Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Sep 1.
Published in final edited form as: Drug Alcohol Depend. 2021 Jul 21;226:108907. doi: 10.1016/j.drugalcdep.2021.108907

Factors associated with obtaining employment among opioid use disorder patients enrolled in a therapeutic workplace intervention

August F Holtyn 1, Forrest Toegel 1, Matthew D Novak 1, Kenneth Silverman 1
PMCID: PMC8355098  NIHMSID: NIHMS1726184  PMID: 34311206

Abstract

Background.

Unemployment is common among adults who have substance use disorder that often persists during treatment and recovery. We identified patient characteristics that were associated with obtaining employment among unemployed adults in opioid use disorder treatment.

Methods.

This analysis used data from participants (N=91) who were enrolled in a randomized controlled trial evaluating the effectiveness of a therapeutic workplace in promoting drug abstinence and employment. After a 3-month training period (Phase 1), participants were randomly assigned to a study group and could work for 12 months with an employment specialist who assisted participants in obtaining employment (Phase 2). A logistic regression model was used to identify patient characteristics that were associated with obtaining employment.

Results.

Of the 91 participants, 39 (42.9%) obtained employment. Compared to participants who did not obtain employment, participants who obtained employment worked more days in the therapeutic workplace during the training period (Phase 1) [OR (95% CI)=1.072 (1.015–1.132), p=.014], provided more opiate- and cocaine-negative urine samples while seeking employment [OR (95% CI)=1.015 (1.002–1.027), p=.025], and reported not usually being unemployed at study intake [OR (95% CI)=0.229 (0.080–0.652), p=.007].

Conclusions.

Our analyses suggest that among unemployed adults in opioid use disorder treatment, those with the lowest rates of therapeutic workplace attendance, lowest rates of drug abstinence while seeking employment, and relatively long histories of unemployment are the least likely to obtain employment. These relations are potentially addressable at a practical level, and future research could build on these findings to improve the effectiveness of employment-based interventions.

Keywords: employment, unemployment, incentives, therapeutic workplace, opioid use disorder

1. Introduction

Unemployment is common among adults who have substance use disorder (SUD) and has been associated with worse treatment outcomes (Henkel, 2011). For example, unemployed adults in SUD treatment are more likely to continue using drugs during treatment (Ginexi et al., 2003), are less likely to stay in treatment (O’Connor et al., 2020), and are more likely to relapse following treatment (Brewer et al., 1998) compared to those who are employed. Furthermore, employment problems often persist during treatment and recovery (Eddie et al., 2020; Hubbard et al., 2003; Miguel et al., 2019), and a history of substance use may negatively impact an individual’s labor market outcomes (Earnshaw et al., 2013; Richardson and Epp, 2016; Woods and Joseph, 2015). Providing employment interventions in conjunction with SUD treatment may be needed to help unemployed adults with histories of substance misuse obtain employment (Holtyn et al., 2015; Holtyn et al., 2017; Silverman et al., 2016b).

The therapeutic workplace is an employment intervention that was designed to address the interrelated problems of unemployment and substance use. Under this two-phase intervention, therapeutic workplace participants are paid for engaging in job-skills training in Phase 1 and for performing real jobs in Phase 2. To promote drug abstinence, participants must provide drug-negative urine samples to earn maximum pay in both phases (Silverman, 2004; Silverman et al., 2016a; Silverman et al., 2018; Silverman et al., 2019). In a randomized controlled trial, we evaluated the effectiveness of the therapeutic workplace intervention in promoting community employment and polydrug (opiate and cocaine) abstinence in unemployed adults in opioid use disorder treatment (Holtyn et al., 2020). Some participants obtained employment during the 12-month intervention evaluation period while others did not. The identification of factors that were associated with obtaining employment could provide valuable information that could be used to improve the effectiveness of employment-based interventions. The purpose of the present study was to identify treatment and patient characteristics that were associated with obtaining employment.

2. Material and Methods

This study describes a secondary analysis of data collected during a randomized controlled trial. After a 3-month abstinence initiation and training period (Phase 1), participants were randomly assigned to a Control group or an Abstinence-Contingent Wage Supplement group (Phase 2). After random assignment, all participants were invited to work for 12 months with an employment specialist who assisted participants in obtaining employment in a community job. The primary outcomes and detailed methodology have been reported previously (Holtyn et al., 2020; Toegel et al., 2020). Methodological details relevant to the present analyses have been reproduced here for reference.

2.1. Setting and Participants

The controlled trial was conducted in a therapeutic workplace at the Center for Learning and Health in Baltimore, Maryland. The therapeutic workplace contained workrooms with individual workstations and a urinalysis laboratory. Participants were enrolled in the trial from November 2015 through April 2018. Eligible applicants were 18 years or older, were unemployed, were enrolled in or eligible for methadone or buprenorphine maintenance treatment, provided an opioid-positive (opiates, methadone, or buprenorphine) urine sample, self-reported an interest in gaining employment, and lived in or near Baltimore City. Applicants were excluded if they reported current suicidal or homicidal ideation; had active hallucinations, delusions, or thought disorder; were currently considered a prisoner; or had physical limitations that prevented typing. The Johns Hopkins Medicine Institutional Review Board approved the study and all participants provided written informed consent.

At study intake, participants were an average (SD) of 47.7 (10.3) years old. More than half of the participants identified as male (54.9%) and Black (56.0%). Participants had an average (SD) of 11.5 (1.8) years of education and reported that their longest full-time job was an average (SD) of 6.6 (5.6) years. Participants were enrolled in methadone (93.4%) or buprenorphine (6.6%) treatment. Nearly half of the participants reported engaging in injection drug use (46.2%), and most participants provided opiate- and cocaine-positive urine samples (62.6%).

2.2. Phase 1 Procedures

Participants were invited to work in the therapeutic workplace for 3 months. During that time, they could work for up to 4 hours every weekday on computer-based programs that taught basic education and data-entry skills. They could earn about $10 per hour, and earnings were paid through reloadable credit cards. Participants were asked to provide urine samples three times per week on mandatory collection days (typically on Mondays, Wednesdays, and Fridays). Urine samples were judged positive for opiates or cocaine if they contained at least 300 ng/ml of morphine (opiates) or 150 ng/ml of benzoylecgonine (cocaine). If a participant’s urine sample tested positive for opiates or cocaine, they could still work but for decreased pay. To maintain maximum pay, they had to provide opiate- and cocaine-negative urine samples on mandatory collection days. Participants who attended the workplace on at least 10 of the last 20 days of Phase 1 were eligible for Phase 2.

2.3. Phase 2 Procedures

Phase 2 lasted 12 months. Prior to beginning the phase, participants were randomly assigned to an Abstinence-Contingent Wage Supplement group (n=44) or a Control group (n=47) using a computerized urn randomization procedure that balanced groups based on characteristics that could influence outcome: (1) had a high school diploma or GED (yes/no); (2) percentage of opiate-positive urine samples collected at monthly assessments throughout Phase 1 (percentage ≥ the rolling median, yes/no); and (3) percentage of cocaine-positive urine samples collected at monthly assessments throughout Phase 1 (percentage ≥ the rolling median, yes/no).

2.3.1. Abstinence-Contingent Wage Supplement Group

Abstinence-Contingent Wage Supplement participants were invited to attend the therapeutic workplace for up to 4 hours every weekday and could work with an employment specialist to seek employment in a community job. Participants in this group could earn stipends for working with the employment specialist (up to $10 per hour for up to 20 hours per week) and wage supplements for working in a community job (up to $8 per hour for up to 40 hours per week). To earn the maximum amount in stipends or wage supplements, participants had to provide opiate- and cocaine-negative urine samples under a routine drug-testing schedule (three times per week) that became random and more intermittent over time.

2.3.2. Control Group

Control participants were invited to attend the therapeutic workplace for up to 4 hours every weekday and could work with an employment specialist to seek employment in a community job. However, participants in this group did not receive stipends for working with the employment specialist.

2.4. Study Assessments

Assessments were conducted at study intake and every month during Phases 1 and 2. At these assessments, participants completed the Addiction Severity Index–Lite (ASI–Lite) (McLellan et al., 1985) and provided a urine sample that was tested for drugs. Participants earned $30 for intake and monthly assessments, which was paid through reloadable credit cards.

2.5. Data Analyses

The primary outcome of interest was employment (ever employed based on self-reported employment on the ASI–Lite) during the 12-month intervention evaluation period (Phase 2). Logistic regression was used to identify variables that were associated with obtaining employment. For these analyses, we selected an a priori subset of variables to examine based on the literature and our experiences implementing employment-based reinforcement in previous trials (Holtyn et al., 2016; Silverman et al., 2019). Specifically, we selected five variables: (1) reported receiving pension, disability, or social security support on the ASI–Lite at study intake; (2) reported being usually unemployed in the past three years on the ASI–Lite at study intake; (3) percentage of opiate- and cocaine-negative urine samples at the Phase 1 monthly assessments; (4) percentage of days worked in the therapeutic workplace during Phase 1; and (5) percentage of opiate- and cocaine-negative urine samples at the Phase 2 monthly assessments. Urine samples that were not collected due to a missed monthly assessment were imputed as the adverse outcome (i.e., opiate- and cocaine-positive urine). Analyses included all randomized participants and were adjusted for covariates used for stratification. The magnitude of effect was expressed using odds ratios with 95% confidence intervals. Stata Statistical Software: Release 15 (College Station, TX; StataCorp LLC) was used to perform these analyses.

3. Results and Discussion

Of the 91 participants, 39 (42.9%) obtained employment and 52 (57.1%) did not during the year-long intervention. This finding underscores the persistent nature of unemployment among adults in SUD treatment and recovery, and highlights the need for further development of effective employment interventions. Employment is one of the most widely acknowledged social determinants of health and well-being (Bambra, 2011) and has long been recognized as an important SUD treatment outcome (Magura and Marshall, 2020; Richardson and Epp, 2016).

Table 1 shows results of the adjusted logistic regression analysis. Three of the five variables selected a priori were associated with obtaining employment. Participants were more likely to obtain employment if they worked more days in the therapeutic workplace during the training phase (Phase 1), provided more opiate- and cocaine-negative urine samples during Phase 2, or were not usually unemployed during the three years prior to study intake. Urinalysis results during Phase 1 and financial support from pension, benefits, or social security were not significantly associated with obtaining employment. Figure 1 provides individual participant data for the three variables that were collected during Phases 1 and 2 for participants who did and did not obtain employment.

Table 1.

Adjusted logistic regression results.

Obtained Employment
Yes (n=39) No (n=52) Adjusted Odds Ratio 95% CI p-value
Variable
Phase 1 Opiate- & Cocaine-Negative Urine, % 74.4 58.3 1.012 1.000–1.024 .057
Phase 1 Days Attended Therapeutic Workplace, % 87.2 82.2 1.072 1.015–1.132 .014
Phase 2 Opiate- & Cocaine-Negative Urine, % 63.0 47.9 1.015 1.002–1.027 .025
Pension, Benefits, or Social Security Support, % 25.6 51.9 0.367 0.136–0.995 .052
Usually Unemployed in the Past 3 Years, % 56.4 86.5 0.229 0.080–0.652 .007
Open in a new tab

Figure 1.

Figure 1.

Open in a new tab

Results for the three variables that were collected during Phases 1 and 2 for participants who did and did not obtain employment. Within each panel, dots show percentages for individual participants and bars show group means.

Development of effective employment interventions for adults with SUD is an important consideration for the SUD treatment field. This study suggests that among unemployed adults in opioid use disorder treatment, those with the lowest rates of therapeutic workplace attendance during the training phase (Phase 1), lowest rates of drug abstinence while seeking employment, and relatively long histories of unemployment are the least likely to obtain employment. These relations are potentially addressable at a practical level; below we describe how future research and practice could build on these findings to improve the effectiveness of employment-based interventions.

The finding that participants who had higher rates of therapeutic workplace attendance in Phase 1 were more likely to obtain employment in Phase 2 is consistent with prior findings indicating that therapeutic workplace attendance predicts treatment outcomes (Donlin et al., 2008; Holtyn et al., 2016). These findings suggest that promoting high rates of engagement early in treatment may improve future employment outcomes. Increasing workplace attendance and the reinforcing value of the workplace – such as by changing the type of work to meet each patient’s interests or by increasing the hourly pay rate – could be examined as potential methods to improve treatment engagement and subsequent employment outcomes. Tailoring the implementation of the employment-based intervention in this manner may benefit a larger proportion of unemployed adults with SUD.

Illicit drug use was associated with worse employment outcomes. Those with the highest rates of illicit opioid and cocaine use were the least likely to obtain employment. For patients who continue to engage in high rates of illicit drug use, increasing the intensity or changing the type of SUD treatment may be crucially important to improving drug use and may have the added benefit of improving employment outcomes. Worth noting, however, several non-abstinent participants were successful at obtaining employment. Thus, we do not suggest withholding employment services based on ongoing substance use. Instead, treatment providers should consider and enable their patients to become aware of the functional relationship between their substance use and employment outcomes.

Participants who were usually unemployed were less likely to obtain employment. Prior research has shown that therapeutic workplace participants with histories of unemployment had limited educational credentials and skills that may be needed to obtain gainful employment (Holtyn et al., 2015; Silverman et al., 1995). This finding could have implications for selecting approaches to employment interventions. Education-focused interventions that seek to establish relevant job skills may be needed to promote employment in participants who report relatively long histories of unemployment. In contrast, a quick-entry approach, which seeks to promote employment as quickly as possible, may be better suited for recently employed individuals. These clients may have work-related skills that can be recovered, updated, or refined once employed. Nevertheless, how participants report their recent employment or unemployment could be used to select appropriate immediate employment goals and intervention approaches (i.e., education focused or quick entry).

Identification of patient characteristics that are associated with obtaining employment could be used to improve treatment effectiveness. However, potential conclusions from the data in this study are limited given that the observed relations were descriptive associations, as opposed to experimentally controlled variables. We cannot know for certain whether the variables identified in this study make participants more likely to gain employment, or whether the variables are simply associated with some unknown factor that influenced employability (e.g., geographical location and job availability; access to reliable transportation and childcare services; criminal justice involvement; stigma associated with opioid use disorder). Nevertheless, some of these factors could be systematically controlled in future studies, which would allow for an experimental analysis of these associations.

Highlights.

  • Patient characteristics may predict employment outcomes and help inform treatment

  • Patients who obtained employment worked more days in a therapeutic workplace

  • Illicit opioid and cocaine use were associated with worse employment outcomes

  • Patients who were usually unemployed were less likely to obtain employment

Acknowledgements

This research was supported by the National Institute on Drug Abuse of the National Institutes of Health under grants R01 DA037314 and T32 DA07209. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The National Institutes of Health had no part in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; and in the decision to submit the article for publication.

We are grateful to our staff who helped to conduct this study: Jacqueline Hampton, Andrew Rodewald, Sarah Pollock, Meghan Arellano, India Harper, and Calvin Jackson.

Footnotes

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

Conflict of Interest. The authors declare that there are no conflicts of interest.

References

  1. Bambra C 2011. Work, worklessness and the political economy of health inequalities. J. Epidemiol. Community. Health. 65, 746–750. [DOI] [PubMed] [Google Scholar]
  2. Brewer DD, Catalano RF, Haggerty K, Gainey RR, Fleming CB, 1998. A meta-analysis of predictors of continued drug use during and after treatment for opiate addiction. Addiction. 93(1), 73–92. [PubMed] [Google Scholar]
  3. Donlin WD, Knealing TW, Needham M, Wong CJ, Silverman K, 2008. Attendance rates in a workplace predict subsequent outcome of employment-based reinforcement of cocaine abstinence in methadone patients. J. Appl. Behav. Anal. 41(4), 499–516. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Earnshaw V, Smith L, Copenhaver M, 2013. Drug addiction stigma in the context of methadone maintenance therapy: an investigation into understudied sources of stigma. Int. J. Ment. Health. Addict. 11(1), 110–122. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Eddie D, Vilsaint CL, Hoffman LA, Bergman BG, Kelly JF, Hoeppner BB, 2020. From working on recovery to working in recovery: Employment status among a nationally representative US sample of individuals who have resolved a significant alcohol or other drug problem. J. Sub. Abuse. Treat. 108000. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Ginexi EM, Foss MA, Scott CK, 2003. Transitions from treatment to work: Employment patterns following publicly funded substance abuse treatment. J. Drug. Issues. 33(2), 497–518. [Google Scholar]
  7. Henkel D, 2011. Unemployment and substance use: a review of the literature (1990-2010). Curr. Drug. Abuse. Rev. 4(1), 4–27. [DOI] [PubMed] [Google Scholar]
  8. Holtyn AF, DeFulio A, Silverman K, 2015. Academic skills of chronically unemployed drug-addicted adults. J. Vocat. Rehabil. 42(1), 67–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Holtyn AF, Donlin Washington W, Knealing TW, Wong CJ, Kolodner K, Silverman K, 2016. Behavioral factors predicting response to employment-based reinforcement of cocaine abstinence in methadone patients. Transl. Issues. Psychol. Sci. 2(2), 192–202. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Holtyn AF, Jarvis BP, Silverman K, 2017. Behavior analysts in the war on poverty: A review of the use of financial incentives to promote education and employment. J. Exp. Anal. Behav. 107(1), 9–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Holtyn AF, Toegel F, Subramaniam S, Jarvis BP, Leoutsakos J-M, Fingerhood M, Silverman K, 2020. Abstinence-contingent wage supplements to promote drug abstinence and employment: a randomised controlled trial. J. Epidemiol. Community. Health. 74(5), 445–452. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Hubbard RL, Craddock SG, Anderson J, 2003. Overview of 5-year followup outcomes in the drug abuse treatment outcome studies (DATOS). J. Sub. Abuse. Treat. 25(3), 125–134. [DOI] [PubMed] [Google Scholar]
  13. Magura S, Marshall T (2020). The Effectiveness of Interventions Intended to Improve Employment Outcomes for Persons with Substance Use Disorder: An Updated Systematic Review. Sub. Use. Misuse. 55(13), 2230–2236. [DOI] [PubMed] [Google Scholar]
  14. McLellan AT, Luborsky L, Cacciola J, Griffith J, Evans F, Barr HL, O’Brien CP, 1985. New data from the Addiction Severity Index: reliability and validity in three centers. J. Nerv. Ment. 173(7), 412–423. [DOI] [PubMed] [Google Scholar]
  15. Miguel AQ, Kiluk BD, Roos CR, Babuscio TA, Nich C, Mari JJ, Carroll KM, 2019. Change in employment status and cocaine use treatment outcomes: A secondary analysis across six clinical trials. J. Sub. Abuse. Treat. 106, 89–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. O’Connor AM, Cousins G, Durand L, Barry J, Boland F, 2020. Retention of patients in opioid substitution treatment: A systematic review. PloS.One. 15(5), e0232086. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Richardson L, Epp S 2016. Substance use disorders, employment and the return to work. In Handbook of return to work (pp. 667–692). Springer, Boston, MA. [Google Scholar]
  18. Silverman K, 2004. Exploring the limits and utility of operant conditioning in the treatment of drug addiction. Behav. Anal. 27(2), 209–230. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Silverman K, Holtyn AF, Jarvis BP, 2016a. A potential role of anti-poverty programs in health promotion. Prev. Med. 92, 58–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Silverman K, Holtyn AF, Morrison R, 2016b. The therapeutic utility of employment in treating drug addiction: Science to application. Transl. Issues. Psychol. Sci. 2(2), 203. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Silverman K, Holtyn AF, Subramaniam S, 2018. Behavior analysts in the war on poverty: Developing an operant antipoverty program. Exp. Clin. Psychopharm. 26(6), 515. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Silverman K, Holtyn AF, Toegel F, 2019. The utility of operant conditioning to address poverty and drug addiction. Perspect. Behav. Sci. 42(3), 525–546. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Toegel F, Holtyn AF, Subramaniam S, Silverman K, 2020. Effects of time-based administration of abstinence reinforcement targeting opiate and cocaine use. J. Appl. Behav. Anal. 53, 1726–1741. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Woods JS, Joseph H 2015. Stigma from the viewpoint of the patient. J. Addict. Dis. 34, 238–247. [DOI] [PubMed] [Google Scholar]

RESOURCES