Work and Recovery from Substance Use Disorder in the VA: Characteristics of Veterans Seeking Work Therapy and Vocational Rehabilitation

Abstract

Objective:

Substance use disorders (SUDs) are prevalent in Veterans and associated with a wide range of deleterious effects. Helping Veterans with SUD identify and engage in meaningful and constructive goal-directed activities may be key to promoting recovery. Work, particularly, has been identified as beneficial. Despite the effects of therapeutic work activities, we still do not have a good understanding of who is enrolling in these services.

Methods:

This study provides an in-depth descriptive look at psychiatric diagnoses, internal factors, disability level, and demographic characteristics of Veterans (N=78) early in the course of recovery from SUD who expressed an interest in engaging in work services as part of their recovery.

Results:

Compared to normative scores, our sample exhibited deficits in multiple areas of functioning including global disability, physical health, social relationships, physical safety, self-esteem, and substance use. Of the 78 Veterans, 17 (22%) chose competitive employment and 61 (78%) chose therapeutic work activity. Compared to Veterans who chose competitive employment, Veterans in the therapeutic work activity branch reported higher rates of medical disabilities that impair the ability to work, were more likely to be retired from employment, had more previous treatment program participation, and had longer periods of abstinence from alcohol.

Conclusions and Implications for Practice:

These findings, though descriptive and preliminary, help us better understand Veterans who are interested in engaging in work activities to promote their recovery, their preference for work therapy versus vocational rehabilitation services, and what additional services they may find most appealing based on their characteristics.

Introduction

SUDs have devastating effects on functioning and are associated with a wide range of negative consequences including homelessness, financial hardship, social and occupational problems, physical and mental health difficulties, and even death (Blakey et al., 2021; Petrakis et al., 2011; Teeters et al., 2017). Substance use is a common problem for adults in the United States with the rate of past-year substance use disorder (SUD) observed to be 17.3%. This rate is even higher for Veterans, at 18.1% (National Survey on Drug Use and Health, 2022). The prevalence of SUD in Veterans has continued to rise since 2009 (Hoggatt et al., 2023). Regardless of presence of diagnosed SUD, compared to demographically matched non-Veterans, Veterans are also more likely to report greater rates of past-month alcohol use, past-month heavy alcohol use, driving under the influence, and cannabis use (Wagner et al., 2007). Relative to non-Veterans with SUD, Veterans with SUD also exhibit lower physical and mental health functioning (Boden & Hoggatt, 2018).

SUDs are chronic conditions, with a relapse rate of about 50% one-year post treatment (McLellan et al., 2002). This rate jumps to 75% in Veterans (Curran et al., 2000). Likelihood of relapse is particularly high early in treatment after individuals complete intensive substance use programs and transition to less structured aftercare (Ross et al., 1995). Some have theorized that high rates of relapse are in part due to the deleterious impact of substance use on social, interpersonal, and occupational activities. As these activities become less meaningful and enjoyable, the person may return to substance use as the one activity that still brings some enjoyment and pleasure, however temporary. In line with this reasoning, it has also been suggested that helping people with SUD identify and engage in meaningful, goal-directed, and enjoyable activities that do not entail substance use may be key to promoting recovery (McKay, 2016).

Indeed, interventions that involve engagement in goal-directed activities have a long history as an adjunctive to traditional substance use treatment. While there are numerous enjoyable goal-directed activities that may aid individuals in recovery, work particularly been identified as beneficial, as it provides structure, socialization, a sense of meaning and purpose, a non-sick identity, and a sense of belonging. In therapeutic communities for SUD, a work structured day has long been recommended as part of standard treatment (Center for Substance Abuse Treatment, 2006; DeLeon et al., 2015; Vanderplasschen et al., 2014).

Several studies have explored the benefits of work in enhancing substance use outcomes, commonly in the form of either supported employment (SE) or therapeutic work activities. SE services exhibit benefits including increased rates of competitive employment in multiple samples including Veterans with PTSD, formerly incarcerated Veterans, and Veterans with spinal cord injury (Probyn et al., 2021). Adding SE on to vocational rehabilitation group programming resulted in higher rates of employment in formerly-incarcerated Veterans with SUD (LePage et al., 2016). However, perceived barriers to SE exist for Veterans, which include substance use, stress, mental health, and medical concerns (Kukla et al., 2016).

Therapeutic work activities are an avenue by which Veterans can engage in work services in an environment that may be more accommodating to the concerns inherent in SE. While therapeutic work activities are not intended to lead to competitive employment, they nevertheless have positive outcomes for Veterans, including reduced alcohol use and psychiatric symptoms and higher rates of employment (Rosenheck & Seibyl, 1997). Adding therapeutic work activities to standard rehabilitation for individuals with SUD leads to better engagement in outpatient substance use treatment, fewer substance-related problems and physical symptoms related to substance use, and lower rates of homelessness and incarceration (Kashner et al., 2002). In our own work, we have also noted the therapeutic effects of work activities, including finding participants who completed a substance use day program and subsequently engaged in a 3-month therapeutic work program had abstinence rates significantly higher than what has been reported in the literature, at approximately 90% days of abstinence at the end of the therapeutic work program and in the 30 days prior to a 6-month follow-up (Bell et al., 2020). In a follow-up study, we extended our assessments to 12 months, and used retrospective chart review to evaluate clinical outcomes in the same group of participants and compare them to a historical comparison sample of participants who also completed a 21-day substance use day program but were either not offered, or did not engage, in therapeutic work activity. Good clinical outcomes were observed for 62–77% of participants engaged in therapeutic work activity but only 27% of the sample that did not engage in work activity. In addition, over 50% of those receiving therapeutic work activity sustained sobriety at 12 months, versus only 18% of those who did not engage in work activity (Bell et al., 2020).

Some research has been conducted comparing SE to therapeutic work activity. At discharge, slightly more Veterans in SE (35%) than therapeutic work services (30%) were competitively employed (Resnick, 2017). An RCT found 67% of Veterans with PTSD assigned to SE gained competitive employment versus 57% of Veterans assigned to therapeutic work activity (Davis et al., 2018). While SE is associated with higher rates of employment compared to work therapy, it should be noted that work therapy is a therapeutic program with a greater emphasis on non-vocational outcomes including well-being, interpersonal functioning, and enhanced efficacy of concurrent therapeutic services.

Despite the positive impacts of engagement in SE and therapeutic work activities, we still do not have a good understanding of who is enrolling in these services. That is, there remains a need to identify what factors characterize individuals who take advantage of these types of services. A better understanding of who is seeking work services as part of their recovery can allow us to better match interventions, customize Veterans’ recovery journeys, and make comprehensive rehabilitation services more Veteran-centered.

The literature on predictors of engagement in work activity in Veterans with SUDs is mostly limited to information about demographics (i.e., sex, age, race, diagnosis, education level, housing status; Drebing et al., 2002; Drebing et al., 2004; Kerrigan et al., 2004). For example, one study found that Veterans with SUD and comorbid psychiatric disorders spent less time in therapeutic work activities and had a lower vocational rehabilitation program completion rate than Veterans with SUD alone (Drebing et al., 2002). Another study examined employment outcomes from VA Therapeutic and Supported Employment Services between 2006 and 2010 (Abraham et al., 2017). This study considered a range of demographic variables, military history, mental health and SUD diagnoses, and housing stability. Descriptive characteristics of the sample included Veterans who were primarily male, White, separated/divorced, and homeless. Many had SUDs and mental health diagnoses, most commonly depression. One aim of these studies has been to inform clinical decision-making regarding rehabilitation for Veterans in different diagnostic groups and make recommendations to programs (i.e., ensure adequate supports are in place for Veterans with comorbid diagnoses, consider additional safeguards to reduce risks of dropout during treatment).

In another study of Veterans referred to vocational rehabilitation programs, having a psychiatric disorder coupled with alcohol use disorder (AUD) was associated with a lower likelihood of participating in vocational rehabilitation compared to having an AUD alone (Sprong et al. 2023). They concluded that an improved understanding is needed of the unique challenges and obstacles faced by Veterans with comorbid disorders who are referred to but do not enroll in vocational rehabilitation.

Unfortunately, little information on other potentially important variables (e.g., self-efficacy, level of disability) is available. We are aware of only one study that expanded beyond typical demographic characteristics to examine psychological factors that were associated with Veteran engagement in work services, which found Veterans with higher employment hope were more likely to enroll in vocational rehabilitation (Sprong et al., 2024).

The current study extends previous research by providing an in-depth descriptive look at psychiatric diagnoses, internal characteristics, disability level, and demographic characteristics of Veterans early in the course of recovery from SUD who express an interest in engaging in work services as part of their recovery. We gain further nuance by investigating differences and similarities in characteristics across two groups – Veterans seeking competitive employment versus Veterans who do not want competitive employment but are interested in therapeutic work services. These findings may help us better understand Veterans who are interested in work services and what additional needs they may have based on their profiles and characteristics.

Methods

Participants

Data was obtained for 78¹ Veterans enrolled in SUD treatment who had consented to participate in a randomized clinical trial to test the effects of therapeutic and competitive work services on improving SUD outcomes in newly recovering Veterans engaged in outpatient treatment.

Potentially eligible participants were in the early stages of recovery from SUD, as indicated by either recently completing an intensive outpatient substance use treatment program, re-engaging in treatment following a relapse, or endorsing interest in addressing or changing substance use. Inclusion criteria were: 1) 18 years of age and older, 2) meeting DSM-V criteria for any current (past year) SUD (e.g., opioids, cocaine, marijuana, amphetamine, or alcohol), and 3) competent to give written informed consent and HIPAA authorization. Exclusion criteria were: 1) involvement in a legal case that may lead to incarceration during study period, 2) developmental disability or medical illness that would prevent therapeutic work activity, as judged by clinical staff providing this service, 3) plans to relocate outside geographic area that would make follow-up unlikely, and 4) lack of willingness to provide contact information of someone who could help us reach the participant in the event that study staff were unable to maintain contact directly.

Procedures

All procedures were reviewed and approved by the Institutional Review Board, and all participants provided written informed consent. In line with eligibility criteria for different work services, the parent study employed a two-branch design: 1) a branch for Veterans who expressed interested in vocational services to obtain competitive employment (i.e. VHA Compensated Work Therapy, resume and job search assistance, and/or referral to VA Vocational Rehabilitation Services), and 2) a branch for Veterans who were not interested in pursuing competitive employment but were interested in therapeutic work activity. The latter branch provided a work experience akin to the VHA’s recently discontinued Incentive Work Therapy (IWT) program.²

Participants were recruited from addiction and mental health clinics throughout a northeastern Veterans Health Administration (VHA) medical center. Following consent, participants were administered a diagnostic interview, along with a brief demographic and psychosocial interview to confirm study eligibility. If deemed eligible, they completed remaining baseline measures. Participants were paid $50 for the baseline assessment.

Recruitment for the larger clinical trial began December 2021. Data analyzed in the current study was collected between December 2021 and November 2024. All measures were administered by a trained master’s level research assistant. For self-report measures, all items were read aloud to participants, and the research assistant noted their responses.

Measures

Baseline assessments included demographic, psychosocial, diagnostic, substance use, addiction severity, symptom, medical condition, quality of life, self-esteem, and self-efficacy measures.

Sociodemographic characteristics

Sociodemographic characteristics that were collected included age, gender, race/ethnicity, years of education, marital status, and housing status.

Financial and employment

Information on disability compensation (i.e., military service connection, SSDI, SSI), other financial supports, retirement status, and employment patterns were collected via participant self-report.

Clinical and substance use diagnoses

The Quick Structured Clinical Interview for DSM-5 (QuickSCID-5; First & Williams, 2021), which was adapted from the Structured Clinical Interview (SCID; First et al., 2015), was used to screen for psychiatric disorders and SUDs. We also assessed for the presence of antisocial personality disorder (ASPD). The master’s level research assistant administering the QuickSCID-5 was trained on diagnostic evaluations, differential diagnosis, and the use of this instrument by a licensed psychologist (JMF).

Service utilization

Number of psychiatric and drug/alcohol hospitalizations (past year and lifetime), number of times the individual participated in intensive SUD treatment programs, and previous participation in work services were collected via participant self-report.

Substance use

In addition to SUD diagnoses, detailed information on substance use (including months of lifetime use and longest period of abstinence) was assessed via an adapted Addiction Severity Index (McLellan et al., 1992).

Psychological characteristics

Several different areas of psychological characteristics were collected via participant self-report spanning symptoms, functional difficulty, self-esteem, self-efficacy, quality of life, global disability, and pain.

Symptoms and functioning.

The Behavior and Symptom Identification Scale (BASIS-24; Cameron et al., 2007) was used to assess transdiagnostic symptoms and associated functional difficulty. The BASIS-24 is a 24-item Likert self-report scale that assesses the frequency and severity of symptoms, with higher scores indicating more serious symptoms or functional difficulties. There is an overall score, with six subscale scores: depression and functioning, interpersonal problems, psychotic symptoms, alcohol/drug use, emotional liability, and self-harm. Each subscale and overall mean score can range from 0 – 4.

Self-esteem.

The Rosenberg Self-Esteem Scale (RSES; Rosenberg, 1965) was used to measure self-esteem. This 10-item Likert self-report measure asks the examinee to indicate to what extent they agree or disagree with statements about their self-worth, usefulness, self-satisfaction, and self-respect. Responses can range from 0 – 3, with a total score ranging from 0 – 30, and higher scores indicating higher self-esteem.

Self-efficacy.

The New General Self-Efficacy Scale was used to assess self-efficacy (NGSES; Chen et al., 2001). This 8-item Likert self-report measure asks examinees to rate their confidence in their capabilities. Responses can range from 1 – 5 with higher scores indicating greater self-efficacy.

Quality of life.

The World Health Organization Quality of Life – BREF (WHOQOL-BREF; Skevington et al., 2004) was used to assess quality of life. This 26-item Likert self-report encompasses domains of physical health, psychological health, social relationships, and environment. Scores range from 0 – 100 with higher scores indicating better quality of life.

Health and disability.

The brief 12-item World Health Organization Disability Assessment Schedule (WHODAS-2; Ustun et al., 2010) was used to assess health and disability. The WHODAS-2 is well validated measure assessing functioning in 6 domains: cognition, mobility, self-care, getting along, life activities, and participation. Higher scores represent higher global disability ratings, indicating worse overall functioning.

Pain.

The Pain, Enjoyment, and General Activity Scale (PEG; Krebs et al., 2009), derived from the Brief Pain Inventory, was used to assess pain intensity and interference. The PEG consists of three items that assess pain intensity, interference with life enjoyment, and interference with general activity, and asks examinees to rate pain severity from 1 – 10 in the past week. Higher scores represent greater self-reported pain levels.

Data Analysis and Results

Categorical variables were summarized using frequencies and percentages. Continuous variables were summarized with descriptive statistics (mean, SD, minimum, maximum). To assist in the interpretation of information about symptom, psychological, and functional characteristics, normative information for these measures was obtained (see Table 2 for sources from which normative scores were drawn).

Table 2.

Sample and normative scores on psychological functioning measures

	Mean	Normative Mean(SD)	Observed Range	Possible Range
RSES Self-esteem ↑	20.15(6.68)	22.62(5.80)a	4 – 30	0 – 30
NGSES Self-efficacy ↑	4.28(.65)	3.87(.54)b	1.88 – 5.00	1 – 5
WHO-QOL BREF ↑
Physical health	66.13(20.77)	73.5(18.1)c	21 – 100	0 – 100
Psychological health	67.71(21.73)	70.6(14.0)c	21 – 100	0 – 100
Social relationships	56.74(26.28)	71.5(18.2)c	0 – 100	0 – 100
Environment	67.85(17.78)	75.1(13.0)c	16 – 97	0 – 100
WHODAS-2 Global disability ↓	11.13(8.68)	3.1(5.3)d	0 – 34	0 – 48
PEG Pain interference ↓	2.33(2.26)	4.1(3.1)e	0 – 7.33	0 – 10

Note.

↓ = lower scores on this measure represent less functional disability

↑ = higher scores on this measure represent better functioning

^anormative scores drawn from Sinclair et al., 2010.

^bnormative scores drawn from Chen et al., 2001.

^cscores drawn from Hawthorne et al., 2006.

^dnormative scores drawn from Andrews et al., 2009.

^escores drawn from Krebs et al., 2009.

Demographic information for the sample overall

Please refer to Table 1 for sample characteristics. On average, participants were 54 years old, primarily male (91.0%), and diverse in terms of race and ethnicity. The most common primary SUD diagnosis was alcohol use disorder (65.4%). Nearly three quarters of the sample (73.1%) met criteria for at least one psychiatric disorder diagnosis, the most common being major depressive disorder (25.6%) and post-traumatic stress disorder (24.4%). Over half the sample (56.4%) had participated in work services before and most (80.8%) had participated in an intensive SUD program in the past. Most Veterans (78.2%) were stably housed upon entry to the study. Almost three-quarters (70.5%) were receiving compensation for disability with the average service connection rating greater than 50%. Retirement from employment was endorsed by one-third (33.3%) of the sample. The most common referral source to the study was the VHA intensive outpatient substance use day program (48.6%), followed by the VHA outpatient substance use clinic (39.3%), and VHA methadone maintenance clinic (2.8%).

Table 1.

Sample characteristics (N = 78)

	Mean (SD) or Percentage	Range
Demographics
Age (Years)	Mean(SD) = 54.63(13.19)
Years of Education	Mean(SD) = 13.03(1.15)
Reported Gender
Male	91.0%
Female	7.7%
Other	1.3%
Race
American Indian/Alaska Native	2.6%
Black/African American	53.8%
White	37.2%
Mixed Race	5.1%
Unknown/Not Reported	1.3%
Ethnicity
Non-Hispanic/Latino	93.6%
Marital Status
Single, never married	38.5%
Married/Separated/Divorced/Widowed	61.5%
Primary Substance Use Diagnosis
Alcohol use disorder	65.4%
Cannabis use disorder	2.6%
Cocaine use disorder	25.6%
Opioid use disorder	2.6%
Other substance use disorder	2.6%
Substance Use (in months)
Lifetime use – AUDa	369.67(213.65)	1 – 707
Lifetime use – Other SUDb	146.26(136.64)	0 – 660
Longest period of abstinence – Primary AUDc	36.67(38.63)	0 – 168
Longest period of abstinence – Primary Other SUDd	95.90(45.12)	0 – 576
Hospitalizations
# of lifetime psychiatric hospitalizations	3.09(7.03)	0 – 50
# of lifetime drug/alcohol hospitalizations	8.83(16.64)	0 – 100
# of times previously participated in intensive drug/alcohol treatment program	4.63(5.51)	0 – 30
Psychiatric Diagnosis
Any Psychiatric Disorder	73.1%
Bipolar Disorder	10.3%
Generalized Anxiety Disorder	3.8%
Major Depressive Disorder	25.6%
Panic Disorder	2.6%
Post-Traumatic Stress Disorder	24.4%
Schizoaffective Disorder/Schizophrenia	5.1%
Anti-Social Personality Disorder	17.9%
Self-Reported Medical Conditions
Medical illness that impairs ability to work	29.5%
Diagnosed pain condition	59.0%
Previous Work Participation
Previously participated in Incentive Work Therapy	23.1%
Previously participated in Compensated Work Therapy	34.6%
Previously participated in Supported Employment	17.9%
Previously participated in other work therapy services	9.0%
Stable Housing – Past 90 Days
Yes	78.2%
Financial
Receiving compensation for disability (SSI, SSDI, SC)	70.5%
Service Connection Percentage	Mean(SD) = 57.23(37.39)
Employment (past 3 years) e
Full-time competitive	23.1%
Part-time competitive	32.1%
Irregular/part-time/odd jobs	10.3%
Unemployed	33.3%
Student training program	1.3%
Retired from employment	33.3%

Note. For substance use variables assessing longest period of abstinence, analyses were conducted for Veterans who had a primary substance use diagnosis for that particular substance (i.e., longest period of abstinence from alcohol was assessed for Veterans who had a primary diagnosis of alcohol use disorder). Thus, sample sizes for each analysis vary.

^aN = 78.

^bN = 21 – 67.

^cN = 51.

^dN = 23.

^eSome participants reported multiple types of employment in the past 3 years.

Differences in psychological functioning in our sample vs normative scores

Compared to normative scores, our sample exhibited deficits in multiple areas of functioning (see Table 2). Particularly striking were Veterans’ reported difficulties with global disability, physical health, social relationships, physical safety/security, self-esteem, and substance use; in all of these areas, Veterans exhibited worse functioning than what was reported for non-age corrected normative samples. Compared to norms, our sample reported worse global disability, t(8900) = 13.226, p <.0001, worse physical health, t(942) = 3.400, p < .001, worse social relationships, t(942) = 6.575, p <.0001, worse environment, t(942) = 4.558, p < .0001, lower self-esteem, t(579) = 3.426, p <.001, and higher alcohol/drug use, t(123) = 5.593, p < .0001. In two areas, our sample reported better functioning than normative samples. First, in our sample pain interference was lower, t(714) = 4.885, p <.0001 than the normative sample, which was a sample of Veterans recruited from ambulatory care settings. Second, self-efficacy was higher, t(399) = 5.773, p < .0001 in our sample compared to a normative sample, which was a sample of undergraduate students.

Summary of similarities between treatment arms on characteristics

Next, the sample was split by treatment branch (those interested in therapeutic work activity versus competitive employment) to examine whether there were differences in characteristics. Of the 78 Veterans, 17 (22%) elected to be in the competitive employment branch and 61 (78%) in the therapeutic work activity branch.

Distributions of baseline demographic and clinical characteristics were summarized and compared between branches using chi-square or t-tests. In cases where the sample size was small, the Fisher’s Exact Test was used. All primary tests were two-sided and considered statistically significant at the alpha = 0.05 threshold. Results are summarized in Table 3.

Table 3.

Statistical comparisons between competitive employment and therapeutic work activity branches

	Competitive Employment Branch (N=17)	Therapeutic Work Activity Branch (N=61)	Statistic
Demographics
Age (Years)	52.94(14.86)	55.10(12.78)	t(76) = −.59, p = .554
Years of Education	13.29(1.61)	12.95(.98)	t(76) = 1.10, p = .277
Reported Gender
Male	94.1%	90.2%	χ2 (2) = .39, p = .821
Female	5.9%	8.2%
Other	0%	1.6%
Race
American Indian/Alaska Native	5.9%	1.6%	χ2 (4) = 2.46, p = .653
Black/African American	41.2%	57.4%
White	47.1%	34.4%
Mixed Race	5.9%	4.9%
Unknown/Not Reported	0%	1.6%
Ethnicity
Non-Hispanic/Latino	94.1%	93.4%	χ2 (1) = .01, p = .920
Marital Status
Single, never married	47.1%	36.1%	χ2 (1) = .68, p = .410
Married/Separated/Divorced/Widowed	52.9%	63.9%
Substance Use (in months)
Lifetime use – AUDa	377.00(225.82)	367.62(212.04)	t(76) = .16, p = .874
Lifetime use – Other SUDb	89.07(103.24)	136.22(144.21)	t(76) = 1.26, p = .212
Longest period of abstinence – Primary AUDc	17.46(26.27)	43.24(40.23)	t(49) = −2.15, p <.05 *
Longest period of abstinence – Primary Other SUDd	150.00(76.37)	67.00(128.34)	t(19) = .89, p = .387
Primary Substance Use Diagnosis
Alcohol use disorder	76.5%	62.3%	χ2 (5) = 2.97, p = .705
Cannabis use disorder	0%	3.2%
Cocaine use disorder	17.6%	27.9%
Opioid use disorder	5.9%	1.6%
Other substance use disorder	0%	3.2%
Hospitalizations
# of lifetime psychiatric hospitalizations	1.12(3.22)	3.64(7.69)	t(76) = −1.32, p = .192
# of lifetime drug/alcohol hospitalizations	3.53(3.56)	10.31(18.49)	t(72.214) = −2.69, p<.01 **
# of times participated in intensive drug/alcohol treatment programs	3.12(3.71)	5.05(5.87)	t(76) = −1.29, p = .203
Psychological Functioning
BASIS-24 total	1.00(.59)	1.04(.74)	t(76) = −.23, p = .820
Functioning subscale	1.10(.80)	1.14(.10)	t(76) = −.16, p = .875
Interpersonal problems subscale	1.16(.87)	1.12(.96)	t(76) = .16, p = .872
Psychotic symptoms subscale	.37(.59)	.62(.85)	t(76) = −1.12, p = .267
Alcohol/drug use subscale	.95(.76)	1.04(.88)	t(76) = −.38, p = .704
Emotional lability subscale	1.38(1.04)	1.42(1.14)	t(76) = −.13, p = .897
Self-harm subscale	.06(.24)	.08(.36)	t(76) = −.20, p = .841
RSES Self-esteem	21.12(5.36)	19.89(7.02)	t(76) = .67, p = .505
NGSES Self-efficacy	4.29(.56)	4.28(.68)	t(76) = .08, p = .938
WHOQOL-BREF Physical health	65.88(18.80)	66.20(21.44)	t(76) = −.06, p = .956
WHOQOL-BREF Psychological health	66.82(19.51)	67.95(22.45)	t(76) = −.19, p = .851
WHOQOL-BREF Social relationships	55.41(24.38)	57.11(26.97)	t(76) = −.24, p = .815
WHOQOL-BREF Environment	63.06(17.23)	69.18(17.84)	t(76) = −1.26, p = .211
WHODAS global disability	9.88(6.72)	11.48(9.17)	t(76) = −.67, p = .507
PEG-3	2.92(2.59)	2.02(2.10)	t(21) = .90, p = .379
Psychiatric Diagnosis
Bipolar Disorder	7.1%	16.3%	χ2 (6) = 6.00, p = .423
Generalized Anxiety Disorder	11.8%	1.6%
Major Depressive Disorder	29.5%	24.3%
Panic Disorder	0%	3.3%
Post-Traumatic Stress Disorder	23.5%	24.6%
Schizoaffective Disorder/Schizophrenia	14.3%	4.7%
Any Psychiatric Disorder	82.4%	70.5%	χ2 (1) = .95, p = .330
Anti-Social Personality Disorder	11.8%	19.7%	χ2 (1) = .56, p = .452
Self-Reported Medical Conditions
Medical illness that impairs ability to work	5.9%	36.1%	χ2 (1) = 5.83, p <.05*
Diagnosed pain condition	58.8%	59.0%	χ2 (1) = .03, p = .864
Previous Work Participation
Previously participated in IWT	5.9%	27.9%	χ2 (1) = 3.62, p = .057+
Previously participated in CWT	29.4%	36.1%	χ2 (1) = .26, p = .610
Previously participated in supportive employment	0%	23.0%	χ2 (1) = 4.76, p < .05*
Previously participated in other work therapy services	5.9%	9.8%	χ2 (1) = .254, p = .614
Stable Housing – Past 90 Days
Yes	82.4%	77.0%	χ2 (1) = .22, p = .639
Financial
Receiving compensation for disability (SSI, SSDI, SC)	64.7%	72.1%	χ2 (1) = .35, p = .553
Service Connection Percentage	53.33(41.23)	58.26(36.90)	t(41) = −.35, p = .730
Employment (past 3 years)
Full-time competitive	29.4%	21.3%	χ2 (4) = 1.23, p = .873
Part-time competitive	35.3%	31.1%
Irregular/part-time/odd jobs	5.9%	11.5%
Unemployed	29.4%	34.4%
Student training program	0%	1.6%
Retired from employment	5.9%	41.0%	χ2 (1) = 7.37, p < .01**

Note.

⁺p<.10.

^*p<.05.

^**p<.01.

For substance use variables assessing longest period of abstinence, analyses were conducted for Veterans who had a primary substance use diagnosis for that particular substance (i.e., longest period of abstinence from alcohol was assessed for Veterans who had a primary diagnosis of alcohol use disorder). Thus, sample sizes for each analysis vary.

^aN = 78.

^bN = 21 – 67.

^cN = 51.

^dN = 23.

The two groups were not significantly different on demographics (age, race, gender), psychological functioning (self-esteem, self-efficacy), diagnoses (substance use, psychiatric), housing status, and employment patterns.

Summary of significant differences between treatment arms

Compared to the competitive employment group, the therapeutic work activity group had higher reported rates of medical disabilities that impair the ability to work, were more likely to be retired from employment, had more instances of prior participation in intensive SUD treatment programs, and had higher rates of previous participation in both supported employment as well as therapeutic work activity (i.e., IWT), though the latter finding was just shy of reaching statistical significance. The therapeutic work activity group also had longer periods of abstinence from alcohol in their past compared to the competitive employment group.

Discussion

The current study extends previous research by providing an in-depth descriptive look at characteristics of Veterans who elect to engage in work services as part of their recovery from AUD and SUD. Previous studies have mostly focused on the impact of co-occurring substance use and psychiatric conditions on engagement in vocational rehabilitation services. Our study expands upon this research by specifically characterizing Veterans who are early in the course of recovery from substance use and who are interested in taking advantage of programs which have been associated with better SUD outcomes, namely work services. Our study is further unique by examining a host of psychological functioning, service utilization, and disability indices.

Additionally, our study examined differences in these characteristics across two groups – Veterans seeking competitive employment and Veterans who are not interested in employment but wish to engage in therapeutic work activity. While the two groups exhibited similarities in many characteristics, some significant differences emerged, though given the small sample size, conclusions should be considered provisional rather than conclusive until further research is conducted. These preliminary findings may help guide future development of recovery interventions that are matched to Veteran areas of need, help us better understand the population of Veterans who may be interested in engaging in work activities to promote their recovery, and what additional services they may want or need based on their characteristics.

Veterans with SUDs who are interested in work services are high service utilizers

Veterans with SUDs who expressed interest in work services have complex, multifaceted needs. Our sample was similar to broader samples of Veterans enrolled in work service programs where SUD was not an inclusion criteria in terms of demographic characteristics including gender, age, race, marital status, years of education, and psychiatric diagnoses (e.g., Drebing et al., 2004). Our sample experienced high rates of hospitalizations and physical impairment. Psychiatric diagnoses were common, most notably depressive disorders and PTSD. Over 20% of the Veterans in our sample reported experiencing current housing instability. Compared to norms, our sample exhibited deficits in psychological functioning, self-esteem, quality of life, and other disability indices. In sum, Veterans with SUDs who are interested in work services may need interventions that can bolster psychological characteristics like self-esteem and target these complex needs.

To promote recovery, both competitive and therapeutic work services are needed

It appears both competitive employment and therapeutic work services are sought by Veterans early in the course of recovery from SUD. A portion of Veterans may specifically seek therapeutic work activity even when vocational rehabilitation services are an option, as suggested by a study, which reported that more Veterans enrolled in VA therapeutic work activity (57.5%) compared to supported employment services (21.0%; Abraham et al., 2017); that is 73.2% of those interested in work chose VA therapeutic work services. This is similar to our finding that 78% of our participants preferred VA therapeutic work services. Another study reported that Veterans in VA vocational rehabilitation programs vary widely in terms of goals, with about half identifying competitive employment as a goal at all and only 5% identifying competitive employment as their primary goal (Drebing et al., 2004). For Veterans not interested in competitive employment, we need a service that will provide the psychosocial benefits of participating in therapeutic work activity, but also reflects their preference for employment options. If VHA is only offering competitive employment services, many Veterans who want productive, meaningful activities but not necessarily competitive employment will be left out. Thus, both competitive and non-competitive work options are necessary to align with Veterans’ individualized goals and needs.

In addition to offering different types of work services based on Veteran preference, there are additional factors that could be incorporated into comprehensive treatment to further align with patient-centered, recovery-oriented care principles. One element is shared decision-making, a process where providers and patients collaborate on treatment decisions. In Veterans, the inclusion of shared-decision making is associated with lower levels of opioid misuse (Somohano et al., 2023) and higher rates of treatment engagement (Hessinger et al., 2018), and is preferred by Veterans versus making decisions completely independently or leaving decisions solely to their provider (Park et al., 2014). In the context of work activities, shared decision-making may include a discussion around available services, the differences in these services (i.e., pay, flexibility), and how each of these services may or may not affect the Veteran’s current standing including service connection.

For Veterans on recovery journeys, changing substance use patterns can be an arduous and multifaceted process. Recovery from SUD has been characterized as a process that involves improving one’s life, moving toward wellness, and maximizing functioning across a variety of domains (psychosocial, physical, mental, spiritual; Brophy et al., 2023). A major component of this recovery process is discovering meaning and purpose in life (El-Guebaly, 2012). Work services, whether competitive employment or therapeutic work activity, can provide meaning and purpose in addition to more foundational benefits including having a structured schedule, avoiding boredom, and keeping the mind and body occupied.

Veterans seeking therapeutic work activity may differ from those seeking competitive employment

When comparing Veterans who sought competitive employment with those who were not interested in competitive employment but nevertheless wanted to engage in therapeutic work activity, the groups were similar in some respects (i.e., self-esteem, psychological health, psychiatric diagnoses). However, there were some notable differences that might help us better understand the characteristics of Veterans in our sample who preferred therapeutic work activity versus those who chose competitive employment.

Despite being similar in age, Veterans seeking non-competitive vocational services were more likely to be retired from employment than those seeking competitive employment. Research suggests that competitive employment was not a common goal for Veterans of retirement age (Drebing et al., 2004). These Veterans are looking for meaningful, goal-directed activities but perhaps are searching for a work environment that is more accommodating and provides greater flexibility or enjoy “giving back” to VA. A work therapy program on the campus of the VA gives them a chance to interact with other Veterans and to feel camaraderie they knew during military service. Further, they may be retired due to age or physical health status and perceive that they are not able to work competitively. These Veterans may also have retired for medical or health reasons, as we found that Veterans seeking therapeutic work activity were more likely to endorse a medical disability that impairs their ability to work and had more lifetime hospitalizations for substance use.

Overall, many Veterans had prior experience with substance use treatment and work services. Re-engaging with work services in the current study likely means they found these services helpful in the past. Examining the findings more closely, Veterans choosing therapeutic work activity had more prior experience with SE than those who chose competitive employment. While we did not obtain specific information that would explain why Veterans who had previous experience with services with the goal of obtaining competitive employment would now be more interested in non-competitive work activity, we speculate that this might be driven by aspects of the competitive employment setting that may be more challenging for Veterans including finding these settings more stressful, less supportive, or more demanding for those with physical or psychiatric restrictions. They may even be associated with previous substance use and be a trigger for return to use. Indeed, prior research has indicated that Veteran goals for vocational rehabilitation often include working in a setting that accommodates physical or psychiatric limitations, allows participants to meet family responsibilities and stay in treatment, and that is perceived as less stressful and helps maintain sobriety and build coping skills (Drebing et al., 2004).

Limitations

There are several limitations to the current study. Notably, the small sample size may reduce the confidence in the results we observed. From a statistical standpoint, this means some analyses may have been underpowered to detect effects, resulting in Type II error. Nevertheless, we did discover several significant differences between groups. Future studies should attempt to replicate these findings with larger Veteran samples. This would also allow for more sophisticated analyses (i.e., combined models) to reveal when variables are distinctively associated with group membership. In addition, it would be important to know why Veterans chose therapeutic work activities over competitive employment, or vice versa. In the future, qualitative interviews could be used to explore reasons behind such choices.

Second, electing to participate in the study served as an index of interest in vocational services. Veterans who were not interested in work services were not enrolled in the study. Thus, we do not have a comparison group of Veterans who were not interested in pursuing those services. A comparison with Veterans who are not interested in any kind of vocational services could offer a better understanding of what factors may make it more likely that Veterans are interested in seeking out work services and offer clues to how existing SUD services might be refined or supplemented to enhance these factors/predispositions.

In addition, the composition of the sample may affect the generalizability of the findings. For instance, an inclusion criterion of the current study was a plan to stay in the geographic area after treatment. Thus, our findings can’t be generalized to Veterans who are planning to move out of state, whether to return to a different geographic area or seek longer-term treatment. Another consideration are the mixed characteristics (i.e., psychiatric diagnoses) of the sample. The heterogeneity of the sample could have several impacts on generalizability including the introduction of statistical noise (potentially masking relationships) and confounding variables (possibly complicating interpretation). This could mean that the results observed in this study may not be replicated in a more homogenous sample.

Lastly, most measures in the current study were self-report measures. A common question is whether Veterans will accurately self-report on sensitive information (i.e., substance use). Prior research indicates that Veteran self-reports of substance use demonstrate high concordance with objective measures including urine screens (Calhoun et al., 2000). In our sample, we also observed a wide range of scores across measures, which further suggests that Veterans reported accurately and honestly.

Conclusion

Work services can act as a way for Veterans to engage in meaningful, goal-directed activity as part of their SUD recovery journey. Many Veterans in our sample chose therapeutic work activity over competitive employment, suggesting that for many, engaging in work is not simply for a paycheck but represents something deeper – perhaps a means of social connection, achievement, or self-efficacy. In sum, a subset of Veterans want non-competitive, therapeutic work services, which can be an integral component in SUD recovery.

Better understanding the characteristics of Veterans seeking work activity as part of their recovery from SUD could inform interventions for this population. For instance, this knowledge may help VA develop supports that are targeted and relevant. Our sample exhibited deficits in psychological functioning compared to normative samples. One example is self-esteem. Based on this finding, explicitly incorporating methods of increasing self-esteem into work services may be indicated. These augmentations may include eliciting personal strengths (Seligman et al., 2005), self-compassion exercises (Neff, 2003), and building supportive connections (Leary & Baumeister, 2000).

Impact and Implications Statement:

For Veterans in recovery from substance use, work can be a meaningful, goal-directed activity that can enhance functioning and promote abstinence. This study described characteristics of Veterans who were seeking competitive employment and those who were not interested in employment but wished to engage in therapeutic work activity as part of their recovery journey. These findings provide deeper knowledge of who is seeking work services as part of their recovery and how those who were interested in competitive employment differ from those seeking therapeutic work activity, which may allow us to better match interventions, customize Veteran’s recovery journeys, identify potential treatment targets, and make comprehensive rehabilitation services more Veteran-centered.