Emily R. Edwards, PhD

Personality disorders (PDs) are enduring patterns of internal experience and behavior that differ from cultural norms and expectations, are inflexible and pervasive, have their onset in adolescence or early adulthood, and lead to distress or impairment. Ten PDs are included in the Diagnostic and Statistical Manual of Mental Disorders (Fifth Edition): paranoid, schizoid, schizotypal, borderline, antisocial, histrionic, narcissistic, avoidant, dependent, and obsessive-compulsive.¹ These disorders impose a high burden on patients, families, health care systems, and broader economic systems.^2,3 Up to 1 in 7 persons in the community and 50% of those receiving outpatient mental health treatment experience a PD.^4,5 These conditions are associated with an increased risk of adverse events, including suicide attempt and death by suicide, criminal-legal involvement, homelessness, substance use, underemployment, relational issues, and high utilization of psychiatric services.^6-9 PDs are routinely underassessed, underdocumented, and undertreated in clinical settings, and consistently receive less research funding than other, less prevalent forms of psychopathology. ^10-12 As a result, there is limited understanding of clinical needs of individuals experiencing PDs.

MILITARY VETERANS WITH PERSONALITY DISORDERS

Underacknowledgment of PDs and their associated difficulties may be especially pronounced in veteran populations. Due to longstanding etiological theories that implicate childhood trauma and adolescent onset in pathology development, PDs are traditionally considered pre-existing conditions or developmental abnormalities by the US Department of Defense and US Department of Veterans Affairs (VA). As a result, PDs are therefore deemed incompatible with military service and ineligible for service-connected disability benefits.^13-15 Such determinations allowed PD pathology to be used as grounds for discharge for 26,000 service members from 2001 to 2007, or 2.6% of total enlisted discharges during that period.^13,15,16

Despite this structural discrimination, recent research suggests veterans may be more likely to experience PD pathology than the general population.17 For example, a 2021 epidemiological survey in a community-based veteran sample found elevated rates of borderline, antisocial, and schizotypal PDs (6%-13%).⁶ In contrast, only 0.8% to 5.0% of veteran electronic health records (EHRs) have a documented PD diagnosis.^8,18,19 Such elevations in PD pathology within veteran samples imply either a disproportionately high prevalence among enlistees (and therefore missed during recruitment procedures) or onset following military service, possibly due to exposure to traumatic events and/ or occupational stress.¹⁷ Due to the relative infancy of research in this area and a lack of longitudinal studies, etiology and course of illness for personality pathology in veterans remains largely unclear.

Structural underacknowledgment of PDs among military personnel has contributed to their underrepresentation in research on veteran populations. PD-focused research with veterans is rare, despite a rapid increase in broader empirical attention paid to these conditions in nonveteran samples.²⁰ A recent meta-analysis of veterans with PDs identified 27 studies that included basic prevalence statistics. PDs were rarely a primary focus for these studies, and most were limited to veterans seen in Veterans Health Administration (VHA) settings.¹⁷ The literature also paints a bleak picture, suggesting veterans who experience PDs are at higher risk for suicide attempt and death by suicide, criminal-legal involvement, and homelessness. They also tend to experience more severe comorbid psychopathological symptoms and more often use high-intensity mental health services (eg, care within emergency departments or psychiatric inpatient settings) than veterans without PD pathology.^6,8,18,19,21 However, PD pathology does not appear to impede the effectiveness of treatment for veterans.^22-24 The implications of PD pathology on broader psychosocial functioning and health care needs certify a need for additional research that examines patterns of personality pathology, particularly in veterans outside the VHA.

METHODS

This study aims to enhance understanding of veterans affected by PDs and offer insight and guidance for treatment of these conditions in federal and nonfederal treatment settings. Previous research has been largely limited to VHA care-receiving samples; the longstanding stigma against PDs by the US military and VA may contribute to biased diagnosis and documentation of PDs in these settings. A large sample of veterans receiving community-based mental health care was therefore used to explore aims of the current study. This study specifically examined demographic patterns, diagnostic comorbidity, psychosocial outcomes, and treatment care settings among veterans with and without a PD diagnosis. Consistent with previous research, we hypothesized that veterans with a PD diagnosis would have more severe mental health comorbidities, poorer psychosocial outcomes, and receive care in higher intensity settings relative to veterans without a diagnosis.

Data for the sample were drawn from the Mental Health Client-Level Data, a publicly available national dataset of nearly 7 million patients who received mental health treatment services provided or funded through state mental health agencies in 2022.²⁵ The analytic sample included about 2.5 million patients for whom veteran status and data around the presence or absence of a PD diagnosis were available. Of these patients, 104,198 were identified as veterans. Veteran patients were identified as predominantly male (63%), White (71%), non-Hispanic (90%), and never married (54%).

Measures

The parent dataset included demographic, clinical, and psychosocial outcome information reported by treatment facilities to individual state administrative systems for each patient who received services. To protect patient privacy, only nonprotected health information is included, and efforts were made throughout compilation of the parent dataset to ensure patient privacy (eg, limiting detail of information disseminated for public access). Because the parent dataset does not include protected health information, studies using these data are considered exempt from institutional review board oversight.

Demographic information. This study reviewed veteran status, sex, race, ethnicity, age, education, and marital status. Veteran status was defined by whether the patient was aged ≥ 18 years and had previously served (but was not currently serving) in the military. Patients with a history of service in the National Guard or Military Reserves were only classified as veterans if they had been called or ordered to active duty while serving. Sex was operationalized dichotomously as male or female; no patients were identified as intersex, transgender, or other gender identities.

Clinical information. Up to 3 mental health diagnoses were reported for each patient and included the following disorders: personality, trauma and attention-deficit/hyperactivity, stressor, anxiety, conduct, delirium/dementia, bipolar, depressive, oppositional defiant, pervasive developmental, schizophrenia or other psychotic, and alcohol or substance use. Mental health diagnosis categories were generated for the parent dataset by grouping diagnostic codes corresponding to each category. To protect patient privacy, more detailed diagnostic information was not available as part of the parent dataset. Although the American Psychiatric Association recognizes 10 distinct PDs, the exact nature of PD diagnoses was not included within the parent dataset. PD diagnoses were coded to reflect the presence or absence of any such diagnosis.

A substance use problem designation was also provided for patients according to various identification methods, including substance use disorder (SUD) diagnosis, substance use screening results, enrollment in a substance use program, substance use survey, service claims information, and other related sources of information. A severe mental illness or serious emotional disturbance designation was provided for patients meeting state definitions of these designations. Context(s) of service provision were coded as inpatient state psychiatric hospital, community-based program, residential treatment center, judicial institution, or other psychiatric inpatient setting.

Psychosocial outcome information. Patient employment and residential status were also included in analyses. Each reflected status at the time of discharge from services or end of reporting period; employment status was only provided for patients receiving treatment in community-based programs.

Data Analysis

Descriptive statistics and X² analyses were used to compare demographic, clinical, and psychosocial outcome variables between patients with and without PD diagnoses. These analyses were calculated for both the 104,198 veterans and the 2,222,306 nonveterans aged ≥ 18 years in the dataset. Given the sample size, a conservative α of .01 was used to determine statistical significance.

RESULTS

In this sample of persons receiving state-funded mental health care, veterans were significantly less likely than nonveterans to have a documented PD diagnosis (2.1% vs 3.6%, X² [1] = 647.49; P < .01). PD diagnoses were more common among White (risk ratio [RR], 1.11), non-Hispanic (RR, 1.03) veterans who were in middle to late adulthood (RR, 1.16-1.40), more educated (RR, 1.35), and divorced or widowed (RR, 1.43), and less common among Black/African American (RR, 0.78) or Puerto Rican (RR, 0.32) veterans who were in early adulthood (RR, 0.31-0.79), less educated (RR, 0.64-0.89), and currently married (RR, 0.89) or never married (RR, 0.86). Veteran men and women were equally likely to have a PD diagnosis (RR, 1.03) (Table 1). Among nonveterans, men were less likely than women to have a PD diagnosis (RR, 0.79), and PD diagnoses were most common among persons in middle adulthood (RR, 1.06-1.15) (eAppendix 1).

Veterans with a PD diagnosis were more likely than those without a diagnosis to have more diagnoses (RR, 2.96-8.49) and to have comorbid trauma or related stressor (RR, 1.33), or bipolar (RR, 1.56) or psychotic (RR, 1.15) disorder diagnoses, but less likely to have comorbid depressive disorder (RR, 0.82). Although veterans with and without a PD diagnosis were similarly likely to have a comorbid SUD (RR, 1.13), those with a PD diagnosis were significantly less likely to be assigned a substance use problem designation (RR, 0.78). PD diagnosis was also more common among veterans who received services in state psychiatric hospitals (RR, 3.05), community-based clinics (RR, 1.06), and judicial institutions (RR, 6.33) and less common among those who received services in other psychiatric inpatient settings (RR, 0.30). No differences were observed for residential treatment settings (RR, 0.79). Among nonveterans, a PD diagnosis was associated with slightly greater odds of a substance use designation (RR, 1.03) (eAppendix 2).

Veterans with a PD diagnosis were also less likely to have full-time employment (RR, 0.73) and more likely to have undifferentiated employment (RR, 2.00) or to be removed from the labor force (RR, 1.35). Veterans with a PD diagnosis were also more likely to reside in nontraditional living conditions (RR, 1.42) and less likely to be residing in a private residence (RR, 0.98), compared with those without PD diagnosis. The rates of homelessness were similar for veterans with and without a PD diagnosis (RR, 0.90) (Table 2). These patterns were similar among nonveterans.

DISCUSSION

This study examined the rate and correlates of PD diagnosis among a large, community-based sample of veterans receiving state-funded mental health care. About 2% of veterans in this sample had a PD diagnosis, with diagnoses more common among veterans who were White, non-Hispanic, aged ≥ 45 years, with higher education, divorced or widowed, also diagnosed with trauma-related, bipolar, and/or psychotic disorders, underemployed, nontraditionally housed, and receiving treatment in state psychiatric hospital, community-based clinic, or judicial system settings.

The observed rate of PD diagnosis in this study aligns with what is typically observed in VHA EHRs.^8,18,19 However, the rate is notably lower than prevalence estimates for psychiatric outpatient settings (about 50%) and in meta-analyses of prevalence among veterans (0.8%-23% for each of the 10 PDs).^4,17,26 Longstanding stigma against PDs may contribute to underdiagnosis. For example, many clinicians are concerned that documentation or disclosure of a PD will interfere with the patient’s ability to access treatment due to stigma and discrimination.^27,28 These fears are not unfounded; even among clinicians, PDs are commonly considered untreatable, and many individuals with PDs are denied access to evidence-based treatments due to the diagnosis.²⁹ In a 2016 survey of community psychiatrists, nearly 1 in 4 reported that they avoid taking patients with a borderline PD diagnosis in their caseloads.²⁸ To date, no studies have been conducted to explore clinicians’ willingness to accept patients with other PDs or, specifically, among veterans.

Despite such widespread stigma, research suggests clinicians' negative attitudes toward PDs can be decreased through antistigma campaigns.³⁰ However, it remains unclear if such efforts also contribute to an increase in clinicians’ willingness to document PD diagnoses. Without accurate identification and documentation, the field’s understanding of PDs will remain limited.

In the current study, veterans with PD diagnoses tended to present with more complex and severe psychiatric comorbidities compared to veterans without such diagnoses. Observed comorbidity of PDs (particularly borderline PD) with trauma-related and bipolar disorders is well established.8 Conversely, co-occurring personality and psychotic disorders—which comprise 16% of veterans with a PD diagnosis in the sample in this study—are not consistently examined in the literature. A 2022 examination of veterans receiving VHA care suggested 12% and 13% of those with a PD diagnosis documented in their EHR also had documented schizophrenia or another psychotic disorder, respectively. PD diagnoses were associated with 6.88- and 9.80-fold increases in risk for comorbid schizophrenia and other psychotic disorder diagnoses, respectively.⁸ Similarly, a recent longitudinal study of nearly 2 million Swedish individuals suggested borderline PD is specifically associated with a > 24-times greater risk of having a comorbid psychotic disorder.³¹ It is therefore possible that the comorbidity between personality and psychotic disorders is quite common despite its relative lack of attention in empirical research.

Veterans with PD diagnoses in this study were also more likely to experience substandard housing, employment challenges, and receive treatment through judicial institutions than those without a PD diagnosis. Such findings are consistent with previous research demonstrating the substantial psychosocial challenges associated with PD diagnosis, even after controlling for comorbid conditions.^7,9 Veterans with PDs may benefit from specialized case management and support to facilitate stable housing and employment and to mitigate the risk of judicial involvement. Some research suggests veterans with PDs may be less likely to gain competitive employment after participating in VA therapeutic and supportive employment services programs, suggesting standard programming may be less suitable for this population.³² Similarly, other research suggests individuals with PDs may benefit more from specialized, intensive services than standard clinical case management.³³ Future research may therefore benefit from clarifying the degree to which adaptations to standard programming could yield beneficial effects for persons with PD diagnoses.

Implications

Cumulatively, the results of this study attest to the necessity for transdiagnostic treatment planning that includes close collaboration between psychotherapeutic, pharmacological, and case management services. Some psychotherapy models for PDs, such as dialectical behavior therapy (DBT), which includes a combination of group skills training, individual therapy, as-needed phone coaching, and therapist consultation, may be successfully adapted to include this collaboration.^34-36 However, implementation of such comprehensive programming often requires extensive clinician training and coordination of resources, which poses implementation challenges.^37-39 In 2021, the VHA began large-scale implementation of PD-specific psychotherapy for veterans with recent suicidal self-directed violence and borderline PD, including DBT, though to date results remain unclear.⁴⁰ Generalist approaches, such as good psychiatric management (GPM), which emphasizes emotional validation, practical problem solving, realistic goal setting, and relationship functioning within the context of standard care appointments, may be more easily implemented in community care settings due to lesser training and resource requirements and can also be adapted to include needed elements of care coordination.^41,42 Both DBT and GPM were initially developed for the treatment of borderline PD. Although DBT has also demonstrated some effectiveness in the treatment of antisocial PD, potential applications of DBT and GPM to other PDs remain largely underdeveloped.^43-46

There are no widely accepted medications for the treatment of PDs. Pharmacotherapy for these conditions typically consists of individualized approaches informed by personal experience that attempt to balance targeting of specific symptoms while minimizing polypharmacy and potential risks (eg, overdose or addiction).^47,48 Despite this, pharmacotherapy is often considered a necessary component in the treatment of bipolar and psychotic disorders, both common comorbidities of PDs found in veterans in this study.^49,50 Careful consideration of complex comorbidities and pharmacotherapy needs is warranted in the treatment of veterans with PDs. Future research may benefit from clarifying clinical guidelines around pharmacotherapy, particularly for observed comorbidities of PDs to trauma, bipolar, and psychotic disorders.

It is important to note the discrepancies in the results of this study surrounding patient substance use. The results suggest a negligible or inverse association between the likelihood of a PD diagnosis and difficulties with substance use among the veterans in this study. However, the unexpectedly low rate of SUD diagnoses (< 6%) suggests that they were likely underdocumented. Research suggests a strong association between personality and SUDs in both veteran and civilian samples.^6,51 Results suggesting a lower prevalence of substance use difficulties among treatment-seeking veterans with PDs should be interpreted with great caution.

Demographically, PD diagnoses were more common among veterans who were White, non-Hispanic, and aged ≥ 45 years, and less common among veterans who were Black/ African American, mixed/unspecified race, Puerto Rican or other non-Mexican Hispanic ethnicity, or aged < 35 years. No significant sex-based differences were observed. These patterns are consistent with research suggesting individuals who identify as Black may be less likely than individuals who identify as White to report PD symptoms, meet criteria for a PD, and have a PD diagnosed even when it is warranted.⁵²

The findings observed in this study with respect to age, however, are notably inconsistent with the literature. Previous research typically suggests a negative association between age and PD pathology; however, a 2020 review of PDs in older adults by Penders et al suggests a prevalence of 11% to 15% in this population.^53,54 Research into PDs most often focuses on adolescent and early adulthood developmental periods, limiting insight into the phenomenology of PDs in middle to late adulthood.⁵⁵ Further, most research into PDs among geriatric populations has focused on psychometric assessment rather than practical treatment guidance.⁵⁴ However, in this study, elevated risk for PD diagnoses was salient throughout middle to late adulthood among veterans; similar, albeit less pronounced patterns were also observed for elevated risk of PD diagnosis in middle adulthood among nonveterans. Such findings suggest clarifying the phenomenology and treatment needs of individuals with PDs in middle to late adulthood may have particularly salient implications for the mental health care of veterans affected by these conditions. As the veteran population advances in age, these needs will present unique challenges if health care systems are unprepared to effectively address them.

Limitations

This study is characterized by several strengths, most notably its use of a large dataset recently collected on a national scale. Few studies outside of the VHA system include samples of > 100,000 treatment-seeking veterans collected on a national scale. Nevertheless, results should be understood within the context of several methodological limitations. However, the dataset was limited to the first 3 diagnoses documented in patients’ EHRs, and many patients had no listed diagnoses. Patients with complex comorbidities may have > 3 diagnoses; for these individuals, data provided an incomplete picture of clinical presentation. This is especially relevant for individuals with PDs, who tend to meet criteria for a range of comorbid conditions.^8,10 The now dated practice of listing PDs on Axis II also increases the chance of clinicians listing PDs after conditions traditionally listed on Axis I (eg, major depressive disorder) in patient charts.⁵⁶ This study’s inclusion of only the first 3 listed diagnoses likely underestimated true PD diagnosis prevalence.

The results of this study must be interpreted as reflecting the prevalence and correlates of receiving a PD diagnosis rather than meeting diagnostic criteria for a PD. Relatedly, PD diagnoses were reported as a single construct, limiting insight into prevalence and correlates of individual PD diagnoses (eg, borderline vs paranoid PDs). Meta-analyses estimates suggest PD prevalence among veterans is likely much higher than observed in this study.¹⁷ Stigma continues to discourage clinicians from documenting and disclosing PD diagnoses even when warranted.^27,28 Continued research should aim to clarify conditions (eg, patient presentation, stigma, or institutional culture) contributing to documentation of PD diagnoses. Given the cross-sectional nature of this study, results cannot speak to longitudinal treatment outcomes or prognosis of persons receiving a PD diagnosis.

Despite its large sample size and national representation, the sampling strategy of this study could have contributed to idiosyncrasies in the dataset. Restriction of data to the persons receiving state-funded mental health services introduces a notable bias to the composition of the sample, which is likely comprised of a disproportionately high number of Medicaid recipients, students, and individuals with chronic illnesses and underrepresentation of persons who pay for mental health services using private insurance or private pay arrangements. As such, although socioeconomic information was not provided within this dataset, one can presume a generally lower socioeconomic status among study participants compared to the community at large. This study also included a proportionally small sample of veterans (3.6% compared to about 6.2% in the broader US population), suggesting veterans may have been underrepresented or underidentified in surveyed mental health care settings.⁵⁷ This study also did not include data around service in active-duty military, national guard, or military reserves; a greater proportion of the sample likely had a history of military service than was represented by veteran status designation. Further, the proportionally high sample of individuals with severe mental illness suggests a likely overrepresentation of such conditions in surveyed settings.

Institutional differences in the practice of assigning diagnoses likely limited statistical power to detect potentially meaningful associations and effects. Structural influences, such as stigma and institutional culture, may have notable effects on documentation practices, particularly for PDs. Future research should aim to replicate observed associations using more controlled diagnostic procedures.

Lastly, even with the use of a more conservative α and a focus on effect sizes to guide interpretation of results, use of multiple bivariate analyses can be presumed to have increased the likelihood of type I error. Given the limited prior research in this area, an exploratory approach to statistical analysis was considered warranted to maximize opportunity for identifying areas in need of additional empirical attention. Continued research using more conservative statistical approaches (eg, multivariate analyses) is needed to determine replicability and generalizability of observed results.

CONCLUSIONS

This study examined the prevalence and correlates of PD diagnoses in a national sample of veterans receiving community-based, state-funded mental health care. About 2% received a PD diagnosis, with diagnoses most common among veterans who were White, non-Hispanic, aged ≥ 45 years, also diagnosed with trauma-based, bipolar, and/or psychotic disorders, underemployed, nontraditionally housed, and receiving treatment in a state psychiatric hospital or judicial system setting. The results attest to a necessity for transdiagnostic treatment planning and care coordination for this population, with particular attention to psychosocial stressors.

Personality disorders (PDs) are enduring patterns of internal experience and behavior that differ from cultural norms and expectations, are inflexible and pervasive, have their onset in adolescence or early adulthood, and lead to distress or impairment. Ten PDs are included in the Diagnostic and Statistical Manual of Mental Disorders (Fifth Edition): paranoid, schizoid, schizotypal, borderline, antisocial, histrionic, narcissistic, avoidant, dependent, and obsessive-compulsive.¹ These disorders impose a high burden on patients, families, health care systems, and broader economic systems.^2,3 Up to 1 in 7 persons in the community and 50% of those receiving outpatient mental health treatment experience a PD.^4,5 These conditions are associated with an increased risk of adverse events, including suicide attempt and death by suicide, criminal-legal involvement, homelessness, substance use, underemployment, relational issues, and high utilization of psychiatric services.^6-9 PDs are routinely underassessed, underdocumented, and undertreated in clinical settings, and consistently receive less research funding than other, less prevalent forms of psychopathology. ^10-12 As a result, there is limited understanding of clinical needs of individuals experiencing PDs.

MILITARY VETERANS WITH PERSONALITY DISORDERS

Underacknowledgment of PDs and their associated difficulties may be especially pronounced in veteran populations. Due to longstanding etiological theories that implicate childhood trauma and adolescent onset in pathology development, PDs are traditionally considered pre-existing conditions or developmental abnormalities by the US Department of Defense and US Department of Veterans Affairs (VA). As a result, PDs are therefore deemed incompatible with military service and ineligible for service-connected disability benefits.^13-15 Such determinations allowed PD pathology to be used as grounds for discharge for 26,000 service members from 2001 to 2007, or 2.6% of total enlisted discharges during that period.^13,15,16

Despite this structural discrimination, recent research suggests veterans may be more likely to experience PD pathology than the general population.17 For example, a 2021 epidemiological survey in a community-based veteran sample found elevated rates of borderline, antisocial, and schizotypal PDs (6%-13%).⁶ In contrast, only 0.8% to 5.0% of veteran electronic health records (EHRs) have a documented PD diagnosis.^8,18,19 Such elevations in PD pathology within veteran samples imply either a disproportionately high prevalence among enlistees (and therefore missed during recruitment procedures) or onset following military service, possibly due to exposure to traumatic events and/ or occupational stress.¹⁷ Due to the relative infancy of research in this area and a lack of longitudinal studies, etiology and course of illness for personality pathology in veterans remains largely unclear.

Structural underacknowledgment of PDs among military personnel has contributed to their underrepresentation in research on veteran populations. PD-focused research with veterans is rare, despite a rapid increase in broader empirical attention paid to these conditions in nonveteran samples.²⁰ A recent meta-analysis of veterans with PDs identified 27 studies that included basic prevalence statistics. PDs were rarely a primary focus for these studies, and most were limited to veterans seen in Veterans Health Administration (VHA) settings.¹⁷ The literature also paints a bleak picture, suggesting veterans who experience PDs are at higher risk for suicide attempt and death by suicide, criminal-legal involvement, and homelessness. They also tend to experience more severe comorbid psychopathological symptoms and more often use high-intensity mental health services (eg, care within emergency departments or psychiatric inpatient settings) than veterans without PD pathology.^6,8,18,19,21 However, PD pathology does not appear to impede the effectiveness of treatment for veterans.^22-24 The implications of PD pathology on broader psychosocial functioning and health care needs certify a need for additional research that examines patterns of personality pathology, particularly in veterans outside the VHA.

METHODS

This study aims to enhance understanding of veterans affected by PDs and offer insight and guidance for treatment of these conditions in federal and nonfederal treatment settings. Previous research has been largely limited to VHA care-receiving samples; the longstanding stigma against PDs by the US military and VA may contribute to biased diagnosis and documentation of PDs in these settings. A large sample of veterans receiving community-based mental health care was therefore used to explore aims of the current study. This study specifically examined demographic patterns, diagnostic comorbidity, psychosocial outcomes, and treatment care settings among veterans with and without a PD diagnosis. Consistent with previous research, we hypothesized that veterans with a PD diagnosis would have more severe mental health comorbidities, poorer psychosocial outcomes, and receive care in higher intensity settings relative to veterans without a diagnosis.

Data for the sample were drawn from the Mental Health Client-Level Data, a publicly available national dataset of nearly 7 million patients who received mental health treatment services provided or funded through state mental health agencies in 2022.²⁵ The analytic sample included about 2.5 million patients for whom veteran status and data around the presence or absence of a PD diagnosis were available. Of these patients, 104,198 were identified as veterans. Veteran patients were identified as predominantly male (63%), White (71%), non-Hispanic (90%), and never married (54%).

Measures

The parent dataset included demographic, clinical, and psychosocial outcome information reported by treatment facilities to individual state administrative systems for each patient who received services. To protect patient privacy, only nonprotected health information is included, and efforts were made throughout compilation of the parent dataset to ensure patient privacy (eg, limiting detail of information disseminated for public access). Because the parent dataset does not include protected health information, studies using these data are considered exempt from institutional review board oversight.

Demographic information. This study reviewed veteran status, sex, race, ethnicity, age, education, and marital status. Veteran status was defined by whether the patient was aged ≥ 18 years and had previously served (but was not currently serving) in the military. Patients with a history of service in the National Guard or Military Reserves were only classified as veterans if they had been called or ordered to active duty while serving. Sex was operationalized dichotomously as male or female; no patients were identified as intersex, transgender, or other gender identities.

Clinical information. Up to 3 mental health diagnoses were reported for each patient and included the following disorders: personality, trauma and attention-deficit/hyperactivity, stressor, anxiety, conduct, delirium/dementia, bipolar, depressive, oppositional defiant, pervasive developmental, schizophrenia or other psychotic, and alcohol or substance use. Mental health diagnosis categories were generated for the parent dataset by grouping diagnostic codes corresponding to each category. To protect patient privacy, more detailed diagnostic information was not available as part of the parent dataset. Although the American Psychiatric Association recognizes 10 distinct PDs, the exact nature of PD diagnoses was not included within the parent dataset. PD diagnoses were coded to reflect the presence or absence of any such diagnosis.

A substance use problem designation was also provided for patients according to various identification methods, including substance use disorder (SUD) diagnosis, substance use screening results, enrollment in a substance use program, substance use survey, service claims information, and other related sources of information. A severe mental illness or serious emotional disturbance designation was provided for patients meeting state definitions of these designations. Context(s) of service provision were coded as inpatient state psychiatric hospital, community-based program, residential treatment center, judicial institution, or other psychiatric inpatient setting.

Psychosocial outcome information. Patient employment and residential status were also included in analyses. Each reflected status at the time of discharge from services or end of reporting period; employment status was only provided for patients receiving treatment in community-based programs.

Data Analysis

Descriptive statistics and X² analyses were used to compare demographic, clinical, and psychosocial outcome variables between patients with and without PD diagnoses. These analyses were calculated for both the 104,198 veterans and the 2,222,306 nonveterans aged ≥ 18 years in the dataset. Given the sample size, a conservative α of .01 was used to determine statistical significance.

RESULTS

In this sample of persons receiving state-funded mental health care, veterans were significantly less likely than nonveterans to have a documented PD diagnosis (2.1% vs 3.6%, X² [1] = 647.49; P < .01). PD diagnoses were more common among White (risk ratio [RR], 1.11), non-Hispanic (RR, 1.03) veterans who were in middle to late adulthood (RR, 1.16-1.40), more educated (RR, 1.35), and divorced or widowed (RR, 1.43), and less common among Black/African American (RR, 0.78) or Puerto Rican (RR, 0.32) veterans who were in early adulthood (RR, 0.31-0.79), less educated (RR, 0.64-0.89), and currently married (RR, 0.89) or never married (RR, 0.86). Veteran men and women were equally likely to have a PD diagnosis (RR, 1.03) (Table 1). Among nonveterans, men were less likely than women to have a PD diagnosis (RR, 0.79), and PD diagnoses were most common among persons in middle adulthood (RR, 1.06-1.15) (eAppendix 1).

Veterans with a PD diagnosis were more likely than those without a diagnosis to have more diagnoses (RR, 2.96-8.49) and to have comorbid trauma or related stressor (RR, 1.33), or bipolar (RR, 1.56) or psychotic (RR, 1.15) disorder diagnoses, but less likely to have comorbid depressive disorder (RR, 0.82). Although veterans with and without a PD diagnosis were similarly likely to have a comorbid SUD (RR, 1.13), those with a PD diagnosis were significantly less likely to be assigned a substance use problem designation (RR, 0.78). PD diagnosis was also more common among veterans who received services in state psychiatric hospitals (RR, 3.05), community-based clinics (RR, 1.06), and judicial institutions (RR, 6.33) and less common among those who received services in other psychiatric inpatient settings (RR, 0.30). No differences were observed for residential treatment settings (RR, 0.79). Among nonveterans, a PD diagnosis was associated with slightly greater odds of a substance use designation (RR, 1.03) (eAppendix 2).

Veterans with a PD diagnosis were also less likely to have full-time employment (RR, 0.73) and more likely to have undifferentiated employment (RR, 2.00) or to be removed from the labor force (RR, 1.35). Veterans with a PD diagnosis were also more likely to reside in nontraditional living conditions (RR, 1.42) and less likely to be residing in a private residence (RR, 0.98), compared with those without PD diagnosis. The rates of homelessness were similar for veterans with and without a PD diagnosis (RR, 0.90) (Table 2). These patterns were similar among nonveterans.

DISCUSSION

This study examined the rate and correlates of PD diagnosis among a large, community-based sample of veterans receiving state-funded mental health care. About 2% of veterans in this sample had a PD diagnosis, with diagnoses more common among veterans who were White, non-Hispanic, aged ≥ 45 years, with higher education, divorced or widowed, also diagnosed with trauma-related, bipolar, and/or psychotic disorders, underemployed, nontraditionally housed, and receiving treatment in state psychiatric hospital, community-based clinic, or judicial system settings.

The observed rate of PD diagnosis in this study aligns with what is typically observed in VHA EHRs.^8,18,19 However, the rate is notably lower than prevalence estimates for psychiatric outpatient settings (about 50%) and in meta-analyses of prevalence among veterans (0.8%-23% for each of the 10 PDs).^4,17,26 Longstanding stigma against PDs may contribute to underdiagnosis. For example, many clinicians are concerned that documentation or disclosure of a PD will interfere with the patient’s ability to access treatment due to stigma and discrimination.^27,28 These fears are not unfounded; even among clinicians, PDs are commonly considered untreatable, and many individuals with PDs are denied access to evidence-based treatments due to the diagnosis.²⁹ In a 2016 survey of community psychiatrists, nearly 1 in 4 reported that they avoid taking patients with a borderline PD diagnosis in their caseloads.²⁸ To date, no studies have been conducted to explore clinicians’ willingness to accept patients with other PDs or, specifically, among veterans.

Despite such widespread stigma, research suggests clinicians' negative attitudes toward PDs can be decreased through antistigma campaigns.³⁰ However, it remains unclear if such efforts also contribute to an increase in clinicians’ willingness to document PD diagnoses. Without accurate identification and documentation, the field’s understanding of PDs will remain limited.

In the current study, veterans with PD diagnoses tended to present with more complex and severe psychiatric comorbidities compared to veterans without such diagnoses. Observed comorbidity of PDs (particularly borderline PD) with trauma-related and bipolar disorders is well established.8 Conversely, co-occurring personality and psychotic disorders—which comprise 16% of veterans with a PD diagnosis in the sample in this study—are not consistently examined in the literature. A 2022 examination of veterans receiving VHA care suggested 12% and 13% of those with a PD diagnosis documented in their EHR also had documented schizophrenia or another psychotic disorder, respectively. PD diagnoses were associated with 6.88- and 9.80-fold increases in risk for comorbid schizophrenia and other psychotic disorder diagnoses, respectively.⁸ Similarly, a recent longitudinal study of nearly 2 million Swedish individuals suggested borderline PD is specifically associated with a > 24-times greater risk of having a comorbid psychotic disorder.³¹ It is therefore possible that the comorbidity between personality and psychotic disorders is quite common despite its relative lack of attention in empirical research.

Veterans with PD diagnoses in this study were also more likely to experience substandard housing, employment challenges, and receive treatment through judicial institutions than those without a PD diagnosis. Such findings are consistent with previous research demonstrating the substantial psychosocial challenges associated with PD diagnosis, even after controlling for comorbid conditions.^7,9 Veterans with PDs may benefit from specialized case management and support to facilitate stable housing and employment and to mitigate the risk of judicial involvement. Some research suggests veterans with PDs may be less likely to gain competitive employment after participating in VA therapeutic and supportive employment services programs, suggesting standard programming may be less suitable for this population.³² Similarly, other research suggests individuals with PDs may benefit more from specialized, intensive services than standard clinical case management.³³ Future research may therefore benefit from clarifying the degree to which adaptations to standard programming could yield beneficial effects for persons with PD diagnoses.

Implications

Cumulatively, the results of this study attest to the necessity for transdiagnostic treatment planning that includes close collaboration between psychotherapeutic, pharmacological, and case management services. Some psychotherapy models for PDs, such as dialectical behavior therapy (DBT), which includes a combination of group skills training, individual therapy, as-needed phone coaching, and therapist consultation, may be successfully adapted to include this collaboration.^34-36 However, implementation of such comprehensive programming often requires extensive clinician training and coordination of resources, which poses implementation challenges.^37-39 In 2021, the VHA began large-scale implementation of PD-specific psychotherapy for veterans with recent suicidal self-directed violence and borderline PD, including DBT, though to date results remain unclear.⁴⁰ Generalist approaches, such as good psychiatric management (GPM), which emphasizes emotional validation, practical problem solving, realistic goal setting, and relationship functioning within the context of standard care appointments, may be more easily implemented in community care settings due to lesser training and resource requirements and can also be adapted to include needed elements of care coordination.^41,42 Both DBT and GPM were initially developed for the treatment of borderline PD. Although DBT has also demonstrated some effectiveness in the treatment of antisocial PD, potential applications of DBT and GPM to other PDs remain largely underdeveloped.^43-46

There are no widely accepted medications for the treatment of PDs. Pharmacotherapy for these conditions typically consists of individualized approaches informed by personal experience that attempt to balance targeting of specific symptoms while minimizing polypharmacy and potential risks (eg, overdose or addiction).^47,48 Despite this, pharmacotherapy is often considered a necessary component in the treatment of bipolar and psychotic disorders, both common comorbidities of PDs found in veterans in this study.^49,50 Careful consideration of complex comorbidities and pharmacotherapy needs is warranted in the treatment of veterans with PDs. Future research may benefit from clarifying clinical guidelines around pharmacotherapy, particularly for observed comorbidities of PDs to trauma, bipolar, and psychotic disorders.

It is important to note the discrepancies in the results of this study surrounding patient substance use. The results suggest a negligible or inverse association between the likelihood of a PD diagnosis and difficulties with substance use among the veterans in this study. However, the unexpectedly low rate of SUD diagnoses (< 6%) suggests that they were likely underdocumented. Research suggests a strong association between personality and SUDs in both veteran and civilian samples.^6,51 Results suggesting a lower prevalence of substance use difficulties among treatment-seeking veterans with PDs should be interpreted with great caution.

Demographically, PD diagnoses were more common among veterans who were White, non-Hispanic, and aged ≥ 45 years, and less common among veterans who were Black/ African American, mixed/unspecified race, Puerto Rican or other non-Mexican Hispanic ethnicity, or aged < 35 years. No significant sex-based differences were observed. These patterns are consistent with research suggesting individuals who identify as Black may be less likely than individuals who identify as White to report PD symptoms, meet criteria for a PD, and have a PD diagnosed even when it is warranted.⁵²

The findings observed in this study with respect to age, however, are notably inconsistent with the literature. Previous research typically suggests a negative association between age and PD pathology; however, a 2020 review of PDs in older adults by Penders et al suggests a prevalence of 11% to 15% in this population.^53,54 Research into PDs most often focuses on adolescent and early adulthood developmental periods, limiting insight into the phenomenology of PDs in middle to late adulthood.⁵⁵ Further, most research into PDs among geriatric populations has focused on psychometric assessment rather than practical treatment guidance.⁵⁴ However, in this study, elevated risk for PD diagnoses was salient throughout middle to late adulthood among veterans; similar, albeit less pronounced patterns were also observed for elevated risk of PD diagnosis in middle adulthood among nonveterans. Such findings suggest clarifying the phenomenology and treatment needs of individuals with PDs in middle to late adulthood may have particularly salient implications for the mental health care of veterans affected by these conditions. As the veteran population advances in age, these needs will present unique challenges if health care systems are unprepared to effectively address them.

Limitations

This study is characterized by several strengths, most notably its use of a large dataset recently collected on a national scale. Few studies outside of the VHA system include samples of > 100,000 treatment-seeking veterans collected on a national scale. Nevertheless, results should be understood within the context of several methodological limitations. However, the dataset was limited to the first 3 diagnoses documented in patients’ EHRs, and many patients had no listed diagnoses. Patients with complex comorbidities may have > 3 diagnoses; for these individuals, data provided an incomplete picture of clinical presentation. This is especially relevant for individuals with PDs, who tend to meet criteria for a range of comorbid conditions.^8,10 The now dated practice of listing PDs on Axis II also increases the chance of clinicians listing PDs after conditions traditionally listed on Axis I (eg, major depressive disorder) in patient charts.⁵⁶ This study’s inclusion of only the first 3 listed diagnoses likely underestimated true PD diagnosis prevalence.

The results of this study must be interpreted as reflecting the prevalence and correlates of receiving a PD diagnosis rather than meeting diagnostic criteria for a PD. Relatedly, PD diagnoses were reported as a single construct, limiting insight into prevalence and correlates of individual PD diagnoses (eg, borderline vs paranoid PDs). Meta-analyses estimates suggest PD prevalence among veterans is likely much higher than observed in this study.¹⁷ Stigma continues to discourage clinicians from documenting and disclosing PD diagnoses even when warranted.^27,28 Continued research should aim to clarify conditions (eg, patient presentation, stigma, or institutional culture) contributing to documentation of PD diagnoses. Given the cross-sectional nature of this study, results cannot speak to longitudinal treatment outcomes or prognosis of persons receiving a PD diagnosis.

Despite its large sample size and national representation, the sampling strategy of this study could have contributed to idiosyncrasies in the dataset. Restriction of data to the persons receiving state-funded mental health services introduces a notable bias to the composition of the sample, which is likely comprised of a disproportionately high number of Medicaid recipients, students, and individuals with chronic illnesses and underrepresentation of persons who pay for mental health services using private insurance or private pay arrangements. As such, although socioeconomic information was not provided within this dataset, one can presume a generally lower socioeconomic status among study participants compared to the community at large. This study also included a proportionally small sample of veterans (3.6% compared to about 6.2% in the broader US population), suggesting veterans may have been underrepresented or underidentified in surveyed mental health care settings.⁵⁷ This study also did not include data around service in active-duty military, national guard, or military reserves; a greater proportion of the sample likely had a history of military service than was represented by veteran status designation. Further, the proportionally high sample of individuals with severe mental illness suggests a likely overrepresentation of such conditions in surveyed settings.

Institutional differences in the practice of assigning diagnoses likely limited statistical power to detect potentially meaningful associations and effects. Structural influences, such as stigma and institutional culture, may have notable effects on documentation practices, particularly for PDs. Future research should aim to replicate observed associations using more controlled diagnostic procedures.

Lastly, even with the use of a more conservative α and a focus on effect sizes to guide interpretation of results, use of multiple bivariate analyses can be presumed to have increased the likelihood of type I error. Given the limited prior research in this area, an exploratory approach to statistical analysis was considered warranted to maximize opportunity for identifying areas in need of additional empirical attention. Continued research using more conservative statistical approaches (eg, multivariate analyses) is needed to determine replicability and generalizability of observed results.

CONCLUSIONS

This study examined the prevalence and correlates of PD diagnoses in a national sample of veterans receiving community-based, state-funded mental health care. About 2% received a PD diagnosis, with diagnoses most common among veterans who were White, non-Hispanic, aged ≥ 45 years, also diagnosed with trauma-based, bipolar, and/or psychotic disorders, underemployed, nontraditionally housed, and receiving treatment in a state psychiatric hospital or judicial system setting. The results attest to a necessity for transdiagnostic treatment planning and care coordination for this population, with particular attention to psychosocial stressors.

Personality disorders (PDs) are enduring patterns of internal experience and behavior that differ from cultural norms and expectations, are inflexible and pervasive, have their onset in adolescence or early adulthood, and lead to distress or impairment. Ten PDs are included in the Diagnostic and Statistical Manual of Mental Disorders (Fifth Edition): paranoid, schizoid, schizotypal, borderline, antisocial, histrionic, narcissistic, avoidant, dependent, and obsessive-compulsive.¹ These disorders impose a high burden on patients, families, health care systems, and broader economic systems.^2,3 Up to 1 in 7 persons in the community and 50% of those receiving outpatient mental health treatment experience a PD.^4,5 These conditions are associated with an increased risk of adverse events, including suicide attempt and death by suicide, criminal-legal involvement, homelessness, substance use, underemployment, relational issues, and high utilization of psychiatric services.^6-9 PDs are routinely underassessed, underdocumented, and undertreated in clinical settings, and consistently receive less research funding than other, less prevalent forms of psychopathology. ^10-12 As a result, there is limited understanding of clinical needs of individuals experiencing PDs.

MILITARY VETERANS WITH PERSONALITY DISORDERS

Underacknowledgment of PDs and their associated difficulties may be especially pronounced in veteran populations. Due to longstanding etiological theories that implicate childhood trauma and adolescent onset in pathology development, PDs are traditionally considered pre-existing conditions or developmental abnormalities by the US Department of Defense and US Department of Veterans Affairs (VA). As a result, PDs are therefore deemed incompatible with military service and ineligible for service-connected disability benefits.^13-15 Such determinations allowed PD pathology to be used as grounds for discharge for 26,000 service members from 2001 to 2007, or 2.6% of total enlisted discharges during that period.^13,15,16

Despite this structural discrimination, recent research suggests veterans may be more likely to experience PD pathology than the general population.17 For example, a 2021 epidemiological survey in a community-based veteran sample found elevated rates of borderline, antisocial, and schizotypal PDs (6%-13%).⁶ In contrast, only 0.8% to 5.0% of veteran electronic health records (EHRs) have a documented PD diagnosis.^8,18,19 Such elevations in PD pathology within veteran samples imply either a disproportionately high prevalence among enlistees (and therefore missed during recruitment procedures) or onset following military service, possibly due to exposure to traumatic events and/ or occupational stress.¹⁷ Due to the relative infancy of research in this area and a lack of longitudinal studies, etiology and course of illness for personality pathology in veterans remains largely unclear.

Structural underacknowledgment of PDs among military personnel has contributed to their underrepresentation in research on veteran populations. PD-focused research with veterans is rare, despite a rapid increase in broader empirical attention paid to these conditions in nonveteran samples.²⁰ A recent meta-analysis of veterans with PDs identified 27 studies that included basic prevalence statistics. PDs were rarely a primary focus for these studies, and most were limited to veterans seen in Veterans Health Administration (VHA) settings.¹⁷ The literature also paints a bleak picture, suggesting veterans who experience PDs are at higher risk for suicide attempt and death by suicide, criminal-legal involvement, and homelessness. They also tend to experience more severe comorbid psychopathological symptoms and more often use high-intensity mental health services (eg, care within emergency departments or psychiatric inpatient settings) than veterans without PD pathology.^6,8,18,19,21 However, PD pathology does not appear to impede the effectiveness of treatment for veterans.^22-24 The implications of PD pathology on broader psychosocial functioning and health care needs certify a need for additional research that examines patterns of personality pathology, particularly in veterans outside the VHA.

METHODS

This study aims to enhance understanding of veterans affected by PDs and offer insight and guidance for treatment of these conditions in federal and nonfederal treatment settings. Previous research has been largely limited to VHA care-receiving samples; the longstanding stigma against PDs by the US military and VA may contribute to biased diagnosis and documentation of PDs in these settings. A large sample of veterans receiving community-based mental health care was therefore used to explore aims of the current study. This study specifically examined demographic patterns, diagnostic comorbidity, psychosocial outcomes, and treatment care settings among veterans with and without a PD diagnosis. Consistent with previous research, we hypothesized that veterans with a PD diagnosis would have more severe mental health comorbidities, poorer psychosocial outcomes, and receive care in higher intensity settings relative to veterans without a diagnosis.

Data for the sample were drawn from the Mental Health Client-Level Data, a publicly available national dataset of nearly 7 million patients who received mental health treatment services provided or funded through state mental health agencies in 2022.²⁵ The analytic sample included about 2.5 million patients for whom veteran status and data around the presence or absence of a PD diagnosis were available. Of these patients, 104,198 were identified as veterans. Veteran patients were identified as predominantly male (63%), White (71%), non-Hispanic (90%), and never married (54%).

Measures

The parent dataset included demographic, clinical, and psychosocial outcome information reported by treatment facilities to individual state administrative systems for each patient who received services. To protect patient privacy, only nonprotected health information is included, and efforts were made throughout compilation of the parent dataset to ensure patient privacy (eg, limiting detail of information disseminated for public access). Because the parent dataset does not include protected health information, studies using these data are considered exempt from institutional review board oversight.

Demographic information. This study reviewed veteran status, sex, race, ethnicity, age, education, and marital status. Veteran status was defined by whether the patient was aged ≥ 18 years and had previously served (but was not currently serving) in the military. Patients with a history of service in the National Guard or Military Reserves were only classified as veterans if they had been called or ordered to active duty while serving. Sex was operationalized dichotomously as male or female; no patients were identified as intersex, transgender, or other gender identities.

Clinical information. Up to 3 mental health diagnoses were reported for each patient and included the following disorders: personality, trauma and attention-deficit/hyperactivity, stressor, anxiety, conduct, delirium/dementia, bipolar, depressive, oppositional defiant, pervasive developmental, schizophrenia or other psychotic, and alcohol or substance use. Mental health diagnosis categories were generated for the parent dataset by grouping diagnostic codes corresponding to each category. To protect patient privacy, more detailed diagnostic information was not available as part of the parent dataset. Although the American Psychiatric Association recognizes 10 distinct PDs, the exact nature of PD diagnoses was not included within the parent dataset. PD diagnoses were coded to reflect the presence or absence of any such diagnosis.

A substance use problem designation was also provided for patients according to various identification methods, including substance use disorder (SUD) diagnosis, substance use screening results, enrollment in a substance use program, substance use survey, service claims information, and other related sources of information. A severe mental illness or serious emotional disturbance designation was provided for patients meeting state definitions of these designations. Context(s) of service provision were coded as inpatient state psychiatric hospital, community-based program, residential treatment center, judicial institution, or other psychiatric inpatient setting.

Psychosocial outcome information. Patient employment and residential status were also included in analyses. Each reflected status at the time of discharge from services or end of reporting period; employment status was only provided for patients receiving treatment in community-based programs.

Data Analysis

Descriptive statistics and X² analyses were used to compare demographic, clinical, and psychosocial outcome variables between patients with and without PD diagnoses. These analyses were calculated for both the 104,198 veterans and the 2,222,306 nonveterans aged ≥ 18 years in the dataset. Given the sample size, a conservative α of .01 was used to determine statistical significance.

RESULTS

In this sample of persons receiving state-funded mental health care, veterans were significantly less likely than nonveterans to have a documented PD diagnosis (2.1% vs 3.6%, X² [1] = 647.49; P < .01). PD diagnoses were more common among White (risk ratio [RR], 1.11), non-Hispanic (RR, 1.03) veterans who were in middle to late adulthood (RR, 1.16-1.40), more educated (RR, 1.35), and divorced or widowed (RR, 1.43), and less common among Black/African American (RR, 0.78) or Puerto Rican (RR, 0.32) veterans who were in early adulthood (RR, 0.31-0.79), less educated (RR, 0.64-0.89), and currently married (RR, 0.89) or never married (RR, 0.86). Veteran men and women were equally likely to have a PD diagnosis (RR, 1.03) (Table 1). Among nonveterans, men were less likely than women to have a PD diagnosis (RR, 0.79), and PD diagnoses were most common among persons in middle adulthood (RR, 1.06-1.15) (eAppendix 1).

Veterans with a PD diagnosis were more likely than those without a diagnosis to have more diagnoses (RR, 2.96-8.49) and to have comorbid trauma or related stressor (RR, 1.33), or bipolar (RR, 1.56) or psychotic (RR, 1.15) disorder diagnoses, but less likely to have comorbid depressive disorder (RR, 0.82). Although veterans with and without a PD diagnosis were similarly likely to have a comorbid SUD (RR, 1.13), those with a PD diagnosis were significantly less likely to be assigned a substance use problem designation (RR, 0.78). PD diagnosis was also more common among veterans who received services in state psychiatric hospitals (RR, 3.05), community-based clinics (RR, 1.06), and judicial institutions (RR, 6.33) and less common among those who received services in other psychiatric inpatient settings (RR, 0.30). No differences were observed for residential treatment settings (RR, 0.79). Among nonveterans, a PD diagnosis was associated with slightly greater odds of a substance use designation (RR, 1.03) (eAppendix 2).

Veterans with a PD diagnosis were also less likely to have full-time employment (RR, 0.73) and more likely to have undifferentiated employment (RR, 2.00) or to be removed from the labor force (RR, 1.35). Veterans with a PD diagnosis were also more likely to reside in nontraditional living conditions (RR, 1.42) and less likely to be residing in a private residence (RR, 0.98), compared with those without PD diagnosis. The rates of homelessness were similar for veterans with and without a PD diagnosis (RR, 0.90) (Table 2). These patterns were similar among nonveterans.

DISCUSSION

This study examined the rate and correlates of PD diagnosis among a large, community-based sample of veterans receiving state-funded mental health care. About 2% of veterans in this sample had a PD diagnosis, with diagnoses more common among veterans who were White, non-Hispanic, aged ≥ 45 years, with higher education, divorced or widowed, also diagnosed with trauma-related, bipolar, and/or psychotic disorders, underemployed, nontraditionally housed, and receiving treatment in state psychiatric hospital, community-based clinic, or judicial system settings.

The observed rate of PD diagnosis in this study aligns with what is typically observed in VHA EHRs.^8,18,19 However, the rate is notably lower than prevalence estimates for psychiatric outpatient settings (about 50%) and in meta-analyses of prevalence among veterans (0.8%-23% for each of the 10 PDs).^4,17,26 Longstanding stigma against PDs may contribute to underdiagnosis. For example, many clinicians are concerned that documentation or disclosure of a PD will interfere with the patient’s ability to access treatment due to stigma and discrimination.^27,28 These fears are not unfounded; even among clinicians, PDs are commonly considered untreatable, and many individuals with PDs are denied access to evidence-based treatments due to the diagnosis.²⁹ In a 2016 survey of community psychiatrists, nearly 1 in 4 reported that they avoid taking patients with a borderline PD diagnosis in their caseloads.²⁸ To date, no studies have been conducted to explore clinicians’ willingness to accept patients with other PDs or, specifically, among veterans.

Despite such widespread stigma, research suggests clinicians' negative attitudes toward PDs can be decreased through antistigma campaigns.³⁰ However, it remains unclear if such efforts also contribute to an increase in clinicians’ willingness to document PD diagnoses. Without accurate identification and documentation, the field’s understanding of PDs will remain limited.

In the current study, veterans with PD diagnoses tended to present with more complex and severe psychiatric comorbidities compared to veterans without such diagnoses. Observed comorbidity of PDs (particularly borderline PD) with trauma-related and bipolar disorders is well established.8 Conversely, co-occurring personality and psychotic disorders—which comprise 16% of veterans with a PD diagnosis in the sample in this study—are not consistently examined in the literature. A 2022 examination of veterans receiving VHA care suggested 12% and 13% of those with a PD diagnosis documented in their EHR also had documented schizophrenia or another psychotic disorder, respectively. PD diagnoses were associated with 6.88- and 9.80-fold increases in risk for comorbid schizophrenia and other psychotic disorder diagnoses, respectively.⁸ Similarly, a recent longitudinal study of nearly 2 million Swedish individuals suggested borderline PD is specifically associated with a > 24-times greater risk of having a comorbid psychotic disorder.³¹ It is therefore possible that the comorbidity between personality and psychotic disorders is quite common despite its relative lack of attention in empirical research.

Veterans with PD diagnoses in this study were also more likely to experience substandard housing, employment challenges, and receive treatment through judicial institutions than those without a PD diagnosis. Such findings are consistent with previous research demonstrating the substantial psychosocial challenges associated with PD diagnosis, even after controlling for comorbid conditions.^7,9 Veterans with PDs may benefit from specialized case management and support to facilitate stable housing and employment and to mitigate the risk of judicial involvement. Some research suggests veterans with PDs may be less likely to gain competitive employment after participating in VA therapeutic and supportive employment services programs, suggesting standard programming may be less suitable for this population.³² Similarly, other research suggests individuals with PDs may benefit more from specialized, intensive services than standard clinical case management.³³ Future research may therefore benefit from clarifying the degree to which adaptations to standard programming could yield beneficial effects for persons with PD diagnoses.

Implications

Cumulatively, the results of this study attest to the necessity for transdiagnostic treatment planning that includes close collaboration between psychotherapeutic, pharmacological, and case management services. Some psychotherapy models for PDs, such as dialectical behavior therapy (DBT), which includes a combination of group skills training, individual therapy, as-needed phone coaching, and therapist consultation, may be successfully adapted to include this collaboration.^34-36 However, implementation of such comprehensive programming often requires extensive clinician training and coordination of resources, which poses implementation challenges.^37-39 In 2021, the VHA began large-scale implementation of PD-specific psychotherapy for veterans with recent suicidal self-directed violence and borderline PD, including DBT, though to date results remain unclear.⁴⁰ Generalist approaches, such as good psychiatric management (GPM), which emphasizes emotional validation, practical problem solving, realistic goal setting, and relationship functioning within the context of standard care appointments, may be more easily implemented in community care settings due to lesser training and resource requirements and can also be adapted to include needed elements of care coordination.^41,42 Both DBT and GPM were initially developed for the treatment of borderline PD. Although DBT has also demonstrated some effectiveness in the treatment of antisocial PD, potential applications of DBT and GPM to other PDs remain largely underdeveloped.^43-46

There are no widely accepted medications for the treatment of PDs. Pharmacotherapy for these conditions typically consists of individualized approaches informed by personal experience that attempt to balance targeting of specific symptoms while minimizing polypharmacy and potential risks (eg, overdose or addiction).^47,48 Despite this, pharmacotherapy is often considered a necessary component in the treatment of bipolar and psychotic disorders, both common comorbidities of PDs found in veterans in this study.^49,50 Careful consideration of complex comorbidities and pharmacotherapy needs is warranted in the treatment of veterans with PDs. Future research may benefit from clarifying clinical guidelines around pharmacotherapy, particularly for observed comorbidities of PDs to trauma, bipolar, and psychotic disorders.

It is important to note the discrepancies in the results of this study surrounding patient substance use. The results suggest a negligible or inverse association between the likelihood of a PD diagnosis and difficulties with substance use among the veterans in this study. However, the unexpectedly low rate of SUD diagnoses (< 6%) suggests that they were likely underdocumented. Research suggests a strong association between personality and SUDs in both veteran and civilian samples.^6,51 Results suggesting a lower prevalence of substance use difficulties among treatment-seeking veterans with PDs should be interpreted with great caution.

Demographically, PD diagnoses were more common among veterans who were White, non-Hispanic, and aged ≥ 45 years, and less common among veterans who were Black/ African American, mixed/unspecified race, Puerto Rican or other non-Mexican Hispanic ethnicity, or aged < 35 years. No significant sex-based differences were observed. These patterns are consistent with research suggesting individuals who identify as Black may be less likely than individuals who identify as White to report PD symptoms, meet criteria for a PD, and have a PD diagnosed even when it is warranted.⁵²

The findings observed in this study with respect to age, however, are notably inconsistent with the literature. Previous research typically suggests a negative association between age and PD pathology; however, a 2020 review of PDs in older adults by Penders et al suggests a prevalence of 11% to 15% in this population.^53,54 Research into PDs most often focuses on adolescent and early adulthood developmental periods, limiting insight into the phenomenology of PDs in middle to late adulthood.⁵⁵ Further, most research into PDs among geriatric populations has focused on psychometric assessment rather than practical treatment guidance.⁵⁴ However, in this study, elevated risk for PD diagnoses was salient throughout middle to late adulthood among veterans; similar, albeit less pronounced patterns were also observed for elevated risk of PD diagnosis in middle adulthood among nonveterans. Such findings suggest clarifying the phenomenology and treatment needs of individuals with PDs in middle to late adulthood may have particularly salient implications for the mental health care of veterans affected by these conditions. As the veteran population advances in age, these needs will present unique challenges if health care systems are unprepared to effectively address them.

Limitations

This study is characterized by several strengths, most notably its use of a large dataset recently collected on a national scale. Few studies outside of the VHA system include samples of > 100,000 treatment-seeking veterans collected on a national scale. Nevertheless, results should be understood within the context of several methodological limitations. However, the dataset was limited to the first 3 diagnoses documented in patients’ EHRs, and many patients had no listed diagnoses. Patients with complex comorbidities may have > 3 diagnoses; for these individuals, data provided an incomplete picture of clinical presentation. This is especially relevant for individuals with PDs, who tend to meet criteria for a range of comorbid conditions.^8,10 The now dated practice of listing PDs on Axis II also increases the chance of clinicians listing PDs after conditions traditionally listed on Axis I (eg, major depressive disorder) in patient charts.⁵⁶ This study’s inclusion of only the first 3 listed diagnoses likely underestimated true PD diagnosis prevalence.

The results of this study must be interpreted as reflecting the prevalence and correlates of receiving a PD diagnosis rather than meeting diagnostic criteria for a PD. Relatedly, PD diagnoses were reported as a single construct, limiting insight into prevalence and correlates of individual PD diagnoses (eg, borderline vs paranoid PDs). Meta-analyses estimates suggest PD prevalence among veterans is likely much higher than observed in this study.¹⁷ Stigma continues to discourage clinicians from documenting and disclosing PD diagnoses even when warranted.^27,28 Continued research should aim to clarify conditions (eg, patient presentation, stigma, or institutional culture) contributing to documentation of PD diagnoses. Given the cross-sectional nature of this study, results cannot speak to longitudinal treatment outcomes or prognosis of persons receiving a PD diagnosis.

Despite its large sample size and national representation, the sampling strategy of this study could have contributed to idiosyncrasies in the dataset. Restriction of data to the persons receiving state-funded mental health services introduces a notable bias to the composition of the sample, which is likely comprised of a disproportionately high number of Medicaid recipients, students, and individuals with chronic illnesses and underrepresentation of persons who pay for mental health services using private insurance or private pay arrangements. As such, although socioeconomic information was not provided within this dataset, one can presume a generally lower socioeconomic status among study participants compared to the community at large. This study also included a proportionally small sample of veterans (3.6% compared to about 6.2% in the broader US population), suggesting veterans may have been underrepresented or underidentified in surveyed mental health care settings.⁵⁷ This study also did not include data around service in active-duty military, national guard, or military reserves; a greater proportion of the sample likely had a history of military service than was represented by veteran status designation. Further, the proportionally high sample of individuals with severe mental illness suggests a likely overrepresentation of such conditions in surveyed settings.

Institutional differences in the practice of assigning diagnoses likely limited statistical power to detect potentially meaningful associations and effects. Structural influences, such as stigma and institutional culture, may have notable effects on documentation practices, particularly for PDs. Future research should aim to replicate observed associations using more controlled diagnostic procedures.

Lastly, even with the use of a more conservative α and a focus on effect sizes to guide interpretation of results, use of multiple bivariate analyses can be presumed to have increased the likelihood of type I error. Given the limited prior research in this area, an exploratory approach to statistical analysis was considered warranted to maximize opportunity for identifying areas in need of additional empirical attention. Continued research using more conservative statistical approaches (eg, multivariate analyses) is needed to determine replicability and generalizability of observed results.

CONCLUSIONS

This study examined the prevalence and correlates of PD diagnoses in a national sample of veterans receiving community-based, state-funded mental health care. About 2% received a PD diagnosis, with diagnoses most common among veterans who were White, non-Hispanic, aged ≥ 45 years, also diagnosed with trauma-based, bipolar, and/or psychotic disorders, underemployed, nontraditionally housed, and receiving treatment in a state psychiatric hospital or judicial system setting. The results attest to a necessity for transdiagnostic treatment planning and care coordination for this population, with particular attention to psychosocial stressors.

In April 2022, the US Department of Veterans Affairs (VA) revised its behavioral health care eligibility policies to provide comprehensive mental and behavioral health care to former service members who received an Other Than Honorable (OTH) discharge characterization upon separation from military service.¹ This policy shift represents a marked expansion in eligibility practices (Table 1 includes amended eligibility criteria).

Since June 2017, eligibility policies allowed veterans with OTH discharges to receive “emergent mental health services” needed to stabilize acute mental health crises related to military service (eg, acute escalations in suicide risk).^2,3 Previously, veterans with OTH discharges were largely ineligible for VA-based health care; these individuals were only able to access Veterans Health Administration (VHA) mental and behavioral health care through limited channels of eligibility (eg, for treatment of military sexual trauma or psychosis or other mental illness within 2 years of discharge).^4,5 The impetus for expansions in eligibility stemmed from VA efforts to reduce the suicide rate among veterans.^6-8 Implications of such expansion extend beyond suicide prevention efforts, with notable promised effects on the care of veterans with criminal-legal involvement. This article highlights potential effects of recent eligibility expansions on veterans with criminal-legal involvement and makes specific recommendations for agencies and organizations serving these veterans.

OTHER THAN HONORABLE DISCHARGE

The US Department of Defense delineates 6 discharge characterizations provided to service members upon separation from military service: honorable, general under honorable conditions, OTH, bad conduct, dishonorable, and uncharacterized. Honorable discharge characterizations are considered to reflect general concordance between service member behavior and military standards; general discharge characterizations reflect some disparity between the service member’s behavior and military standards; OTH, bad conduct, and dishonorable discharge characterizations reflect serious disparities between the service member’s behavior and military standards; and uncharacterized discharge characterizations are given when other discharge characterizations are deemed inappropriate.^9,10 OTH discharge characterizations are typically issued under instances of misconduct, fraudulent entry, security reasons, or in lieu of trial by court martial.^9,10

Recent research suggests that about 85% of service members receive an honorable discharge characterization upon separation from military service, 8% receive general, 6% receive OTH, and 1% receive bad conduct or dishonorable discharges.11 In 2017, the VA estimated there were > 500,000 prior service members with OTH discharge characterizations, which has grown over time (1.9% during the Korean Conflict, 2.5% during the Vietnam War Era, 3.9% during the Cold War, 4.8% in the Persian Gulf War, and 5.8% in the post-9/11 era).^7,11

The OTH discharge characterization is 1 of 3 less than honorable discharges informally referred to as bad papers (ie, OTH, bad conduct, or dishonorable). Former service members receiving these discharge characterizations face significant social stigma and structural discrimination upon military discharge, including significant hurdles to employment and educational pursuits as well as notable social alienation.¹² Due to their discharge characterization, some are viewed as less deserving of the veteran title, and until recently, many did not qualify for the complex legal definition of veteran as established by the Congress.^11,13 Veterans with OTH discharge characterizations have also historically been excluded from services (eg, VHA care),³ benefits (eg, disability compensation),¹⁴ and protections (eg, Uniformed Services Employment and Reemployment Rights Act)¹⁵ offered to veterans with honorable or general discharge characterizations. However, eligibility policies have gradually expanded, providing veterans with OTH discharges with access to VHA-based mental and behavioral health services and VA supportive housing assistance.^1,3,16

Perhaps due to their historical exclusion from VA services, there is limited research available on the behavioral health and associated needs of veterans with OTH discharges. Some scholars argue that historical exclusions have exacerbated underlying difficulties faced by this population, thereby contributing to stark health and social disparities across discharge types.^14,15,17 Studies with large veteran samples, for example, reflect notable demographic and behavioral health differences across discharge types. Compared to routinely discharged veterans, veterans with OTH discharges are significantly more likely to be younger, have lower income, use substances, have a history of criminal-legal involvement, and have mental and physical health difficulties.^18,19

Substantial evidence also suggests a historical racial bias, with service members of color being disproportionately more likely to receive an OTH discharge.¹² Similarly, across all branches of military service, Black service members are significantly more likely to face general or special court martial in military justice proceedings when compared with White service members.²⁰ Service members from gender and sexual minorities are also disproportionately impacted by the OTH designation. Historically, many have been discharged with bad papers due to discriminatory policies, such as Don’t Ask Don’t Tell, which discriminated on the basis of sexual orientation between December 1993 and September 2011, and Directive-type Memorandum-19-004, which banned transgender persons from military service between April 2019 and January 2021.^21,22

There is also significant mental health bias in the provision of OTH discharges, such that OTH characterizations are disproportionately represented among individuals with mental health disorders.^18-20 Veterans discharged from military service due to behavioral misconduct are significantly more likely to meet diagnostic criteria for various behavioral health conditions and to experience homelessness, criminal-legal involvement, and suicidal ideation and behavior compared with routinely-discharged veterans.^23-28

Consistent with their comparatively higher rates of criminal-legal involvement relative to routinely discharged veterans, veterans with OTH discharges are disproportionately represented in criminal justice settings. While veterans with OTH discharges represent only 6% of discharging service members and 2.5% of community-based veterans, they represent 10% of incarcerated veterans.^11,18,23,29 Preliminary research suggests veterans with OTH discharges may be at higher risk for lifetime incarceration, though the association between OTH discharge and frequency of lifetime arrests remains unclear.^18,30

VETERANS TREATMENT COURTS

Given the overrepresentation of veterans with OTH discharges in criminal-legal settings, consideration for this subset of the veteran population and its unique needs is commonplace among problem-solving courts that service veterans. First conceptualized in 2004, Veterans Treatment Courts (VTCs) are specialized problem-solving courts that divert veterans away from traditional judicial court and penal systems and into community-based supervision and treatment (most commonly behavioral health services).^31-34 Although each VTC program is unique in structure, policies, and procedures, most VTCs can be characterized by certain key elements, including voluntary participation, plea requirements, delayed sentencing (often including reduced or dismissed charges), integration of military culture into court proceedings, a rehabilitative vs adversarial approach to decreasing risk of future criminal behavior, mandated treatment and supervision during participation, and use of veteran mentors to provide peer support.^32-35 Eligibility requirements vary; however, many restrict participation to veterans with honorable discharge types and charges for nonviolent offenses.^32,33,35-37

VTCs connect veterans within the criminal-legal system to needed behavioral health, community, and social services.^31-33,37 VTC participants are commonly connected to case management, behavioral health care, therapeutic journaling programs, and vocational rehabilitation.^38,39 Accordingly, the most common difficulties faced by veterans participating in these courts include substance use, mental health, family issues, anger management and/or aggressive behavior, and homelessness.^36,39 There is limited research on the effectiveness of VTCs. Evidence on their overall effectiveness is largely mixed, though some studies suggest VTC graduates tend to have lower recidivism rates than offenders more broadly or persons who terminate VTC programs prior to completion.^40,41 Other studies suggest that VTC participants are more likely to have jail sanctions, new arrests, and new incarcerations relative to nontreatment court participants.⁴² Notably, experimental designs (considered the gold standard in assessing effectiveness) to date have not been applied to evaluate the effectiveness of VTCs; as such, the effectiveness of these programs remains an area in need of continued empirical investigation.

Like all problem-solving courts, VTCs occasionally struggle to connect participating defendants with appropriate care, particularly when encountering structural barriers (eg, insurance, transportation) and/or complex behavioral health needs (eg, personality disorders).^34,43 As suicide rates among veterans experiencing criminal-legal involvement surge (about 150 per 100,000 in 2021, a 10% increase from 2020 to 2021 compared to about 40 per 100,000 and a 1.8% increase among other veterans), efficiency of adequate care coordination is vital.⁴⁴ Many VTCs rely on VTC-VA partnerships and collaborations to navigate these difficulties and facilitate connection of participating veterans to needed services.^32-34,45 For example, within the VHA, Veterans Justice Outreach (VJO) and Health Care for Re-Entry Veterans (HCRV) specialists assist and bridge the gap between the criminal-legal system (including, but not limited to VTCs) and VA services by engaging veterans involved in the criminal-legal system and connecting them to needed VA-based services (Table 2). Generally, VJO specialists support veterans involved with the front end of the criminallegal system (eg, arrest, pretrial incarceration, or participation in VTCs), while HCRV specialists tend to support veterans transitioning back into the community after a period of incarceration.^46,47 Specific to VTCs, VJO specialists typically serve as liaisons between the courts and VA, coordinating VA services for defendants to fulfill their terms of VTC participation.⁴⁶

The historical exclusion of veterans with OTH discharge characterizations from VA-based services has restricted many from accessing VTC programs.³² Of 17 VTC programs active in Pennsylvania in 2014, only 5 accepted veterans with OTstayH discharges, and 3 required application to and eligibility for VA benefits.³³ Similarly, in national surveys of VTC programs, about 1 in 3 report excluding veterans deemed ineligible for VA services.^35,36 When veterans with OTH discharges have accessed VTC programs, they have historically relied on non-VA, community-based programming to fulfill treatment mandates, which may be less suited to addressing the unique needs of veterans.⁴⁸

Veterans who utilize VTCs receive several benefits, namely peer support and mentorship, acceptance into a veteran-centric space, and connection with specially trained staff capable of supporting the veteran through applications for a range of VA benefits (eg, service connection, housing support).^31-33,37 Given the disparate prevalence of OTH discharge characterizations among service members from racial, sexual, and gender minorities and among service members with mental health disorders, exclusion of veterans with OTH discharges from VTCs solely based on the type of discharge likely contributes to structural inequities among these already underserved groups by restricting access to these potential benefits. Such structural inequity stands in direct conflict with VTC best practice standards, which admonish programs to adjust eligibility requirements to facilitate access to treatment court programs for historically marginalized groups.⁴⁹

ELIGIBILITY EXPANSIONS

Given the overrepresentation of veterans with OTH discharge characterizations within the criminal-legal system and historical barriers of these veterans to access needed mental and behavioral health care, expansions in VA eligibility policies could have immense implications for VTCs. First, these expansions could mitigate common barriers to connecting VTC-participating veterans with OTH discharges with needed behavioral health care by allowing these veterans to access established, VA-based services and programming. Expansion may also allow VTCs to serve as a key intercept point for identifying and engaging veterans with OTH discharges who may be unaware of their eligibility for VA-based behavioral health care.

Access to VA health care services is a major resource for VTC participants and a common requirement.³² Eligibility expansion should ease access barriers veterans with OTH discharges commonly face. By providing a potential source of treatment, expansions may also support OTH eligibility practices within VTCs, particularly practices that require participants to be eligible for VA health care.^33,35,36 Some VTCs may continue to determine eligibility on the basis of discharge status and remain inaccessible to veterans with OTH discharge characterizations without program-level policy changes.^32,36,37

Communicating changes in eligibility policies relevant to veterans with OTH discharges may be a challenge, because many of these individuals have no established channels of communication with the VA. Because veterans with OTH discharges are at increased risk for legal system involvement, VTCs may serve as a unique point of contact to help facilitate communication.¹⁸ For example, upon referral to a VTC, veterans with OTH discharges can be identified, VA health care eligibility can be verified, and veterans can connect to VA staff to facilitate enrollment in VA services and care.

VJO specialists are in a favorable position to serve a critical role in utilizing VTCs as a potential intercept point for engaging veterans with OTH discharge characterizations. As outlined in the STRONG Veterans Act of 2022, VJOs are mandated to “spread awareness and understanding of veteran eligibility for the [VJO] Program, including the eligibility of veterans who were discharged from service in the Armed Forces under conditions other than honorable.”50 The Act further requires VJOs to be annually trained in communicating eligibility changes as they arise. Accordingly, VJOs receive ongoing training in a wide variety of critical outreach topics, including changes in eligibility; while VJOs cannot make eligibility determinations, they are tasked with enrolling all veterans involved in the criminal-legal system with whom they interact into VHA services, whether through typical or special eligibility criteria (M. Stimmel, PhD, National Training Director for Veteran Justice Programming, oral communication, July 14, 2023). VJOs therefore routinely serve in this capacity of facilitating VA enrollment of veterans with OTH discharge characterizations.

Recommendations to Veteran-Servicing Judicial Programs

Considering these potential implications, professionals routinely interacting with veterans involved in the criminal-legal system should become familiarized with recent changes in VA eligibility policies. Such familiarization would support the identification of veterans previously considered ineligible for care; provision of education to these veterans regarding their new eligibility; and referral to appropriate VA-based behavioral health care options. Although conceptually simple, executing such an educational campaign may prove logistically difficult. Given their historical exclusion from VA services, veterans with OTH discharge characterizations are unlikely to seek VA-based services in times of need, instead relying on a broad swath of civilian community-based organizations and resources. Usual approaches to advertising VHA health care policy changes (eg, by notifying VA employees and/or departments providing corresponding services or by circulating information to veteran-focused mailing lists and organizations) likely would prove insufficient. Educational campaigns to disseminate information about recent OTH eligibility changes should instead consider partnering with traditionally civilian, communitybased organizations and institutions, such as state bar associations, legal aid networks, case management services, nonveteran treatment court programs (eg, drug courts, or domestic violence courts), or probation/ parole programs. Because national surveys suggest generally low military cultural competence among civilian populations, providing concurrent support in developing foundational veteran cultural competencies (eg, how to phrase questions about military service history, or understanding discharge characterizations) may be necessary to ensure effective identification and engagement of veteran clients.⁴⁸

Programs that serve veterans with criminal-legal involvement should also consider potential relevance of recent OTH eligibility changes to program operations. VTC program staff and key partners (eg, judges, case managers, district attorneys, or defense attorneys), should revisit policies and procedures surrounding the engagement of veterans with OTH discharges within VTC programs and strategies for connecting these veterans with needed services. VTC programs that have historically excluded veterans with OTH discharges due to associated difficulties in locating and connecting with needed services should consider expanding eligibility policies considering recent shifts in VA behavioral health care eligibility.^33,35,36 Within the VHA, VJO specialists can play a critical role in supporting these VTC eligibility and cultural shifts. Some evidence suggests a large proportion of VTC referrals are facilitated by VJO specialists and that many such referrals are identified when veterans involved with the criminal-legal system seek VA support and/or services.³³ Given the historical exclusion of veterans with OTH discharges from VA care, strategies used by VJO specialists to identify, connect, and engage veterans with OTH discharges with VTCs and other services may be beneficial.

Even with knowledge of VA eligibility changes and considerations of these changes on local operations, many forensic settings and programs struggle to identify veterans. These difficulties are likely amplified among veterans with OTH discharge characterizations, who may be hesitant to self-disclose their military service history due to fear of stigma and/or views of OTH discharge characterizations as undeserving of the veteran title.¹² The VA offers 2 tools to aid in identification of veterans for these settings: the Veterans Re-Entry Search Service (VRSS) and Status Query and Response Exchange System (SQUARES). For VRSS, correctional facilities, courts, and other criminal justice entities upload a simple spreadsheet that contains basic identifying information of inmates or defendants in their system. VRSS returns information about which inmates or defendants have a history of military service and alerts VA Veterans Justice Programs staff so they can conduct outreach. A pilot study conducted by the California Department of Corrections and Rehabilitation found that 2.7% of its inmate population self-identified as veterans, while VRSS identified 7.7% of inmates with a history of military service. This difference represented about 5000 previously unidentified veterans.⁵¹ Similarly, community entities that partner with the VA, such as law enforcement or homeless service programs, can be approved to become a SQUARES user and submit identifying information of individuals with whom they interact directly into the SQUARES search engine. SQUARES then directly returns information about the individual’s veteran status and eligibility for VA health care and homeless programs.

Other Eligibility Limitations

VTCs and other professionals looking to refer veterans with OTH discharge characterizations to VA-based behavioral health care should be aware of potential limitations in eligibility and access. Specifically, although veterans with OTH discharges are now broadly eligible for VA-based behavioral health care and homeless programs, they remain ineligible for other forms of health care, including primary care and nonbehavioral specialty care.1 Research has found a strong comorbidity between behavioral and nonbehavioral health concerns, particularly within historically marginalized demographic groups.^52-55 Because these historically marginalized groups are often overrepresented among persons with criminal-legal involvement, veterans with OTH discharges, and VTC participants, such comorbidities require consideration by services or programming designed to support veterans with criminal-legal involvement.^12,56-58 Connection with VA-based health care will therefore continue to fall short of addressing all health care needs of veterans with OTH discharges and effective case management will require considerable treatment coordination between VA behavioral health care practitioners (HCPs) and community-based HCPs (eg, primary care professionals or nonbehavioral HCPs).

Implications for VA Mental Health Care

Recent eligibility expansions will also have inevitable consequences for VA mental health care systems. For many years, these systems have been overburdened by high caseloads and clinician burnout.^59,60 Given the generally elevated rates of mental health and substance use concerns among veterans with OTH discharge characterizations, expansions hold the potential to further burden caseloads with clinically complex, high-risk, high-need clients. Nevertheless, these expansions are also structured in a way that forces existing systems to absorb the responsibilities of providing necessary care to these veterans. To mitigate detrimental effects of eligibility expansions on the broader VA mental health system, clinicians should be explicitly trained in identifying veterans with OTH discharge characterizations and the implications of discharge status on broader health care eligibility. Treatment of veterans with OTH discharges may also benefit from close coordination between mental health professionals and behavioral health care coordinators to ensure appropriate coordination of care between VA- and non–VA-based HCPs.

CONCLUSIONS

Recent changes to VA eligibility policies now allow comprehensive mental and behavioral health care services to be provided to veterans with OTH discharges.1 Compared to routinely discharged veterans, veterans with OTH discharges are more likely to be persons of color, sexual or gender minorities, and experiencing mental health-related difficulties. Given the disproportionate mental health burden often faced by veterans with OTH discharges and relative overrepresentation of these veterans in judicial and correctional systems, these changes have considerable implications for programs and services designed to support veterans with criminallegal involvement. Professionals within these systems, particularly VTC programs, are therefore encouraged to familiarize themselves with recent changes in VA eligibility policies and to revisit strategies, policies, and procedures surrounding the engagement and enrollment of veterans with OTH discharge characterizations. Doing so may ensure veterans with OTH discharges are effectively connected to needed behavioral health care services.

In April 2022, the US Department of Veterans Affairs (VA) revised its behavioral health care eligibility policies to provide comprehensive mental and behavioral health care to former service members who received an Other Than Honorable (OTH) discharge characterization upon separation from military service.¹ This policy shift represents a marked expansion in eligibility practices (Table 1 includes amended eligibility criteria).

Since June 2017, eligibility policies allowed veterans with OTH discharges to receive “emergent mental health services” needed to stabilize acute mental health crises related to military service (eg, acute escalations in suicide risk).^2,3 Previously, veterans with OTH discharges were largely ineligible for VA-based health care; these individuals were only able to access Veterans Health Administration (VHA) mental and behavioral health care through limited channels of eligibility (eg, for treatment of military sexual trauma or psychosis or other mental illness within 2 years of discharge).^4,5 The impetus for expansions in eligibility stemmed from VA efforts to reduce the suicide rate among veterans.^6-8 Implications of such expansion extend beyond suicide prevention efforts, with notable promised effects on the care of veterans with criminal-legal involvement. This article highlights potential effects of recent eligibility expansions on veterans with criminal-legal involvement and makes specific recommendations for agencies and organizations serving these veterans.

OTHER THAN HONORABLE DISCHARGE

The US Department of Defense delineates 6 discharge characterizations provided to service members upon separation from military service: honorable, general under honorable conditions, OTH, bad conduct, dishonorable, and uncharacterized. Honorable discharge characterizations are considered to reflect general concordance between service member behavior and military standards; general discharge characterizations reflect some disparity between the service member’s behavior and military standards; OTH, bad conduct, and dishonorable discharge characterizations reflect serious disparities between the service member’s behavior and military standards; and uncharacterized discharge characterizations are given when other discharge characterizations are deemed inappropriate.^9,10 OTH discharge characterizations are typically issued under instances of misconduct, fraudulent entry, security reasons, or in lieu of trial by court martial.^9,10

Recent research suggests that about 85% of service members receive an honorable discharge characterization upon separation from military service, 8% receive general, 6% receive OTH, and 1% receive bad conduct or dishonorable discharges.11 In 2017, the VA estimated there were > 500,000 prior service members with OTH discharge characterizations, which has grown over time (1.9% during the Korean Conflict, 2.5% during the Vietnam War Era, 3.9% during the Cold War, 4.8% in the Persian Gulf War, and 5.8% in the post-9/11 era).^7,11

The OTH discharge characterization is 1 of 3 less than honorable discharges informally referred to as bad papers (ie, OTH, bad conduct, or dishonorable). Former service members receiving these discharge characterizations face significant social stigma and structural discrimination upon military discharge, including significant hurdles to employment and educational pursuits as well as notable social alienation.¹² Due to their discharge characterization, some are viewed as less deserving of the veteran title, and until recently, many did not qualify for the complex legal definition of veteran as established by the Congress.^11,13 Veterans with OTH discharge characterizations have also historically been excluded from services (eg, VHA care),³ benefits (eg, disability compensation),¹⁴ and protections (eg, Uniformed Services Employment and Reemployment Rights Act)¹⁵ offered to veterans with honorable or general discharge characterizations. However, eligibility policies have gradually expanded, providing veterans with OTH discharges with access to VHA-based mental and behavioral health services and VA supportive housing assistance.^1,3,16

Perhaps due to their historical exclusion from VA services, there is limited research available on the behavioral health and associated needs of veterans with OTH discharges. Some scholars argue that historical exclusions have exacerbated underlying difficulties faced by this population, thereby contributing to stark health and social disparities across discharge types.^14,15,17 Studies with large veteran samples, for example, reflect notable demographic and behavioral health differences across discharge types. Compared to routinely discharged veterans, veterans with OTH discharges are significantly more likely to be younger, have lower income, use substances, have a history of criminal-legal involvement, and have mental and physical health difficulties.^18,19

Substantial evidence also suggests a historical racial bias, with service members of color being disproportionately more likely to receive an OTH discharge.¹² Similarly, across all branches of military service, Black service members are significantly more likely to face general or special court martial in military justice proceedings when compared with White service members.²⁰ Service members from gender and sexual minorities are also disproportionately impacted by the OTH designation. Historically, many have been discharged with bad papers due to discriminatory policies, such as Don’t Ask Don’t Tell, which discriminated on the basis of sexual orientation between December 1993 and September 2011, and Directive-type Memorandum-19-004, which banned transgender persons from military service between April 2019 and January 2021.^21,22

There is also significant mental health bias in the provision of OTH discharges, such that OTH characterizations are disproportionately represented among individuals with mental health disorders.^18-20 Veterans discharged from military service due to behavioral misconduct are significantly more likely to meet diagnostic criteria for various behavioral health conditions and to experience homelessness, criminal-legal involvement, and suicidal ideation and behavior compared with routinely-discharged veterans.^23-28

Consistent with their comparatively higher rates of criminal-legal involvement relative to routinely discharged veterans, veterans with OTH discharges are disproportionately represented in criminal justice settings. While veterans with OTH discharges represent only 6% of discharging service members and 2.5% of community-based veterans, they represent 10% of incarcerated veterans.^11,18,23,29 Preliminary research suggests veterans with OTH discharges may be at higher risk for lifetime incarceration, though the association between OTH discharge and frequency of lifetime arrests remains unclear.^18,30

VETERANS TREATMENT COURTS

Given the overrepresentation of veterans with OTH discharges in criminal-legal settings, consideration for this subset of the veteran population and its unique needs is commonplace among problem-solving courts that service veterans. First conceptualized in 2004, Veterans Treatment Courts (VTCs) are specialized problem-solving courts that divert veterans away from traditional judicial court and penal systems and into community-based supervision and treatment (most commonly behavioral health services).^31-34 Although each VTC program is unique in structure, policies, and procedures, most VTCs can be characterized by certain key elements, including voluntary participation, plea requirements, delayed sentencing (often including reduced or dismissed charges), integration of military culture into court proceedings, a rehabilitative vs adversarial approach to decreasing risk of future criminal behavior, mandated treatment and supervision during participation, and use of veteran mentors to provide peer support.^32-35 Eligibility requirements vary; however, many restrict participation to veterans with honorable discharge types and charges for nonviolent offenses.^32,33,35-37

VTCs connect veterans within the criminal-legal system to needed behavioral health, community, and social services.^31-33,37 VTC participants are commonly connected to case management, behavioral health care, therapeutic journaling programs, and vocational rehabilitation.^38,39 Accordingly, the most common difficulties faced by veterans participating in these courts include substance use, mental health, family issues, anger management and/or aggressive behavior, and homelessness.^36,39 There is limited research on the effectiveness of VTCs. Evidence on their overall effectiveness is largely mixed, though some studies suggest VTC graduates tend to have lower recidivism rates than offenders more broadly or persons who terminate VTC programs prior to completion.^40,41 Other studies suggest that VTC participants are more likely to have jail sanctions, new arrests, and new incarcerations relative to nontreatment court participants.⁴² Notably, experimental designs (considered the gold standard in assessing effectiveness) to date have not been applied to evaluate the effectiveness of VTCs; as such, the effectiveness of these programs remains an area in need of continued empirical investigation.

Like all problem-solving courts, VTCs occasionally struggle to connect participating defendants with appropriate care, particularly when encountering structural barriers (eg, insurance, transportation) and/or complex behavioral health needs (eg, personality disorders).^34,43 As suicide rates among veterans experiencing criminal-legal involvement surge (about 150 per 100,000 in 2021, a 10% increase from 2020 to 2021 compared to about 40 per 100,000 and a 1.8% increase among other veterans), efficiency of adequate care coordination is vital.⁴⁴ Many VTCs rely on VTC-VA partnerships and collaborations to navigate these difficulties and facilitate connection of participating veterans to needed services.^32-34,45 For example, within the VHA, Veterans Justice Outreach (VJO) and Health Care for Re-Entry Veterans (HCRV) specialists assist and bridge the gap between the criminal-legal system (including, but not limited to VTCs) and VA services by engaging veterans involved in the criminal-legal system and connecting them to needed VA-based services (Table 2). Generally, VJO specialists support veterans involved with the front end of the criminallegal system (eg, arrest, pretrial incarceration, or participation in VTCs), while HCRV specialists tend to support veterans transitioning back into the community after a period of incarceration.^46,47 Specific to VTCs, VJO specialists typically serve as liaisons between the courts and VA, coordinating VA services for defendants to fulfill their terms of VTC participation.⁴⁶

The historical exclusion of veterans with OTH discharge characterizations from VA-based services has restricted many from accessing VTC programs.³² Of 17 VTC programs active in Pennsylvania in 2014, only 5 accepted veterans with OTstayH discharges, and 3 required application to and eligibility for VA benefits.³³ Similarly, in national surveys of VTC programs, about 1 in 3 report excluding veterans deemed ineligible for VA services.^35,36 When veterans with OTH discharges have accessed VTC programs, they have historically relied on non-VA, community-based programming to fulfill treatment mandates, which may be less suited to addressing the unique needs of veterans.⁴⁸

Veterans who utilize VTCs receive several benefits, namely peer support and mentorship, acceptance into a veteran-centric space, and connection with specially trained staff capable of supporting the veteran through applications for a range of VA benefits (eg, service connection, housing support).^31-33,37 Given the disparate prevalence of OTH discharge characterizations among service members from racial, sexual, and gender minorities and among service members with mental health disorders, exclusion of veterans with OTH discharges from VTCs solely based on the type of discharge likely contributes to structural inequities among these already underserved groups by restricting access to these potential benefits. Such structural inequity stands in direct conflict with VTC best practice standards, which admonish programs to adjust eligibility requirements to facilitate access to treatment court programs for historically marginalized groups.⁴⁹

ELIGIBILITY EXPANSIONS

Given the overrepresentation of veterans with OTH discharge characterizations within the criminal-legal system and historical barriers of these veterans to access needed mental and behavioral health care, expansions in VA eligibility policies could have immense implications for VTCs. First, these expansions could mitigate common barriers to connecting VTC-participating veterans with OTH discharges with needed behavioral health care by allowing these veterans to access established, VA-based services and programming. Expansion may also allow VTCs to serve as a key intercept point for identifying and engaging veterans with OTH discharges who may be unaware of their eligibility for VA-based behavioral health care.

Access to VA health care services is a major resource for VTC participants and a common requirement.³² Eligibility expansion should ease access barriers veterans with OTH discharges commonly face. By providing a potential source of treatment, expansions may also support OTH eligibility practices within VTCs, particularly practices that require participants to be eligible for VA health care.^33,35,36 Some VTCs may continue to determine eligibility on the basis of discharge status and remain inaccessible to veterans with OTH discharge characterizations without program-level policy changes.^32,36,37

Communicating changes in eligibility policies relevant to veterans with OTH discharges may be a challenge, because many of these individuals have no established channels of communication with the VA. Because veterans with OTH discharges are at increased risk for legal system involvement, VTCs may serve as a unique point of contact to help facilitate communication.¹⁸ For example, upon referral to a VTC, veterans with OTH discharges can be identified, VA health care eligibility can be verified, and veterans can connect to VA staff to facilitate enrollment in VA services and care.

VJO specialists are in a favorable position to serve a critical role in utilizing VTCs as a potential intercept point for engaging veterans with OTH discharge characterizations. As outlined in the STRONG Veterans Act of 2022, VJOs are mandated to “spread awareness and understanding of veteran eligibility for the [VJO] Program, including the eligibility of veterans who were discharged from service in the Armed Forces under conditions other than honorable.”50 The Act further requires VJOs to be annually trained in communicating eligibility changes as they arise. Accordingly, VJOs receive ongoing training in a wide variety of critical outreach topics, including changes in eligibility; while VJOs cannot make eligibility determinations, they are tasked with enrolling all veterans involved in the criminal-legal system with whom they interact into VHA services, whether through typical or special eligibility criteria (M. Stimmel, PhD, National Training Director for Veteran Justice Programming, oral communication, July 14, 2023). VJOs therefore routinely serve in this capacity of facilitating VA enrollment of veterans with OTH discharge characterizations.

Recommendations to Veteran-Servicing Judicial Programs

Considering these potential implications, professionals routinely interacting with veterans involved in the criminal-legal system should become familiarized with recent changes in VA eligibility policies. Such familiarization would support the identification of veterans previously considered ineligible for care; provision of education to these veterans regarding their new eligibility; and referral to appropriate VA-based behavioral health care options. Although conceptually simple, executing such an educational campaign may prove logistically difficult. Given their historical exclusion from VA services, veterans with OTH discharge characterizations are unlikely to seek VA-based services in times of need, instead relying on a broad swath of civilian community-based organizations and resources. Usual approaches to advertising VHA health care policy changes (eg, by notifying VA employees and/or departments providing corresponding services or by circulating information to veteran-focused mailing lists and organizations) likely would prove insufficient. Educational campaigns to disseminate information about recent OTH eligibility changes should instead consider partnering with traditionally civilian, communitybased organizations and institutions, such as state bar associations, legal aid networks, case management services, nonveteran treatment court programs (eg, drug courts, or domestic violence courts), or probation/ parole programs. Because national surveys suggest generally low military cultural competence among civilian populations, providing concurrent support in developing foundational veteran cultural competencies (eg, how to phrase questions about military service history, or understanding discharge characterizations) may be necessary to ensure effective identification and engagement of veteran clients.⁴⁸

Programs that serve veterans with criminal-legal involvement should also consider potential relevance of recent OTH eligibility changes to program operations. VTC program staff and key partners (eg, judges, case managers, district attorneys, or defense attorneys), should revisit policies and procedures surrounding the engagement of veterans with OTH discharges within VTC programs and strategies for connecting these veterans with needed services. VTC programs that have historically excluded veterans with OTH discharges due to associated difficulties in locating and connecting with needed services should consider expanding eligibility policies considering recent shifts in VA behavioral health care eligibility.^33,35,36 Within the VHA, VJO specialists can play a critical role in supporting these VTC eligibility and cultural shifts. Some evidence suggests a large proportion of VTC referrals are facilitated by VJO specialists and that many such referrals are identified when veterans involved with the criminal-legal system seek VA support and/or services.³³ Given the historical exclusion of veterans with OTH discharges from VA care, strategies used by VJO specialists to identify, connect, and engage veterans with OTH discharges with VTCs and other services may be beneficial.

Even with knowledge of VA eligibility changes and considerations of these changes on local operations, many forensic settings and programs struggle to identify veterans. These difficulties are likely amplified among veterans with OTH discharge characterizations, who may be hesitant to self-disclose their military service history due to fear of stigma and/or views of OTH discharge characterizations as undeserving of the veteran title.¹² The VA offers 2 tools to aid in identification of veterans for these settings: the Veterans Re-Entry Search Service (VRSS) and Status Query and Response Exchange System (SQUARES). For VRSS, correctional facilities, courts, and other criminal justice entities upload a simple spreadsheet that contains basic identifying information of inmates or defendants in their system. VRSS returns information about which inmates or defendants have a history of military service and alerts VA Veterans Justice Programs staff so they can conduct outreach. A pilot study conducted by the California Department of Corrections and Rehabilitation found that 2.7% of its inmate population self-identified as veterans, while VRSS identified 7.7% of inmates with a history of military service. This difference represented about 5000 previously unidentified veterans.⁵¹ Similarly, community entities that partner with the VA, such as law enforcement or homeless service programs, can be approved to become a SQUARES user and submit identifying information of individuals with whom they interact directly into the SQUARES search engine. SQUARES then directly returns information about the individual’s veteran status and eligibility for VA health care and homeless programs.

Other Eligibility Limitations

VTCs and other professionals looking to refer veterans with OTH discharge characterizations to VA-based behavioral health care should be aware of potential limitations in eligibility and access. Specifically, although veterans with OTH discharges are now broadly eligible for VA-based behavioral health care and homeless programs, they remain ineligible for other forms of health care, including primary care and nonbehavioral specialty care.1 Research has found a strong comorbidity between behavioral and nonbehavioral health concerns, particularly within historically marginalized demographic groups.^52-55 Because these historically marginalized groups are often overrepresented among persons with criminal-legal involvement, veterans with OTH discharges, and VTC participants, such comorbidities require consideration by services or programming designed to support veterans with criminal-legal involvement.^12,56-58 Connection with VA-based health care will therefore continue to fall short of addressing all health care needs of veterans with OTH discharges and effective case management will require considerable treatment coordination between VA behavioral health care practitioners (HCPs) and community-based HCPs (eg, primary care professionals or nonbehavioral HCPs).

Implications for VA Mental Health Care

Recent eligibility expansions will also have inevitable consequences for VA mental health care systems. For many years, these systems have been overburdened by high caseloads and clinician burnout.^59,60 Given the generally elevated rates of mental health and substance use concerns among veterans with OTH discharge characterizations, expansions hold the potential to further burden caseloads with clinically complex, high-risk, high-need clients. Nevertheless, these expansions are also structured in a way that forces existing systems to absorb the responsibilities of providing necessary care to these veterans. To mitigate detrimental effects of eligibility expansions on the broader VA mental health system, clinicians should be explicitly trained in identifying veterans with OTH discharge characterizations and the implications of discharge status on broader health care eligibility. Treatment of veterans with OTH discharges may also benefit from close coordination between mental health professionals and behavioral health care coordinators to ensure appropriate coordination of care between VA- and non–VA-based HCPs.

CONCLUSIONS

Recent changes to VA eligibility policies now allow comprehensive mental and behavioral health care services to be provided to veterans with OTH discharges.1 Compared to routinely discharged veterans, veterans with OTH discharges are more likely to be persons of color, sexual or gender minorities, and experiencing mental health-related difficulties. Given the disproportionate mental health burden often faced by veterans with OTH discharges and relative overrepresentation of these veterans in judicial and correctional systems, these changes have considerable implications for programs and services designed to support veterans with criminallegal involvement. Professionals within these systems, particularly VTC programs, are therefore encouraged to familiarize themselves with recent changes in VA eligibility policies and to revisit strategies, policies, and procedures surrounding the engagement and enrollment of veterans with OTH discharge characterizations. Doing so may ensure veterans with OTH discharges are effectively connected to needed behavioral health care services.

In April 2022, the US Department of Veterans Affairs (VA) revised its behavioral health care eligibility policies to provide comprehensive mental and behavioral health care to former service members who received an Other Than Honorable (OTH) discharge characterization upon separation from military service.¹ This policy shift represents a marked expansion in eligibility practices (Table 1 includes amended eligibility criteria).

Since June 2017, eligibility policies allowed veterans with OTH discharges to receive “emergent mental health services” needed to stabilize acute mental health crises related to military service (eg, acute escalations in suicide risk).^2,3 Previously, veterans with OTH discharges were largely ineligible for VA-based health care; these individuals were only able to access Veterans Health Administration (VHA) mental and behavioral health care through limited channels of eligibility (eg, for treatment of military sexual trauma or psychosis or other mental illness within 2 years of discharge).^4,5 The impetus for expansions in eligibility stemmed from VA efforts to reduce the suicide rate among veterans.^6-8 Implications of such expansion extend beyond suicide prevention efforts, with notable promised effects on the care of veterans with criminal-legal involvement. This article highlights potential effects of recent eligibility expansions on veterans with criminal-legal involvement and makes specific recommendations for agencies and organizations serving these veterans.

OTHER THAN HONORABLE DISCHARGE

The US Department of Defense delineates 6 discharge characterizations provided to service members upon separation from military service: honorable, general under honorable conditions, OTH, bad conduct, dishonorable, and uncharacterized. Honorable discharge characterizations are considered to reflect general concordance between service member behavior and military standards; general discharge characterizations reflect some disparity between the service member’s behavior and military standards; OTH, bad conduct, and dishonorable discharge characterizations reflect serious disparities between the service member’s behavior and military standards; and uncharacterized discharge characterizations are given when other discharge characterizations are deemed inappropriate.^9,10 OTH discharge characterizations are typically issued under instances of misconduct, fraudulent entry, security reasons, or in lieu of trial by court martial.^9,10

Recent research suggests that about 85% of service members receive an honorable discharge characterization upon separation from military service, 8% receive general, 6% receive OTH, and 1% receive bad conduct or dishonorable discharges.11 In 2017, the VA estimated there were > 500,000 prior service members with OTH discharge characterizations, which has grown over time (1.9% during the Korean Conflict, 2.5% during the Vietnam War Era, 3.9% during the Cold War, 4.8% in the Persian Gulf War, and 5.8% in the post-9/11 era).^7,11

The OTH discharge characterization is 1 of 3 less than honorable discharges informally referred to as bad papers (ie, OTH, bad conduct, or dishonorable). Former service members receiving these discharge characterizations face significant social stigma and structural discrimination upon military discharge, including significant hurdles to employment and educational pursuits as well as notable social alienation.¹² Due to their discharge characterization, some are viewed as less deserving of the veteran title, and until recently, many did not qualify for the complex legal definition of veteran as established by the Congress.^11,13 Veterans with OTH discharge characterizations have also historically been excluded from services (eg, VHA care),³ benefits (eg, disability compensation),¹⁴ and protections (eg, Uniformed Services Employment and Reemployment Rights Act)¹⁵ offered to veterans with honorable or general discharge characterizations. However, eligibility policies have gradually expanded, providing veterans with OTH discharges with access to VHA-based mental and behavioral health services and VA supportive housing assistance.^1,3,16

Perhaps due to their historical exclusion from VA services, there is limited research available on the behavioral health and associated needs of veterans with OTH discharges. Some scholars argue that historical exclusions have exacerbated underlying difficulties faced by this population, thereby contributing to stark health and social disparities across discharge types.^14,15,17 Studies with large veteran samples, for example, reflect notable demographic and behavioral health differences across discharge types. Compared to routinely discharged veterans, veterans with OTH discharges are significantly more likely to be younger, have lower income, use substances, have a history of criminal-legal involvement, and have mental and physical health difficulties.^18,19

Substantial evidence also suggests a historical racial bias, with service members of color being disproportionately more likely to receive an OTH discharge.¹² Similarly, across all branches of military service, Black service members are significantly more likely to face general or special court martial in military justice proceedings when compared with White service members.²⁰ Service members from gender and sexual minorities are also disproportionately impacted by the OTH designation. Historically, many have been discharged with bad papers due to discriminatory policies, such as Don’t Ask Don’t Tell, which discriminated on the basis of sexual orientation between December 1993 and September 2011, and Directive-type Memorandum-19-004, which banned transgender persons from military service between April 2019 and January 2021.^21,22

There is also significant mental health bias in the provision of OTH discharges, such that OTH characterizations are disproportionately represented among individuals with mental health disorders.^18-20 Veterans discharged from military service due to behavioral misconduct are significantly more likely to meet diagnostic criteria for various behavioral health conditions and to experience homelessness, criminal-legal involvement, and suicidal ideation and behavior compared with routinely-discharged veterans.^23-28

Consistent with their comparatively higher rates of criminal-legal involvement relative to routinely discharged veterans, veterans with OTH discharges are disproportionately represented in criminal justice settings. While veterans with OTH discharges represent only 6% of discharging service members and 2.5% of community-based veterans, they represent 10% of incarcerated veterans.^11,18,23,29 Preliminary research suggests veterans with OTH discharges may be at higher risk for lifetime incarceration, though the association between OTH discharge and frequency of lifetime arrests remains unclear.^18,30

VETERANS TREATMENT COURTS

Given the overrepresentation of veterans with OTH discharges in criminal-legal settings, consideration for this subset of the veteran population and its unique needs is commonplace among problem-solving courts that service veterans. First conceptualized in 2004, Veterans Treatment Courts (VTCs) are specialized problem-solving courts that divert veterans away from traditional judicial court and penal systems and into community-based supervision and treatment (most commonly behavioral health services).^31-34 Although each VTC program is unique in structure, policies, and procedures, most VTCs can be characterized by certain key elements, including voluntary participation, plea requirements, delayed sentencing (often including reduced or dismissed charges), integration of military culture into court proceedings, a rehabilitative vs adversarial approach to decreasing risk of future criminal behavior, mandated treatment and supervision during participation, and use of veteran mentors to provide peer support.^32-35 Eligibility requirements vary; however, many restrict participation to veterans with honorable discharge types and charges for nonviolent offenses.^32,33,35-37

VTCs connect veterans within the criminal-legal system to needed behavioral health, community, and social services.^31-33,37 VTC participants are commonly connected to case management, behavioral health care, therapeutic journaling programs, and vocational rehabilitation.^38,39 Accordingly, the most common difficulties faced by veterans participating in these courts include substance use, mental health, family issues, anger management and/or aggressive behavior, and homelessness.^36,39 There is limited research on the effectiveness of VTCs. Evidence on their overall effectiveness is largely mixed, though some studies suggest VTC graduates tend to have lower recidivism rates than offenders more broadly or persons who terminate VTC programs prior to completion.^40,41 Other studies suggest that VTC participants are more likely to have jail sanctions, new arrests, and new incarcerations relative to nontreatment court participants.⁴² Notably, experimental designs (considered the gold standard in assessing effectiveness) to date have not been applied to evaluate the effectiveness of VTCs; as such, the effectiveness of these programs remains an area in need of continued empirical investigation.

Like all problem-solving courts, VTCs occasionally struggle to connect participating defendants with appropriate care, particularly when encountering structural barriers (eg, insurance, transportation) and/or complex behavioral health needs (eg, personality disorders).^34,43 As suicide rates among veterans experiencing criminal-legal involvement surge (about 150 per 100,000 in 2021, a 10% increase from 2020 to 2021 compared to about 40 per 100,000 and a 1.8% increase among other veterans), efficiency of adequate care coordination is vital.⁴⁴ Many VTCs rely on VTC-VA partnerships and collaborations to navigate these difficulties and facilitate connection of participating veterans to needed services.^32-34,45 For example, within the VHA, Veterans Justice Outreach (VJO) and Health Care for Re-Entry Veterans (HCRV) specialists assist and bridge the gap between the criminal-legal system (including, but not limited to VTCs) and VA services by engaging veterans involved in the criminal-legal system and connecting them to needed VA-based services (Table 2). Generally, VJO specialists support veterans involved with the front end of the criminallegal system (eg, arrest, pretrial incarceration, or participation in VTCs), while HCRV specialists tend to support veterans transitioning back into the community after a period of incarceration.^46,47 Specific to VTCs, VJO specialists typically serve as liaisons between the courts and VA, coordinating VA services for defendants to fulfill their terms of VTC participation.⁴⁶

The historical exclusion of veterans with OTH discharge characterizations from VA-based services has restricted many from accessing VTC programs.³² Of 17 VTC programs active in Pennsylvania in 2014, only 5 accepted veterans with OTstayH discharges, and 3 required application to and eligibility for VA benefits.³³ Similarly, in national surveys of VTC programs, about 1 in 3 report excluding veterans deemed ineligible for VA services.^35,36 When veterans with OTH discharges have accessed VTC programs, they have historically relied on non-VA, community-based programming to fulfill treatment mandates, which may be less suited to addressing the unique needs of veterans.⁴⁸

Veterans who utilize VTCs receive several benefits, namely peer support and mentorship, acceptance into a veteran-centric space, and connection with specially trained staff capable of supporting the veteran through applications for a range of VA benefits (eg, service connection, housing support).^31-33,37 Given the disparate prevalence of OTH discharge characterizations among service members from racial, sexual, and gender minorities and among service members with mental health disorders, exclusion of veterans with OTH discharges from VTCs solely based on the type of discharge likely contributes to structural inequities among these already underserved groups by restricting access to these potential benefits. Such structural inequity stands in direct conflict with VTC best practice standards, which admonish programs to adjust eligibility requirements to facilitate access to treatment court programs for historically marginalized groups.⁴⁹

ELIGIBILITY EXPANSIONS

Given the overrepresentation of veterans with OTH discharge characterizations within the criminal-legal system and historical barriers of these veterans to access needed mental and behavioral health care, expansions in VA eligibility policies could have immense implications for VTCs. First, these expansions could mitigate common barriers to connecting VTC-participating veterans with OTH discharges with needed behavioral health care by allowing these veterans to access established, VA-based services and programming. Expansion may also allow VTCs to serve as a key intercept point for identifying and engaging veterans with OTH discharges who may be unaware of their eligibility for VA-based behavioral health care.

Access to VA health care services is a major resource for VTC participants and a common requirement.³² Eligibility expansion should ease access barriers veterans with OTH discharges commonly face. By providing a potential source of treatment, expansions may also support OTH eligibility practices within VTCs, particularly practices that require participants to be eligible for VA health care.^33,35,36 Some VTCs may continue to determine eligibility on the basis of discharge status and remain inaccessible to veterans with OTH discharge characterizations without program-level policy changes.^32,36,37

Communicating changes in eligibility policies relevant to veterans with OTH discharges may be a challenge, because many of these individuals have no established channels of communication with the VA. Because veterans with OTH discharges are at increased risk for legal system involvement, VTCs may serve as a unique point of contact to help facilitate communication.¹⁸ For example, upon referral to a VTC, veterans with OTH discharges can be identified, VA health care eligibility can be verified, and veterans can connect to VA staff to facilitate enrollment in VA services and care.

VJO specialists are in a favorable position to serve a critical role in utilizing VTCs as a potential intercept point for engaging veterans with OTH discharge characterizations. As outlined in the STRONG Veterans Act of 2022, VJOs are mandated to “spread awareness and understanding of veteran eligibility for the [VJO] Program, including the eligibility of veterans who were discharged from service in the Armed Forces under conditions other than honorable.”50 The Act further requires VJOs to be annually trained in communicating eligibility changes as they arise. Accordingly, VJOs receive ongoing training in a wide variety of critical outreach topics, including changes in eligibility; while VJOs cannot make eligibility determinations, they are tasked with enrolling all veterans involved in the criminal-legal system with whom they interact into VHA services, whether through typical or special eligibility criteria (M. Stimmel, PhD, National Training Director for Veteran Justice Programming, oral communication, July 14, 2023). VJOs therefore routinely serve in this capacity of facilitating VA enrollment of veterans with OTH discharge characterizations.

Recommendations to Veteran-Servicing Judicial Programs

Considering these potential implications, professionals routinely interacting with veterans involved in the criminal-legal system should become familiarized with recent changes in VA eligibility policies. Such familiarization would support the identification of veterans previously considered ineligible for care; provision of education to these veterans regarding their new eligibility; and referral to appropriate VA-based behavioral health care options. Although conceptually simple, executing such an educational campaign may prove logistically difficult. Given their historical exclusion from VA services, veterans with OTH discharge characterizations are unlikely to seek VA-based services in times of need, instead relying on a broad swath of civilian community-based organizations and resources. Usual approaches to advertising VHA health care policy changes (eg, by notifying VA employees and/or departments providing corresponding services or by circulating information to veteran-focused mailing lists and organizations) likely would prove insufficient. Educational campaigns to disseminate information about recent OTH eligibility changes should instead consider partnering with traditionally civilian, communitybased organizations and institutions, such as state bar associations, legal aid networks, case management services, nonveteran treatment court programs (eg, drug courts, or domestic violence courts), or probation/ parole programs. Because national surveys suggest generally low military cultural competence among civilian populations, providing concurrent support in developing foundational veteran cultural competencies (eg, how to phrase questions about military service history, or understanding discharge characterizations) may be necessary to ensure effective identification and engagement of veteran clients.⁴⁸

Programs that serve veterans with criminal-legal involvement should also consider potential relevance of recent OTH eligibility changes to program operations. VTC program staff and key partners (eg, judges, case managers, district attorneys, or defense attorneys), should revisit policies and procedures surrounding the engagement of veterans with OTH discharges within VTC programs and strategies for connecting these veterans with needed services. VTC programs that have historically excluded veterans with OTH discharges due to associated difficulties in locating and connecting with needed services should consider expanding eligibility policies considering recent shifts in VA behavioral health care eligibility.^33,35,36 Within the VHA, VJO specialists can play a critical role in supporting these VTC eligibility and cultural shifts. Some evidence suggests a large proportion of VTC referrals are facilitated by VJO specialists and that many such referrals are identified when veterans involved with the criminal-legal system seek VA support and/or services.³³ Given the historical exclusion of veterans with OTH discharges from VA care, strategies used by VJO specialists to identify, connect, and engage veterans with OTH discharges with VTCs and other services may be beneficial.

Even with knowledge of VA eligibility changes and considerations of these changes on local operations, many forensic settings and programs struggle to identify veterans. These difficulties are likely amplified among veterans with OTH discharge characterizations, who may be hesitant to self-disclose their military service history due to fear of stigma and/or views of OTH discharge characterizations as undeserving of the veteran title.¹² The VA offers 2 tools to aid in identification of veterans for these settings: the Veterans Re-Entry Search Service (VRSS) and Status Query and Response Exchange System (SQUARES). For VRSS, correctional facilities, courts, and other criminal justice entities upload a simple spreadsheet that contains basic identifying information of inmates or defendants in their system. VRSS returns information about which inmates or defendants have a history of military service and alerts VA Veterans Justice Programs staff so they can conduct outreach. A pilot study conducted by the California Department of Corrections and Rehabilitation found that 2.7% of its inmate population self-identified as veterans, while VRSS identified 7.7% of inmates with a history of military service. This difference represented about 5000 previously unidentified veterans.⁵¹ Similarly, community entities that partner with the VA, such as law enforcement or homeless service programs, can be approved to become a SQUARES user and submit identifying information of individuals with whom they interact directly into the SQUARES search engine. SQUARES then directly returns information about the individual’s veteran status and eligibility for VA health care and homeless programs.

Other Eligibility Limitations

VTCs and other professionals looking to refer veterans with OTH discharge characterizations to VA-based behavioral health care should be aware of potential limitations in eligibility and access. Specifically, although veterans with OTH discharges are now broadly eligible for VA-based behavioral health care and homeless programs, they remain ineligible for other forms of health care, including primary care and nonbehavioral specialty care.1 Research has found a strong comorbidity between behavioral and nonbehavioral health concerns, particularly within historically marginalized demographic groups.^52-55 Because these historically marginalized groups are often overrepresented among persons with criminal-legal involvement, veterans with OTH discharges, and VTC participants, such comorbidities require consideration by services or programming designed to support veterans with criminal-legal involvement.^12,56-58 Connection with VA-based health care will therefore continue to fall short of addressing all health care needs of veterans with OTH discharges and effective case management will require considerable treatment coordination between VA behavioral health care practitioners (HCPs) and community-based HCPs (eg, primary care professionals or nonbehavioral HCPs).

Implications for VA Mental Health Care

Recent eligibility expansions will also have inevitable consequences for VA mental health care systems. For many years, these systems have been overburdened by high caseloads and clinician burnout.^59,60 Given the generally elevated rates of mental health and substance use concerns among veterans with OTH discharge characterizations, expansions hold the potential to further burden caseloads with clinically complex, high-risk, high-need clients. Nevertheless, these expansions are also structured in a way that forces existing systems to absorb the responsibilities of providing necessary care to these veterans. To mitigate detrimental effects of eligibility expansions on the broader VA mental health system, clinicians should be explicitly trained in identifying veterans with OTH discharge characterizations and the implications of discharge status on broader health care eligibility. Treatment of veterans with OTH discharges may also benefit from close coordination between mental health professionals and behavioral health care coordinators to ensure appropriate coordination of care between VA- and non–VA-based HCPs.

CONCLUSIONS

Recent changes to VA eligibility policies now allow comprehensive mental and behavioral health care services to be provided to veterans with OTH discharges.1 Compared to routinely discharged veterans, veterans with OTH discharges are more likely to be persons of color, sexual or gender minorities, and experiencing mental health-related difficulties. Given the disproportionate mental health burden often faced by veterans with OTH discharges and relative overrepresentation of these veterans in judicial and correctional systems, these changes have considerable implications for programs and services designed to support veterans with criminallegal involvement. Professionals within these systems, particularly VTC programs, are therefore encouraged to familiarize themselves with recent changes in VA eligibility policies and to revisit strategies, policies, and procedures surrounding the engagement and enrollment of veterans with OTH discharge characterizations. Doing so may ensure veterans with OTH discharges are effectively connected to needed behavioral health care services.