ON TIME BUSINESS MANAGEMENT A+ WORK, ON TIME, NO PLAGARIZING; ON TIME

Psychological Services Role of Homelessness and Incarceration in Time-to-Decision on Veterans Affairs Disability Benefits for a Mental Health Condition Hind A. Beydoun, Mayumi Gianoli, Andrew Meisler, and Jack Tsai

Online First Publication, June 30, 2025. https://dx.doi.org/10.1037/ser0000971

CITATION

Beydoun, H. A., Gianoli, M., Meisler, A., & Tsai, J. (2025). Role of homelessness and incarceration in time- to-decision on veterans affairs disability benefits for a mental health condition. Psychological Services. Advance online publication. https://dx.doi.org/10.1037/ser0000971

Role of Homelessness and Incarceration in Time-to-Decision on Veterans Affairs Disability Benefits for a Mental Health Condition

Hind A. Beydoun1, 2, Mayumi Gianoli3, Andrew Meisler3, and Jack Tsai1, 2, 3 1 VA National Center on Homelessness Among Veterans, U.S. Department of Veterans Affairs, Washington, DC, United States

2 Department of Management, Policy, and Community Health, School of Public Health, University of Texas Health Science Center at Houston 3 Department of Psychiatry, Yale School of Medicine

Homelessness and incarceration are major psychosocial concerns in the United States that may be mitigated by access to disability benefits. Data from the Veterans Benefits Administration (VBA) provides a unique opportunity to study how the process of obtaining disability benefits and time-to-decision might be affected by homelessness and/or incarceration among veterans in the Veterans Health Administration (VHA). The purpose of this study was to examine how homelessness and/or incarceration status among VHA-enrolled U.S. veterans might affect outcomes of VBA disability benefits claims for a mental health condition, including time-to-decision and rating outcomes. Using a retrospective cohort design, linked VHA/VBAdata from 2022 to 2023 on 327,047 U.S. veterans (including 7,202 homeless veterans and 280 incarcerated veterans) were analyzed with multivariable Cox regressions. Overall, 85.7% of veterans received award of a Department of Veterans Affairs (VA) service-connected disability for a mental health condition. Although VA service-connected disability for a mental health condition was awarded less frequently for homeless and incarcerated veterans, the time-to-decision for homeless veterans—with or without incarceration—was twice as fast as nonhomeless and nonincarcerated veterans, after adjustment for demographic and clinical characteristics. In conclusion, although homeless and/or incarcerated veterans were less likely to receive VA service-connected disability for a mental health condition than other VA disability-seeking veterans, homeless veterans had faster decision times on VA disability applications. These findings inform ongoing efforts to ensure veterans with diverse backgrounds and needs receive the VBA benefits for which they qualify.

Public Significance Statement This study investigated the association of homelessness and incarceration status with outcomes of Veterans Benefits Administration disability benefits claims for mental health conditions among veterans enrolled in Veterans Health Administration. Service-connected disability awards for homeless and incarcerated veterans were less frequent, but the time-to-decision was faster than for other veterans, informing efforts by policymakers to ensure equitable distribution of disability benefits.

Keywords: benefits, disability, incarceration, homelessness, veteran

John R. McQuaid served as action editor. Hind A. Beydoun https://orcid.org/0000-0002-9704-328X The Veterans Health Administration and Veterans Benefits Administra-

tion databases are not public use, and access to these databases is limited to employees of the U.S. Department of Veterans Affairs (VA) with proper approvals. Accordingly, the study data cannot be made available to re- searchers due to data sharing agreement limitations. Analytic methods can be obtained upon request with approval from the VA National Center on Homelessness Among Veterans. A Quality Improvement project in the Veterans Health Administration Homeless Programs Office was undertaken in partnership with the Compensation and Pension service at the VA Connecticut Healthcare System. Institutional Review Board approval was waived given the nonresearch nature of this Quality Improvement project. The authors have no relevant financial or nonfinancial interests to disclose.

The authors declare that no funds, grants, or other support were received during the preparation of this article. Hind A. Beydoun played a lead role in data curation, formal analysis,

methodology, software, writing–original draft, and writing–review and editing and an equal role in conceptualization. Mayumi Gianoli played a lead role in conceptualization, investigation, and project administration and an equal role in data curation, methodology, supervision, validation, writing– original draft, and writing–review and editing. Andrew Meisler played an equal role in investigation, writing–original draft, and writing–review and editing. Jack Tsai played an equal role in investigation, methodology, re- sources, supervision, writing–original draft, and writing–review and editing. Correspondence concerning this article should be addressed to Jack Tsai,

VA National Center on Homelessness Among Veterans, U.S. Department of Veterans Affairs, Homeless Programs Office, 810 Vermont Avenue, NW Washington, DC 20420, United States. Email: [email protected]

Psychological Services

In the public domain ISSN: 1541-1559 https://doi.org/10.1037/ser0000971

1

Homelessness and incarceration are interrelated psychosocial con- cerns affecting military veterans in the United States (Cusack & Montgomery, 2017; Elbogen et al., 2023). Previous studies have estimated that veterans make up 8%–10% of those who are incar- cerated, with incarcerated veteransmaking up 5.3%of the population in federal prisons and 7.9% in state prisons (Bronson et al., 2015; Elbogen et al., 2012; Maruschak et al., 2021), despite representing approxi- mately 6% of the U.S. population ≥18 years of age (U.S. Census Bureau, 2021). Several factors have been found to place veterans at increased risk for incarceration including posttraumatic stress disorder (PTSD; Backhaus et al., 2016; Saxon et al., 2001) and substance use disorders (SUDs; Elbogen et al., 2012). Studies have also shown that, historically, veterans represent a disproportionately high percentage of the homeless population in the United States (Tsai &Rosenheck, 2015) In their meta-analysis of 31 studies, Tsai and Rosenheck found that the strongest risk factors for homelessness among veterans were substance use and mental health disorders, followed by low income and other income-related factors (Tsai & Rosenheck, 2015). Not surprisingly, various studies have found that incarceration and homelessness are often related such as a study that found that 27% of incarcerated veterans reported having experienced homelessness in the 12 months preceding their arrest (Elbogen et al., 2023). Clearly, there is a common set of characteristics that place veterans

at increased risk for both homelessness and incarceration. Taking these interconnected risk factors into consideration, it follows that U.S. veterans with disabling conditions related to military service (i.e., “service-related”), including mental health problems, would be at increased risk for homelessness and incarceration because of psychosocial impairment and its associated financial hardship. To address these risk factors, improving access to disability benefits and services from the Department of Veterans Affairs (VA)—including income, housing, and health care—is a critical area for intervention. For example, Murdoch and colleagues found that, 10 years after attaining VA service-connected disability benefits for PTSD, veterans were less likely to be indigent and/or homeless (Murdoch et al., 2011). In addition to financial benefits, some have argued that the disability examination process itself can serve as an entry point for veterans to learn about and access various VA services (Rosen et al., 2013). U.S. veterans who are homeless or incarcerated are important

priority populations for the VA (U.S. Department of Veterans Affairs, 2023). TheVA consists of three branches, namely, theVeteransHealth Administration (VHA), the Veterans Benefits Administration (VBA), and the National Cemetery Administration. While VHA operates a national network of VA medical centers and clinics to provide health care and social services to veterans, VBA administers the program for receiving and processing applications for service-connected disability benefits. Both VHA and VBA offer potentially important resources that can help address homelessness and criminal justice involvement among U.S. veterans (Edens et al., 2011; Fargo et al., 2017). For both incarcerated and homeless veterans, not having the disability benefits they may be entitled to places them at elevated risk for repeated cycles of adverse outcomes. The VHA operates a large national continuum of homeless service

programs (Tsai & Byrne, 2019; Tsai, Kasprow, & Rosenheck, 2013) which includes the two major programs to connect criminal justice- involved veterans to health care through the Veterans Justice Outreach (VJO) and Health Care for Re-Entry Veterans (HCRV) programs. While theVJO program focuses on diverting justice-involved veterans from incarceration to VA services, such as through veterans treatment

courts (Blue-Howells et al., 2013; Tsai et al., 2017), the HCRV program helps incarcerated veterans leaving correctional facilities reenter their communities and connect to needed VA care (Finlay et al., 2017; Tsai, Rosenheck, et al., 2013). Presumably, both VA’s homeless and criminal justice programs can help veterans not only connect toVA health care and social services but help them obtainVA disability benefits if they are eligible. However, to date, there has been no formal examination about the process and extent towhich homeless and justice-involved veterans are connected to VA disability benefits.

In this study, we linked nationwide data sets from VHA and VBA to identify factors related to the disability claims process amongU.S. veterans who were experiencing homelessness or incarceration. The primary aim of the study was to examine processes and outcomes of disability claims applications for a mental health condition among U.S. veterans who were homeless and/or incarcerated compared to U.S. veterans who were not, including (a) rate of claim completion; (b) time-to-decision regarding VA service connection; (c) outcome of disability decision (positive or negative); and (d) level of monetary disability benefit awarded as a result of this claim (i.e., the service- connected rating). The secondary aim was to examine the association of homelessness and/or incarceration with time-to-decision regarding service connection, controlling for demographic, clinical, mental health, and substance use characteristics. To our knowledge, this is the first study to explore the potential role that homelessness and incarceration play in VA’s disability claims process.

Method

Study Design, Setting, and Procedure

A Quality Improvement project in the VHA Homeless Programs Office was undertaken in partnership with the Compensation and Pension (C&P) service at the VA Connecticut Healthcare System. Institutional Review Board approval was waived given the non- research nature of this Quality Improvement project. A retrospective cohort design was applied, whereby VHA and VBA databases were linked together using veteran identifiers for the purpose of examining VA disability claims and outcomes among veterans experiencing homelessness and/or incarceration. In partnership with the C&P service, VBA data elements pertaining to U.S. veterans who applied for disability benefits for amental health condition betweenDecember 14, 2022, and December 12, 2023, were requested and obtained through VBA’s Office of Performance Analysis and Integrity. These data elements included social security number, date of request sub- mitted by the U.S. veteran, date that the examination request was sent by VBA (form 2507), type of examination (mental health or PTSD, initial or review), date of examination, diagnoses rendered, date that decisions were made, nature of decisions (VA service-connected disability and %), and veteran geographic region (Veterans Integrated Service Network). The main exposure variables (i.e., homelessness and/or incarceration at the time of disability claim submission) were defined based on flag indicators recorded within the VBA database. Homelessness and incarceration were coded based on records made at that time and do not necessarily indicate veterans were homeless or incarcerated the entire study period. Time-to-decision regarding disability claims was defined based on a binary disability claim status (“complete” coded as 1 vs. “pending” coded as 0) variable and a time- to-event variable calculated as the number of days to closing a dis- ability claim among veterans with a “complete” status or the number

2 BEYDOUN, GIANOLI, MEISLER, AND TSAI

of days to disability claim pending among veterans with a “pending” status. These VBA data were linked to VHA data stored in the

Corporate Data Warehouse, containing medical records data from U.S. veterans who sought VA health care at an index date between January 1, 2022, and December 31, 2022. Data elements from VHA included demographic characteristics (e.g., sex, age, race, marital status, combat service, military sexual trauma) and clinical characteristics, including prior VHA-based disability rating, SUDs diagnoses (i.e., tobacco, alcohol, and/or drugs [opioids, cannabis, stimulants, sedatives/hypnotics/anxiolytics, hallucinogens, in- halants, other psychoactive substances]), mental health disorder diagnoses (i.e., major depressive disorder, bipolar disorder, other mood disorder, PTSD, other stress disorders, generalized anxiety disorder, other anxiety disorders, schizophrenia, other psychotic disorder, and/or other mental health disorders) and the Charlson Comorbidity Index (CCI) score. The 17-item CCI captures the number of major chronic medical conditions based on International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) diagnostic codes and predicts cumulative 1-year mortality risk (Sundararajan et al., 2004). Inpatient and outpatient diagnoses used to define SUDs, mental health disorders, and the CCI score were obtained within 12 months after index dates that ranged between June 1, 2022, and December 31, 2022. ICD-10-CM diagnostic codes used to define SUDs and mental health disorders are described elsewhere (Beydoun et al., 2024).

Study Population

The VBA cohort requesting disability benefits was further restricted to those U.S. veterans who met the following eligibility criteria: (a) requested disability benefits throughout the VA between December 14, 2022, and December 12, 2023; (b) sought health care services at least once throughout the VA between January 1, 2022, and December 31, 2023; (c) had matching VBA/VHA records; and (d) had no missing data on demographic and clinical characteristics. The original VBA data set consisted of 437,786 records, and the original VHAdata set consisted of 5,756,894 records. The final linked VBA/VHA data set consisted of 327,047 records corresponding to unique U.S. veterans who have applied for and/or received VA disability benefits and met prespecified study eligibility criteria. Of note, demographic and clinical characteristics obtained from theVHA database were examined as covariates in regression models for the role of homelessness and incarceration in time-to-decision on VA disability benefits. As such, the time period of the VHA database used (January 1, 2022, to December 31, 2022) preceded that of the VBA database (December 14, 2022, to December 12, 2023).

Statistical Analysis

Descriptive, bivariate, and multivariable statistical analyses were performed using SAS Version 9.2 (SAS Institute, North Carolina). Summary statistics included measures of central tendency and dis- persion for continuous variables, as well as frequencies with per- centages for categorical variables. Bivariate analyses were performed using chi-square and independent samples t test, as appropriate. Multivariable Cox proportional hazards regression models were used to calculate hazard ratios (HR) with 95% confidence intervals (CI) for

predictors of time-to-decision regarding disability benefits. Two- tailed statistical tests were assessed at α = .05.

Results

Table 1 describes demographic, clinical, SUDs, and mental health disorders characteristics among veterans in the study sample by homelessness and incarceration status at the time of VBA request. Homeless veterans applying for VA disability benefits for a mental health condition were more likely to be female, younger, non-White, unmarried, to have either no or 0% prior disability rating based on VHA data, to have experienced military sexual trauma, and to be diagnosed with substance use and/or mental health disorders, but less likely to have served in a combat zone than their nonhomeless counterparts. Similarly, incarcerated veterans applying for VA dis- ability benefits for a mental health condition were more likely to be male, younger, White, unmarried, to have no or 0% prior disability rating based onVHAdata, to have experiencedmilitary sexual trauma, and to be diagnosed with substance use and/or mental health disorders, but less likely to have served in a combat zone than their non- incarcerated counterparts. Finally, the mean CCI score, reflecting 1-year mortality risk associated with comorbidities, was lower among the homeless versus the nonhomeless veterans and among the incarcerated versus the nonincarcerated veterans.

Table 2 presents a flowchart of veterans in the study sample applying for VBA disability benefits for a mental health condition by homelessness and incarceration status at the time of VBA request. Of 327,047 records of VBA claims, 66.7% were complete and 33.3% were pending. The average number of days from application to claim completion was 104.2 days. Overall, 85.7% (280,556 of 327,047 claims) resulted in an award of a VA service-connected disability. Overall, the claim status was more likely to be completed among homeless versus nonhomeless veterans (77.84% vs. 66.47%), and the average time-to-decision was significantly shorter among homeless versus nonhomeless veterans (72.26 days vs. 105.06 days), as well as among incarcerated versus nonincarcerated veterans (84.98 days vs. 104.24 days). The award of VA service connection was less frequent among homeless versus nonhomeless (83.70% vs. 85.83%) and among incarcerated versus nonincarcerated veterans (72.14% vs. 85.80%) based on VBA data. Table A1 presents the distribution of time-to-decision by geographical region, irrespective of homelessness and/or incarceration status. Using a subset of 283,100 with non- missing data on time-to-decision and geographical region, results suggested that there were statistically significant differences in time-to-decision according to region, with the lowest mean value for time-to-decision in the Northeast region and the highest mean value for time-to-decision in the Pacific region.

Table A2 presents percentage of VA service-connected dis- ability rating for a mental health condition based on VBA data by homelessness and incarceration status at the time of VBA request. The distribution of service-connected disability rating according to homelessness and/or incarceration indicated that, among veterans who received VBA benefits, the incarcerated group had the lowest rating, and those who were neither homeless nor incarcerated had the highest rating.

Table 3 presents multivariable Cox regression models for demo- graphic, clinical, and homelessness and/or incarceration characteristics as predictors of time-to-decision for disability benefits. Homeless veterans—without incarceration (HR= 1.93, 95%CI [1.88, 1.99]) and

TIME-TO-DECISION ON DISABILITY BENEFITS 3

with incarceration (HR = 2.19, 95% CI [1.46, 3.29])—were nearly twice as likely to have a completed disability claim status than their nonhomeless and nonincarcerated counterparts. Other key predictors of a completed disability claims status included age ≥80 years (vs. <50 years; HR= 1.35, 95%CI [1.31, 1.38]), age 75–79 years (vs.<50 years; HR = 1.08, 95% CI [1.07, 1.09]), and prior VA service- connected disability rating based on VHA data (50%–100% vs. none/ 0%: HR = 1.19, 95% CI [1.15, 1.22]; 10%–40% vs. none/0%: HR = 1.15, 95% CI [1.12, 1.18]).

Discussion

This large cohort study used linked VHA/VBA data to evaluate the relationship between homelessness and/or incarceration status and the

process of obtaining VA disability benefits among U.S. veterans. Veterans experiencing homelessness had their claims processed faster than those not experiencing homelessness, which may be due to VA’s national priority to prevent and end veteran homelessness. VHA has various homeless and criminal justice programs to help connect veterans with VA benefits and health care services. In addition, VBA has implemented priority processing for U.S. veterans experiencing homelessness or those “at risk” for homelessness. Veterans’ claims are identified as “priority” when veterans self-identify as being homeless or “at risk” for homelessness (Veterans Benefits Administration, 2016). Although the current data do not provide direct evidence of this, given these policy initiatives, it is plausible that both examiners and VBA raters accelerated the turnaround of claims for veterans identified as homeless. It is worth mentioning though that we are not

Table 1 Demographic, Clinical, Mental Health, and Substance Use Characteristics of the Study Sample by Current Homelessness and Incarceration Status (n = 327,047)

Characteristic

N (%) or M ± SD

Total

Currently homeless Currently incarcerated

Yes No Yes No

Overall 327,047 7,202 319,845 280 326,767 Sex p = .04 p < .0001 Male 292,743 (89.51) 6,395 (88.79) 286,348 (89.53) 274 (97.86) 292,469 (89.50) Female 34,304 (10.49) 807 (11.21) 33,497 (10.47) 6 (2.14) 34,298 (10.50)

Age (years) p < .0001 p < .0001 M ± SD 55.72 ± 16.46 49.20 ± 14.71 55.86 ± 16.46 48.33 ± 12.62 55.72 ± 16.46

p < .0001 p < .0001 <50 122,303 (37.40) 3,645 (50.61) 118,658 (37.10) 153 (54.64) 122,150 (37.38) 50–64 65,158 (19.92) 1,571 (21.81) 63,587 (19.88) 79 (28.21) 65,079 (19.92) 65–74 47,560 (14.54) 1,351 (18.76) 46,209 (14.45) 33 (11.79) 47,527 (14.54) 75–79 81,690 (24.98) 573 (7.96) 81,117 (25.36) 14 (5.00) 81,676 (25.00) 80+ 10,336 (3.16) 62 (0.86) 10,274 (3.21) 1 (0.36) 10,335 (3.16)

Race/ethnicity p < .0001 p = .0004 Non-Hispanic White 160,516 (49.08) 2,945 (40.89) 157,571 (49.26) 137 (48.93) 160,379 (49.08) Non-Hispanic Black 72,758 (22.25) 2,570 (35.68) 70,188 (21.94) 68 (24.29) 72,690 (22.25) Hispanic 30,754 (9.40) 377 (5.23) 30,377 (9.50) 8 (2.86) 30,746 (9.41) Mixed race/other 12,956 (3.96) 252 (3.50) 12,704 (3.97) 8 (2.86) 12,948 (3.96) Missing 50,063 (15.31) 1,058 (14.69) 49,005 (15.32) 59 (21.07) 50,004 (15.30)

Marital status p < .0001 p < .0001 Married 191,506 (58.56) 1,782 (24.74) 189,724 (59.32) 88 (31.43) 191,418 (58.58) Single/never married 53,807 (16.45) 2,327 (32.31) 51,480 (16.10) 79 (28.21) 53,728 (16.44) Separated/divorced 66,742 (20.41) 2,866 (39.79) 63,876 (19.97) 101 (36.07) 66,641 (20.39) Widowed 5,573 (1.70) 126 (1.75) 5,447 (1.70) 3 (1.07) 5,570 (1.70) Missing 9,419 (2.88) 101 (1.40) 9,318 (2.91) 9 (3.21) 9,410 (2.88)

Disability ratinga p < .0001 p < .0001 None/0% 7,698 (2.35) 464 (6.44) 7,234 (2.26) 19 (6.79) 7,679 (2.35) 10%–40% 43,246 (13.22) 1,216 (16.88) 42,030 (13.14) 77 (27.50) 43,169 (13.21) 50%–100% 276,103 (84.42) 5,522 (76.67) 270,581 (84.60) 184 (65.71) 275,919 (84.44)

Combat service p < .0001 p = .21 Yes 29,095 (8.90) 522 (7.25) 28,573 (8.93) 19 (6.79) 29,076 (8.90) No 297,952 (91.10) 6,680 (92.75) 291,272 (91.07) 261 (93.21) 297,691 (91.10)

Military sexual trauma p < .0001 p < .0001 Yes 21,642 (6.62) 1,058 (14.69) 20,584 (6.44) 36 (12.86) 21,606 (6.61) No 305,405 (93.38) 6,144 (85.31) 299,261 (93.56) 244 (87.14) 305,161 (93.39)

Charlson Comorbidity Index p < .0001 p < .0001 M ± SD 0.76 ± 1.46 0.66 ± 1.42 0.77 ± 1.46 0.41 ± 0.93 0.76 ± 1.46

Mental health disorders p < .0001 p < .0001 Yes 130,195 (39.81) 4,312 (59.87) 125,883 (39.36) 151 (53.93) 130,044 (39.80) No 196,852 (60.19) 2,890 (40.13) 193,962 (60.64) 129 (46.07) 196,723 (60.20)

Substance use disorders p < .0001 p < .0001 Yes 33,034 (10.10) 2,610 (36.24) 30,424 (9.51) 96 (34.29) 32,938 (10.08) No 294,013 (89.90) 4,592 (63.76) 289,421 (90.49) 184 (65.71) 293,829 (89.92)

a Disability rating as indicated in the Veterans Health Administration records.

4 BEYDOUN, GIANOLI, MEISLER, AND TSAI

able to determine how many homeless and/or incarcerated veterans who qualify for VA service-connected disability did not apply. There may be various barriers we could not study since we only had VBA data on veterans who had filed claims, and there are likely many veterans who had not filed claims. Among veterans who applied for VA disability benefits, homeless

and incarcerated veterans were less likely to be awarded a service- connected disability rating than their nonhomeless and non- incarcerated veteran counterparts, respectively. This is noteworthy because it has been assumed that barriers to VA service connection among these groups related mostly to access, whereas the data indicate that, even with completed claims, there may be other complicating factors that require further study (Chong et al., 2018; Lucksted et al., 2022; Pena et al., 2024; Stupart et al., 2008). For example, there may be examiner bias or misconceptions regarding the cause of impairments for homeless/incarcerated veterans (e.g., criminological in nature and not necessarily related to disability or their military service) that may influence decisions (Lucksted et al., 2022). Examiner bias has been studied in other areas of forensic psychology (Chong et al., 2018; Pena et al., 2024; Stupart et al., 2008) and warrants investigation in the VA disability examination process. It is possible that, for some veterans, factors related to their homelessness and incarceration may obscure other service-related conditions and impairments. Thus, identifying a causal nexus (required for service connection) betweenmilitary service and current diagnosis may be more difficult for these veterans. Further, appor- tioning symptoms and impairments to a VA service-related condition relative to other factors is difficult. Given the putative relationship between combat zone and traumatic event exposures, it is noteworthy that both homeless and incarcerated veterans were less likely than their counterparts to have served in a combat zone, potentially ex- plaining the lower VA-connected disability rating among the homeless and/or incarcerated veteran populations.

There were several limitations of this study. First, analyses were restricted to the 2022–2023 VBA/VHA linked databases and veterans who had both VBA and VHA records. It would be informative to expand the dates to examine the relationship between incarceration and/or homelessness status and timeline for and outcomes of disability benefits processing over a longer period. Analysis of data for an expanded time frame could provide evidence of the effectiveness of VA’s increase in services for homeless and incarcerated veterans in recent years, thereby reducing barriers toVBAbenefits. A greater time frame would also permit a more thorough analysis of those claims that were taking longer to process as some can take months or years due to repeat submissions and reviews for the same veteran. Second, the study relied on VBA-recorded information of homelessness and/or incarceration, rather than operational definitions developed by the VHA, as described elsewhere (Tsai et al., 2022). Furthermore, it is not clear from the VBA database if homeless veterans had received homeless services through the VA or if incarcerated veterans were seen by a Veteran Justice Program specialist. The linked VBA/VHA database is not sufficiently granular to allow the ascertainment of protocols for cancellation of appointments which may vary over time. Cancellation may be initiated by the examiner, the veteran, or both, whereas no-shows may or may not result in cancellations, and if cancellations were by the provider, reschedulingmay have occurred at a later time point. Also, the linked VBA/VHA database did not have information on whether veterans had an appointment with VHA or contract examiners. Over the past 8 years, many C&P disability examinations have been outsourced to contract examination (i.e., medical disability examination) companies (Disabled American Veterans letter to Congress). Thus, future studies should evaluate whether reliance on VHA examiners for the C&P disability exami- nation process can reduce barriers to accessing VBA services among homeless and incarcerated veterans. Third, mental health and SUD diagnoses are important factors related to homelessness, incarceration,

Table 2 Flowchart for Veteran Disability Benefits in the Study Sample by Current Homelessness and Incarceration Status (n = 327,047)

Study sample

N (%) or M ± SD

Total

Currently homeless Currently incarcerated

Yes No Yes No

Overalla 327,047 (100%) 7,202 (2.20%) 319,848 (97.80%) 280 (0.09%) 326,767 (99.91%) Appointment scheduled p < .0001 p < .0001 Yes 242,641 (74.19) 4,635 (64.36) 238,006 (74.41) 148 (52.86) 242,493 (74.21)

Type of exam—Initial p < .0001 p = .21 Yes 24,018 (7.34) 828 (11.50) 23,190 (7.25) 26 (9.29) 23,992 (7.34)

Type of exam—Review p < .0001 p = .13 Yes 25,907 (7.92) 822 (11.41) 25,085 (7.84) 29 (10.36) 25,878 (7.92)

Claim status p < .0001 p = .88 Completed 218,218 (66.72) 5,606 (77.84) 212,612 (66.47) 92 (32.86) 108,737 (33.28) Pending 108,829 (33.28) 1,596 (22.16) 107,233 (33.53) 188 (67.14) 218,030 (66.72)

Days to claim completionb p < .0001 p < .0001 M ± SD 104.22 ± 59.32 72.26 ± 52.64 105.06 ± 59.26 84.98 ± 57.17 104.24 ± 59.32 Range 1–357 2–345 1–357 4–252 1–357

Days claim pendingc p < .0001 p = .48 M ± SD 121.35 ± 73.15 102.24 ± 76.76 121.63 ± 73.06 127.43 ± 82.96 121.34 ± 73.14 Range 1–364 1–348 1–364 2–307 1–364

Decision made p < .0001 p < .0001 Service-connected 280,556 (85.78) 6,028 (83.70) 274,528 (85.83) 202 (72.14) 280,354 (85.80)

a For all veterans in the study sample, a claim has been submitted and established, and an examination has been requested. b Among veterans with closed claims. c Among veterans with claims pending.

TIME-TO-DECISION ON DISABILITY BENEFITS 5

as well as eligibility for VA disability benefits. However, we did not find they added significantly to our regression models as predictors of VA disability claims outcome, warranting further investigation. Future studies with more fine-grained clinical assessment data may improve modeling strategies for understanding the roles of home- lessness and incarceration in disability benefits processing for U.S. veterans. Also, future studies should collect data that compares time-to-decision on disability benefits among incarcerated veter- ans based on their enrollment in the Veteran Justice Program (VJO or HCRV). Finally, it is worth noting that study findings can only be generalized to veterans who applied for disability benefits and

that these veterans may differ in terms of their demographic and clinical characteristics—including combat service—from those who did not apply for disability benefits.

Conclusions

In conclusion, although negative disability decisions and rating outcomes were more commonly observed among homeless and/or incarcerated U.S. veterans, time-to-decision on disability benefits was shorter for homeless veterans. Further research is needed on how to better integrate VBA and VHA services that target homeless and incarcerated subpopulations of veterans to achieve equitable access to these services for veterans who qualify for benefits.

References

Backhaus, A., Gholizadeh, S., Godfrey, K. M., Pittman, J., & Afari, N. (2016). The many wounds of war: The association of service-related and clinical characteristics with problems with the law in Iraq and Afghanistan veterans. International Journal of Law and Psychiatry, 49(Pt. B), 205– 213. https://doi.org/10.1016/j.ijlp.2016.10.007

Beydoun, H. A., Szymkowiak, D., Kinney, R., Beydoun, M. A., Zonderman, A. B., & Tsai, J. (2024). Is the risk of Alzheimer’s disease and related dementias among U.S. veterans influenced by the intersectionality of housing status, HIV/AIDS, hepatitis C, and psychiatric disorders? The Journals of Gerontology: Series A, 79(8), Article glae153. https://doi.org/ 10.1093/gerona/glae153

Blue-Howells, J. H., Clark, S. C., van den Berk-Clark, C., & McGuire, J. F. (2013). The U.S. Department of Veterans Affairs Veterans Justice pro- grams and the sequential intercept model: Case examples in national dissemination of intervention for justice-involved veterans. Psychological Services, 10(1), 48–53. https://doi.org/10.1037/a0029652

Bronson, J. C. A., Noonan, M., & Berzofsky, M. (2015). Veterans in prison and jail, 2011–12. U.S. Department of Justice, Office of Justice Programs, Bureau of Justice Statistics. https://bjs.ojp.gov/content/pub/pdf/vpj11 12.pdf

Chong, L., Taylor, S., Haywood, M., Adelstein, B. A., & Shulruf, B. (2018). Examiner seniority and experience are associated with bias when scoring communication, but not examination, skills in objective structured clinical examinations in Australia. Journal of Educational Evaluation for Health Professions, 15, Article 17. https://doi.org/10.3352/jeehp.2018.15.17

Cusack, M., & Montgomery, A. E. (2017). Examining the bidirectional association between veteran homelessness and incarceration within the context of permanent supportive housing. Psychological Services, 14(2), 250–256. https://doi.org/10.1037/ser0000110

Edens, E. L., Kasprow,W., Tsai, J., & Rosenheck, R. A. (2011). Association of substance use and VA service-connected disability benefits with risk of homelessness among veterans. The American Journal on Addictions, 20(5), 412–419. https://doi.org/10.1111/j.1521-0391.2011.00166.x

Elbogen, E. B., DuBois, C. M., Finlay, A. K., Clark, S., Kois, L. E., Wagner, H. R., & Tsai, J. (2023). How often does homelessness precede criminal arrest in veterans? Results from the U.S. survey of prison inmates. American Journal of Orthopsychiatry, 93(6), 486–493. https://doi.org/10 .1037/ort0000693

Elbogen, E. B., Johnson, S. C., Newton, V. M., Straits-Troster, K., Vasterling, J. J., Wagner, H. R., & Beckham, J. C. (2012). Criminal justice involvement, trauma, and negative affect in Iraq and Afghanistan war era veterans. Journal of Consulting and Clinical Psychology, 80(6), 1097– 1102. https://doi.org/10.1037/a0029967

Fargo, J. D., Brignone, E., Metraux, S., Peterson, R., Carter, M. E., Barrett, T., Palmer, M., Redd, A., Samore, M. H., & Gundlapalli, A. V. (2017). Homelessness following disability-related discharges from active duty

Table 3 Multivariable Cox Regression Models for Demographic, Clinical, Mental Health, Substance Use, Homelessness, and/or Incarceration Characteristics as Predictors of Time-to-Decision for Veteran Disability Benefits in the Study Sample (n = 327,047)

Characteristic Hazard ratio

[95% confidence interval]

Sex Male Reference Female 0.96 [0.94, 0.97]

Age (years) <50 Reference 50–64 0.95 [0.94, 0.96] 65–74 0.98 [0.96, 0.99] 75–79 1.08 [1.07, 1.09] 80+ 1.35 [1.31, 1.38]

Race/ethnicity Non-Hispanic White Reference Non-Hispanic Black 0.99 [0.98, 1.00] Hispanic 1.05 [1.03, 1.06] Mixed race/other 0.94 [0.92, 0.96] Missing 1.00 [0.99, 1.02]

Marital status Married Reference Single/never married 1.03 [1.02, 1.04] Separated/divorced 1.01 [0.99, 1.02] Widowed 1.02 [0.99, 1.05] Missing 1.05 [1.02, 1.08]

Disability ratinga

None/0% Reference 10%–40% 1.15 [1.12, 1.18] 50%–100% 1.19 [1.15, 1.22]

Combat service Yes 0.97 [0.95, 0.98] No Reference

Military sexual trauma Yes 0.95 [0.93, 0.97] No Reference

Charlson Comorbidity Index 1.01 [1.01, 1.02] Mental health disorders Yes 0.99 [0.98, 1.00] No Reference

Substance use disorders Yes 1.03 [1.01, 1.04] No Reference

Homelessness and/or incarceration None Reference Homeless only 1.93 [1.88, 1.99] Incarcerated only 1.12 [0.96, 1.30] Both 2.19 [1.46, 3.29]

a Disability rating as indicated in the Veterans Health Administration records.

6 BEYDOUN, GIANOLI, MEISLER, AND TSAI

military service in Afghanistan and Iraq. Disability and Health Journal, 10(4), 592–599. https://doi.org/10.1016/j.dhjo.2017.03.003

Finlay, A. K., Stimmel, M., Blue-Howells, J., Rosenthal, J., McGuire, J., Binswanger, I., Smelson, D., Harris, A. H., Frayne, S. M., Bowe, T., & Timko, C. (2017). Use of Veterans Health Administration mental health and substance use disorder treatment after exiting prison: The health care for reentry veterans program. Administration and Policy in Mental Health and Mental Health Services Research, 44(2), 177–187. https://doi.org/10 .1007/s10488-015-0708-z

Lucksted, A., Plater, L., Glieger, J., Deriggs, D. J., Kuykendall, L., Taylor, M., & Drapalski, A. (2022). VJP strategic objectives related to stigma- tization of veterans with criminal justice history. https://mirecc.va.gov/ MIRECC/visn5/education/VJP/Y1-Report-Executive-Summary.asp

Maruschak, L. M., Bronson, J., & Alper, M. (2021). Veterans in prison: Survey of Prison Inmates, 2016. Department of Justice, Office of Justice Programs, Bureau of Justice Statistics. https://bjs.ojp.gov/library/publica tions/veterans-prison-survey-prison-inmates-2016

Murdoch, M., Sayer, N. A., Spoont, M. R., Rosenheck, R., Noorbaloochi, S., Griffin, J. M., Arbisi, P. A., & Hagel, E. M. (2011). Long-term outcomes of disability benefits in U.S. veterans with posttraumatic stress disorder. Archives of General Psychiatry, 68(10), 1072–1080. https://doi.org/10 .1001/archgenpsychiatry.2011.105

Pena,M.M., Stoiloff, S., Sparacino, M., & Schreiber Compo, N. (2024). The effects of cognitive bias, examiner expertise, and stimulus material on forensic evidence analysis. Journal of Forensic Sciences, 69(5), 1740– 1757. https://doi.org/10.1111/1556-4029.15565

Rosen, M. I., Afshartous, D. R., Nwosu, S., Scott, M. C., Jackson, J. C., Marx, B. P., Murdoch, M., Sinnott, P. L., & Speroff, T. (2013). Racial differences in veterans’ satisfaction with examination of disability from posttraumatic stress disorder. Psychiatric Services, 64(4), 354–359. https://doi.org/10.1176/appi.ps.201100526

Saxon, A. J., Davis, T. M., Sloan, K. L., McKnight, K. M., McFall, M. E., & Kivlahan, D. R. (2001). Trauma, symptoms of posttraumatic stress disorder, and associated problems among incarcerated veterans. Psychiatric Services, 52(7), 959–964. https://doi.org/10.1176/appi.ps.52.7.959

Stupart, D., Goldberg, P., Krige, J., & Khan, D. (2008). Does examiner bias in undergraduate oral and clinical surgery examinations occur? South African Medical Journal, 98(10), 805–807. https://www.ncbi.nlm.nih.go v/pubmed/19115759

Sundararajan, V., Henderson, T., Perry, C., Muggivan, A., Quan, H., & Ghali, W. A. (2004). New ICD-10 version of the Charlson Comorbidity Index predicted in-hospital mortality. Journal of Clinical Epidemiology, 57(12), 1288–1294. https://doi.org/10.1016/j.jclinepi.2004.03.012

Tsai, J., & Byrne, T. H. (2019). National utilization patterns of veterans affairs homelessness programs in the era of housing first. Psychiatric Services, 70(4), 309–315. https://doi.org/10.1176/appi.ps.201800393

Tsai, J., Flatley, B., Kasprow,W. J., Clark, S., & Finlay, A. (2017). Diversion of veterans with criminal justice involvement to treatment courts: Participant characteristics and outcomes. Psychiatric Services, 68(4), 375– 383. https://doi.org/10.1176/appi.ps.201600233

Tsai, J., Kasprow, W. J., & Rosenheck, R. A. (2013). Latent homeless risk profiles of a national sample of homeless veterans and their relation to program referral and admission patterns. American Journal of Public Health, 103(Suppl. 2), S239–S247. https://doi.org/10.2105/AJPH.2013 .301322

Tsai, J., & Rosenheck, R. A. (2015). Risk factors for homelessness among U.S. veterans. Epidemiologic Reviews, 37(1), 177–195. https://doi.org/10 .1093/epirev/mxu004

Tsai, J., Rosenheck, R. A., Kasprow, W. J., & McGuire, J. F. (2013). Risk of incarceration and other characteristics of Iraq and Afghanistan era veterans in state and federal prisons. Psychiatric Services, 64(1), 36–43. https:// doi.org/10.1176/appi.ps.201200188

Tsai, J., Szymkowiak, D., & Jutkowitz, E. (2022). Developing an operational definition of housing instability and homelessness in Veterans Health Administration’s medical records. PLOS ONE, 17(12), Article e0279973. https://doi.org/10.1371/journal.pone.0279973

U.S. Census Bureau. (2021). American Community Survey: B21001 (table): Sex by age by veteran status for the civilian population 18 years and over. https:// data.census.gov/table?q=B21001:+SEX+BY+AGE+BY+VETERAN+ST ATUS+FOR+THE+CIVILIAN+POPULATION+18+YEARS+AND +OVER&g=010XX00US&tid=ACSDT1Y2021.B21001

U.S. Department of Veterans Affairs. (2023). Fiscal years (FY) 2022–28 learning agenda supplement on homelessness.

Veterans Benefits Administration. (2016). Section D. Claims that require priority processing [M21-1, Part III, Sub part ii]. https://view.officeapps.live .com/op/view.aspx?src=https%3A%2F%2Fwww.benefits.va.gov%2FWAR MS%2Fdocs%2Fadmin21%2Fm21_1%2Fmr%2Fpart3%2Fsubptii%2Fch0 1%2FM21-1III_ii_1_SecD.doc&wdOrigin=BROWSELINK

(Appendix follows)

TIME-TO-DECISION ON DISABILITY BENEFITS 7

Appendix

Appendix Tables

Received November 14, 2024 Revision received April 25, 2025

Accepted May 1, 2025 ▪

Table A1 Time-to-Decision According to Geographical Region (n = 283,100)

Geographical region N M (SD)a

Continental 48,618 111.08 (64.87) Northeast 108,042 109.99 (64.39) Pacific 59,100 112.20 (65.03) Southeast 67,340 109.92 (63.96)

a One-way analysis of variance test: F = 18.66, p < .0001.

Table A2 VA-Connected Disability Rating Among Veterans Who Received Veterans Benefits Administration Disability Benefits for a Mental Health Condition by Homelessness and/or Incarceration Status

Homelessness and/or incarceration n M (SD)a

Total p < .0001 Homeless only 5,928 79.0 (23.9) Incarcerated only 176 68.2 (29.7) Both 26 77.3 (19.3) Neither 272,363 79.1 (23.9)

Note. VA = U.S. Department of Veterans Affairs. a Disability rating as indicated in the Veterans Benefits Administration records.

8 BEYDOUN, GIANOLI, MEISLER, AND TSAI

Reproduced with permission of copyright owner. Further reproduction prohibited without permission.