoned male heroin users and methadone treat- ment: Mortality, re-incarceration and hepatitis C infection. Addiction, 100, 820 – 828. http://dx.doi .org/10.1111/j.1360-0443.2005.01050.x

Finlay, A. K., Binswanger, I., Timko, C., Rosenthal, J., Clark, S., Blue-Howells, J., . . . Harris, A. H. S. (2016). Receipt of pharmacotherapy for alcohol use disorder by male justice-involved U.S. Veter- ans Health Administration patients. Criminal Jus- tice Policy Review. Advance online publication. http://dx.doi.org/10.1177/0887403416644011

Finlay, A. K., Harris, A. H. S., Rosenthal, J., Blue- Howells, J., Clark, S., McGuire, J., . . . Binswanger, I. (2016). Receipt of pharmacother- apy for opioid use disorder by justice-involved U.S. Veterans Health Administration patients. Drug and Alcohol Dependence, 160, 222–226. http://dx.doi.org/10.1016/j.drugalcdep.2016.01 .013

Finlay, A. K., Smelson, D., Sawh, L., McGuire, J., Rosenthal, J., Blue-Howells, J., . . . Harris, A. H. S. (2016). U.S. Department of Veterans Affairs Vet- erans Justice Outreach program: Connecting jus- tice-involved veterans with mental health and sub- stance use disorder treatment. Criminal Justice Policy Review, 27, 203–222. http://dx.doi.org/10 .1177/0887403414562601

Finlay, A. K., Stimmel, M., Blue-Howells, J., Rosenthal, J., McGuire, J., Binswanger, I., . . . Timko, C. (2017). Use of Veterans Health Admin- istration mental health and substance use disorder

treatment after exiting prison: The Health Care for Reentry Veterans program. Administration and Policy in Mental Health and Mental Health Ser- vices Research, 44, 177–187. http://dx.doi.org/10 .1007/s10488-015-0708-z

Goins, R. T., Williams, K. A., Carter, M. W., Spen- cer, M., & Solovieva, T. (2005). Perceived barriers to health care access among rural older adults: A qualitative study. The Journal of Rural Health, 21, 206 –213. http://dx.doi.org/10.1111/j.1748-0361 .2005.tb00084.x

Gryczynski, J., Kinlock, T. W., Kelly, S. M., O’Grady, K. E., Gordon, M. S., & Schwartz, R. P. (2012). Opioid agonist maintenance for probation- ers: Patient-level predictors of treatment retention, drug use, and crime. Substance Abuse, 33, 30 –39. http://dx.doi.org/10.1080/08897077.2011.616816

Gryczynski, J., Schwartz, R. P., Salkever, D. S., Mitchell, S. G., & Jaffe, J. H. (2011). Patterns in admission delays to outpatient methadone treat- ment in the United States. Journal of Substance Abuse Treatment, 41, 431– 439. http://dx.doi.org/ 10.1016/j.jsat.2011.06.005

Howerton, A., Byng, R., Campbell, J., Hess, D., Owens, C., & Aitken, P. (2007). Understanding help seeking behaviour among male offenders: Qualitative interview study. British Medical Jour- nal, 334, 303. http://dx.doi.org/10.1136/bmj.39059 .594444.AE

Hudson, T. J., Fortney, J. C., Williams, J. S., Austen, M. A., Pope, S. K., & Hayes, C. J. (2014). Effect of rural residence on use of VHA mental health care among OEF/OIF veterans. Psychiatric Ser- vices, 65, 1420 –1425. http://dx.doi.org/10.1176/ appi.ps.201300041

Johnson, C. E., Bush, R. L., Harman, J., Bolin, J., Evans Hudnall, G., & Nguyen, A. M. (2015). Vari- ation in utilization of health care services for rural VA enrollees with mental health-related diagno- ses. The Journal of Rural Health, 31, 244 –253. http://dx.doi.org/10.1111/jrh.12105

Jonas, D. E., Amick, H. R., Feltner, C., Bobashev, G., Thomas, K., Wines, R., . . . Garbutt, J. C. (2014). Pharmacotherapy for adults with alcohol use dis- orders in outpatient settings: A systematic review and meta-analysis. Journal of the American Med- ical Association, 311, 1889 –1900. http://dx.doi .org/10.1001/jama.2014.3628

Kondo, K., Low, A., Everson, T., Gordon, C. D., Veazie, S., Lozier, C. C., . . . Kansagara, D. (2017). Health disparities in veterans: A map of the evidence. Medical Care, 55(Suppl. 9, Suppl. 2), S9 –S15. http://dx.doi.org/10.1097/MLR .0000000000000756

Marlow, E., White, M. C., & Chesla, C. A. (2010). Barriers and facilitators: Parolees’ perceptions of community health care. Journal of Correctional

58 FINLAY ET AL.

Health Care, 16, 17–26. http://dx.doi.org/10.1177/ 1078345809348201

Mattick, R. P., Breen, C., Kimber, J., & Davoli, M. (2009). Methadone maintenance therapy versus no opioid replacement therapy for opioid dependence. Cochrane Database of Systematic Reviews, 3, CD002209.

McGuire, J., Rosenheck, R. A., & Kasprow, W. J. (2003). Health status, service use, and costs among veterans receiving outreach services in jail or com- munity settings. Psychiatric Services, 54, 201– 207. http://dx.doi.org/10.1176/appi.ps.54.2.201

Medicode (Firm). (1996). ICD-9-CM: International classification of diseases, 9th revision, clinical modification. Salt Lake City, Utah: Medicode.

Mohamed, S., Neale, M., & Rosenheck, R. A. (2009). VA intensive mental health case management in urban and rural areas: Veteran characteristics and service delivery. Psychiatric Services, 60, 914 – 921. http://dx.doi.org/10.1176/ps.2009.60.7.914

Mott, J. M., Grubbs, K. M., Sansgiry, S., Fortney, J. C., & Cully, J. A. (2015). Psychotherapy utili- zation among rural and urban veterans from 2007 to 2010. The Journal of Rural Health, 31, 235– 243. http://dx.doi.org/10.1111/jrh.12099

Mowbray, O., McBeath, B., Bank, L., & Newell, S. (2016). Trajectories of health and behavioral health services use among community corrections- involved rural adults. Social Work Research, 40, 7–18. http://dx.doi.org/10.1093/swr/svv048

Oser, C. B., Leukefeld, C. G., Tindall, M. S., Garrity, T. F., Carlson, R. G., Falck, R., . . . Booth, B. M. (2011). Rural drug users: Factors associated with substance abuse treatment utilization. Interna- tional Journal of Offender Therapy and Compar- ative Criminology, 55, 567–586. http://dx.doi.org/ 10.1177/0306624X10366012

Quest, T. L., Merrill, J. O., Roll, J., Saxon, A. J., & Rosenblatt, R. A. (2012). Buprenorphine therapy for opioid addiction in rural Washington: The ex- perience of the early adopters. Journal of Opioid Management, 8, 29 –38. http://dx.doi.org/10.5055/ jom.2012.0093

Ratcliffe, M., Burd, C., Holder, K., & Fields, A. (2016). Defining rural at the U.S. Census Bureau (ACSGEO-1). Retrieved from https://www.census .gov/library/publications/2016/acs/acsgeo-1.html

Rawson, R. A., Marinelli-Casey, P., Anglin, M. D., Dickow, A., Frazier, Y., Gallagher, C., . . . Meth- amphetamine Treatment Project Corporate Au- thors. (2004). A multi-site comparison of psycho- social approaches for the treatment of methamphetamine dependence. Addiction, 99, 708 –717. http://dx.doi.org/10.1111/j.1360-0443 .2004.00707.x

Rosenblatt, R. A., Andrilla, C. H., Catlin, M., & Larson, E. H. (2015). Geographic and specialty distribution of U.S. physicians trained to treat opi- oid use disorder. Annals of Family Medicine, 13, 23–26. http://dx.doi.org/10.1370/afm.1735

Sigmon, S. C. (2014). Access to treatment for opioid dependence in rural America: Challenges and fu- ture directions. Journal of the American Medical Association Psychiatry, 71, 359 –360. http://dx.doi .org/10.1001/jamapsychiatry.2013.4450

Teich, J., Ali, M. M., Lynch, S., & Mutter, R. (2017). Utilization of mental health services by veterans living in rural areas. The Journal of Rural Health, 33, 297–304. http://dx.doi.org/10.1111/jrh.12221

Tracy, S. W., Trafton, J. A., Weingardt, K. R., Aton, E. G., & Humphreys, K. (2007). How are sub- stance use disorders addressed in VA psychiatric and primary care settings? Results of a national survey. Psychiatric Services, 58, 266 –269. http:// dx.doi.org/10.1176/ps.2007.58.2.266

U.S. Department of Health and Human Services. (2016). Facing addiction in America: The Surgeon General’s report on alcohol, drugs, and health. Washington, DC: Author.

Vaughn, M. G., DeLisi, M., Beaver, K. M., Perron, B. E., & Abdon, A. (2012). Toward a criminal justice epidemiology: Behavioral and physical health of probationers and parolees in the United States. Journal of Criminal Justice, 40, 165–173. http://dx.doi.org/10.1016/j.jcrimjus.2012.03.001

Wainwright, V., McDonnell, S., Lennox, C., Shaw, J., & Senior, J. (2017). Treatment barriers and support for male ex-Armed Forces personnel in prison: Professional and service user perspectives. Qualitative Health Research, 27, 759 –769. http:// dx.doi.org/10.1177/1049732316636846

Received June 13, 2017 Revision received March 1, 2018

Accepted March 3, 2018 �

59RURAL AND JUSTICE-INVOLVED VETERANS