Abstract
Background:
Unemployment is a chronic problem among treatment seeking substance users and is associated with poor treatment response. Most studies that have examined the relationship between employment and treatment outcomes for substance use disorders have done so by considering employment at only one specific point in time (e.g., upon entering treatment). There is a lack of research on how change in employment status over time is associated with substance use treatment outcomes. The aim of this study was to evaluate both static employment status and change in employment status over time as predictors of cocaine use treatment outcomes.
Methods
We utilized data pooled from six randomized clinical trials evaluating treatment for cocaine use disorders (n=553). Multiple general linear mixed models were conducted to determine the association of baseline, end-of-treatment, and change in employment status (from baseline to end-of-treatment) with treatment outcomes.
Results:
Treatment outcomes did not differ by baseline employment status but were significantly better for those employed versus unemployed at the end-of-treatment. In regard to change in employment status over time, those who were unemployed at baseline and acquired employment by end-of-treatment had significantly better treatment outcomes during active treatment and follow-up, as compared to those who were unemployed at baseline and remained unemployed by end-of-treatment.
Conclusion:
Our findings suggest that end-of-treatment employment status may be an important marker of good outcome among those unemployed at treatment entry and support the incorporation of interventions designed to promote employment by substance use disorders treatment programs.
Keywords: Randomized clinical trial, Cocaine use disorders, treatment outcomes, employment, psychosocial intervention, pharmacological intervention
1. INTRODUCTION
Unemployment is an important factor to consider in the development, maintenance, and recovery from substance use disorders (SUD). For example, the development or worsening of SUDs increases the risk of job loss and the maintenance of unemployment status (Baldwin, Marcus, & De Simone, 2010; De Simone, 2002; Dooley, Catalano, & Hough, 1992), while job loss/unemployment increases the risk of development or worsening of SUDs (Claussen, 1999; Fergusson, Horwood, & Lynskey, 1997; Green, Doherty, Reisinger, Chilcoat, & Ensminger, 2010; Hammer, 1992).
Not surprisingly, rates of employment are extremely low among treatment seeking substance users (Platt, 1995; Sterling et al., 2001), with recent data from the U.S. Drug and Alcohol Services Information System (DASIS) pointing that only a third (31%) of substance users were employed at the time of treatment entry (Substance Abuse and Mental Health Services Administration, 2008). Furthermore, evidence shows that the low rates of employment among substance users tend to persist long after treatment completion, with one multi-site study demonstrating that rates of employment at one and five-year follow-ups were 43% and 54% respectively (Hubbard, Craddock, & Anderson, 2003).
This is particularly worrisome since employment is a consistent predictor of treatment success for persons with SUD and has been consistently associated with retention in treatment, increased abstinence, lower rates of relapse, less probation violations and less involvement in criminal activities across different treatment modalities (Arria, 2003; Jenner, Lennox, Hargrave, Lennings, & Andrew, 1998; Platt, 1995; Scorzelli, 2007; Sterling et al., 2001; Substance Abuse and Mental Health Services Administration, 2008). Furthermore, employment is not only considered an important outcome among clinicians and researchers (A. T. McLellan, Lewis, O’Brien, & Kleber, 2000; A.T. McLellan et al., 1996) but is also consistently cited as a top priority by substance users in all stages of recovery (A.B. Laudet, Magura, Vogel, & Knight, 2002).
Although employment is commonly used to evaluate treatment success and is known to be an important predictor of treatment response, most studies that have looked at employment during SUD treatments (Arria, 2003; Jenner et al., 1998; Platt, 1995; Scorzelli, 2007; Sterling et al., 2001; Substance Abuse and Mental Health Services Administration, 2008) have done so by determining employment status at single time-points (e.g. baseline, end-of-treatment, follow-ups). However, employment status may fluctuate over time and transitions in employment status may be a salient factor in the recovery process. To our knowledge, no studies have explored the association between changes in employment status during the course of treatment and treatment outcomes, both short- and long-term.
To better understand the relationship between employment and treatment outcomes, this study had three main objectives. First, we evaluated the relation between baseline employment status and several demographic and clinical variables among individuals participating in clinical trials evaluating treatment for cocaine use disorder (CUD). Second, we evaluated how baseline and end-of-treatment employment status separately predicted cocaine treatment outcomes. Third, we evaluated how change in employment status during the course of treatment predicted cocaine treatment outcomes.
Based on previous evidence, we hypothesized that those employed at baseline and end-of-treatment would have better treatment outcomes when compared to those unemployed at baseline and end-of-treatment, respectively. Additionally, with regard to the change in employment status, we hypothesized that participants who remained employed or transitioned to employed status at the end-of-treatment would achieve better outcomes when compared to participants who remained unemployed or transitioned to unemployed status at end-of-treatment.
2. MATERIAL AND METHODS
2.1. Participant, study design and measures
Data for these analyses was composed of an aggregated sample pooled from six randomized controlled trials (RCT) designed to evaluate pharmacological and/or behavioral outpatient treatments for CUD. All six trials had identical inclusion/exclusion criteria: participants were 18 years or older, seeking outpatient treatment for cocaine use who met DSM-IV criteria for current cocaine abuse or dependence (Diagnostic and Statistical Manual of Mental Disorders, fourth edition, American Psychiatric Association, 1994), and were psychiatrically stable for outpatient treatment. All trials had a common assessment battery that included the Structured Clinical Interview for DSM-IV (First, 1997), the Addiction Severity Index (ASI) (A. T. McLellan et al., 1992), self-reported substance use through Timeline Follow-back method (Robinson, Sobell, Sobell, & Leo, 2014; Sobell & Sobell, 1992), urine toxicology screens at each research visit and collection schedule (pre-treatment; weekly during treatment, end-of-treatment and 1, 3 and 6-month follow-ups), with similar treatment duration (2 of the 6 trials included an 8-week active treatment period while all other trials were 12 weeks). Primary outcome results and CONSORT diagrams for each study have been published (Carroll et al., 2008; Carroll, Kiluk, Nich, Gordon, et al., 2014; Carroll, Nich, DeVito, Shi, & Sofuoglu, 2018; Carroll et al., 2016; Carroll, Nich, Shi, Eagan, & Ball, 2012). For an overview of all six trials, see Table 1.
Table 1.
Overview of the Clinical Trials included in the dataset
| Study | Behavioral treatment conditions | Pharmacological treatment conditions | Treatment duration | Baseline n (%) | End of Treatment n (%) | 1-month follow-up n (%) | 3-month follow-up n (%) | 6-month follow-up n (%) | Primary outcomes citation |
|---|---|---|---|---|---|---|---|---|---|
| 1 | CBT vs IPT | Disulfiram vs placebo | 12 weeks | 119 (100) | 103 (86.5) | 107 (89.9) | 105 (88.2) | 103 (86.5) | Carroll et al., 2004 |
| 2 | CBT4CBT + TAU vs TAU | - | 8 weeks | 39 (100) | 31 (79.5) | 32 (82) | 28 (71.8) | 28 (71.8) | Carroll et al., 2008 |
| 3 | TSF + TAU vs TAU | Disulfiram vs placebo | 12 weeks | 106 (100) | 98 (92.4) | 100 (94.3) | 100 (94.3) | 99 (93.4) | Carroll et al., 2012 |
| 4 | CBT4CBT + TAU vs TAU | - | 8 weeks | 80 (100) | 76 (95) | 77 (96.2) | 76 (95) | 73 (91.2) | Carroll et al., 2014 |
| 5 | CM + CBT vs CBT | Disulfiram vs placebo | 12 weeks | 94 (100) | 81 (86.2) | 81 (86.2) | 81 (86.2) | 78 (83) | Carroll et al., 2016 |
| 6 | CBT4CBT + TAU vs TAU | Galantamine vs placebo | 12 weeks | 115 (100) | 111 (96.5) | 112 (97.4) | 112 (97.4) | 111 (96.5) | Carroll et al., 2018 |
| Total | - | - | - | 553 (100) | 500 (90.4) | 509 (92) | 502 (90.8) | 492 (89) | - |
CBT = Cognitive behavioral therapy; IPT = interpersonal therapy adapted for drug abuse; CBT4CBT = Computer based training for cognitive behavioral therapy; TAU = Treatment as usual; TSF = Twelve-step facilitation; CM = Contingency Management
For the purposes of this study, only participants reporting being half-time employed, full-time employed or unemployed on the ASI baseline assessment were included in these analyses. In order to dichotomize the employment variable, half and full-time employment were computed as ‘employed’. Participants classified as persons with disabilities, students, homemakers, inmates or retirees at baseline were not included (n = 45; 7.5%). As a result, of the 598 subjects included in the six original trials, 553 (92.5%) were included for these analyses. Additionally, of these 553 individuals, 500 were assessed at end-of-treatment (90.4%), with 466 individuals (84.3%) reporting being employed or unemployed at the end- of-treatment. As a result, only these 466 individuals were considered for the analyses involving end-of-treatment employment status and change in employment status (from baseline to end-of-treatment).
The cocaine use outcome measures used in this study were selected for having previously shown to be sensitive to treatment effects and predictive of long-term functional improvements in previous studies conducted with this same dataset (Carroll, Kiluk, Nich, DeVito, et al., 2014; Kiluk et al., 2016; Miguel et al., 2019). These outcome measures were: maximum days of consecutive abstinence; percentage of days abstinent; percentage of urine specimens negative for cocaine; retained and abstinent at the last week of treatment; and 3 or more weeks of continuous abstinence. Despite evidence of its insensitivity to treatment effects, being completely abstinent during treatment was also included as an outcome measure since it is still widely considered as one of the most desirable treatment outcomes (Ciraulo, Piechniczek-Buczek, & Iscan, 2003; Donovan et al., 2012). The maximum days of consecutive abstinence; percentage of days abstinent; and 3 or more weeks of continuous abstinence outcome measures were computed using self-reported cocaine use via the Timeline Follow-back method (Robinson et al., 2014; Sobell & Sobell, 1992). The percentage of urine specimens negative for cocaine outcome was based exclusively on the urine toxicology results. The retained and abstinent at the last week of treatment outcome was computed using the Timeline Follow-back method and the completion of a study visit during the final week of treatment. Finally, the complete abstinence during treatment outcome was computed using the Timeline Follow-back method and zero cocaine-positive urinalysis during treatment.
2.2. Statistical Analyses
In order to compare profile and clinical characteristics among those entering treatment employed or unemployed and evaluate the impact of baseline employment status on treatment outcome responses, multiple Generalized Linear Mixed Models (GLMM) were conducted assigning baseline characteristics and treatment outcome responses collected during treatment and follow-up separately as the dependent variables (targets). For all models, the dichotomous baseline employment status variable was assigned as the independent variable (primary predictor of interest). Similarly, multiple GLMMs were conducted to determine the impact of end-of-treatment employment status on treatment outcome responses assigning each outcome measure separately as the dependent variables and the dichotomous end-of-treatment employment status as the independent variable.
To evaluate the association between change in employment status and treatment outcomes, four different groups were created: unemployed at baseline and at the end-of-treatment (UU); unemployed at baseline and employed at the end-of-treatment (UE); employed at baseline and unemployed at the end-of-treatment (EU); and employed at baseline and at the end-of-treatment (EE). Multiple GLMMs were then conducted where each outcome measure was assigned separately as the dependent variable and the change of employment status assigned as the independent variable. For these analyses, pairwise comparisons were conducted with all 6 possible contrasts (UU vs UE; EE vs EU; and UE vs EE; EU vs UE; UU vs EU; and UU vs EE).
Because the six trials included in this dataset were conducted in different time periods (2004 to 2018), in different treatment settings (e.g. outpatient, methadone clinic, etc.), with different interventions (e.g. CBT, CM, disulfiram, etc.) and different treatment lengths (8-12 weeks), a study protocol variable was included as a random intercept in all GLMM analyses. In order to control for multiple comparisons but still be able to identify all potential subgroup differences (minimizing type II error) the significance level at .01. All statistical analyses were performed with SPSS Statistics software package, version 24.0 (IBM Corporation, Armonk, NY).
3. RESULTS
3.1. Baseline demographics and treatment outcomes by baseline employment status
As can be seen in Table 2, most baseline demographic and clinical features did not differ by baseline employment status when protocol was accounted for. Participants who were employed at treatment entry reported trends towards fewer years of regular cocaine use (t = −2.01; p =.045) and higher odds of submitting a cocaine positive urinalysis at baseline (OR = 1.61; p = .047) compared to those entering treatment unemployed. Those employed at baseline were half as likely to have ever been arrested (OR = .53; p = .002) but had higher odds of being currently on probation/parole (OR = 1.75; p = .007) or having been referred to treatment by criminal justice system (OR = 2.7; p = .002). As expected, the mean ASI employment composite score was significantly lower among those entering treatment employed compared to those entering treatment unemployed (t = −3.4; p = .001).
Table 2:
Baseline demographics by employment status.
| Unemployed n = 342 (61.2%) |
Employed n= 211 (38.2%) |
Estimate/SE | t | OR | P. | |
|---|---|---|---|---|---|---|
| Demographics | ||||||
| Men (N (%)) | 208 (60.8) | 140 (66.4) | - | - | - | ns |
| Age (Mean ± SD) | 38.8 ± 8.6 | 37.1 ± 8.4 | - | - | - | ns |
| Never married/living alone (N (%)) | 252 (73.7) | 143 (67.8) | - | - | - | ns |
| Completed high school (N (%)) | 267 (78.1) | 174 (82.5) | - | - | - | ns |
| Race/ethnic (N (%)) | - | - | - | ns | ||
| White | 180 (52.6) | 121 (57.3) | ||||
| African American | 108 (31.6) | 65 (30.8) | ||||
| Hispanic | 49 (14.3) | 23 (10.9) | ||||
| Others | 5 (1.5) | 2 (0.9) | ||||
| History and pattern of cocaine use | ||||||
| Age of cocaine onset (Mean ± SD) | 20.6 ± 5.6 | 21.3 ± 7 | - | - | - | ns |
| Years of regular cocaine use (Mean ± SD) | 10.3 ± 7.9 | 8.8 ± 7.7 | −1.42/0.7 | −2.01 | - | .045 |
| Days of cocaine use in 28 days prior to treatment (Mean ± SD) | 14.5 ± 8.6 | 14 ± 9.1 | - | - | - | ns |
| Pretreatment cocaine positive urine result (N (%)) | 249 (72.8) | 163 (77.2) | .48/0.24 | - | 1.61 | .047 |
| History of substance abuse treatment | ||||||
| Treatment naïve (N (%)) | 56 (16.4) | 33 (15.6) | - | - | - | ns |
| Number outpatient substance use treatments (Mean ± SD) | 2.2 ± 3.3 | 2.4 ± 3.4 | - | - | - | ns |
| Number inpatient substance use treatments (Mean ± SD) | 2.7 ± 4.3 | 2.9 ± 5.9 | - | - | - | ns |
| Concomitant Psychiatric disorders | ||||||
| Lifetime Alcohol use disorder (N (%)) | 215 (68) | 138 (68.3) | - | - | - | ns |
| Lifetime Major depression disorder (N (%)) | 61 (18) | 38 (18.3) | - | - | - | ns |
| Lifetime Anxiety disorder (N (%)) | 52 (15.4) | 22 (10.5) | - | - | - | ns |
| Lifetime Antisocial personality disorder (N (%)) | 54 (15.9) | 39 (18.7) | - | - | - | ns |
| History of legal problems | ||||||
| Ever Arrested (N (%)) | 134 (39.3) | 58 (27.5) | −.63/0.2 | - | 0.53 | .002 |
| Lifetime number of months incarcerated (Mean ± SD) | 26.3 ± 47.4 | 17.4 ± 42.3 | - | −1.22 | - | ns |
| Currently on probation/parole (N (%)) | 52 (16.4) | 51 (27) | .56/0.21 | - | 1.75 | .007 |
| Currently awaiting charges, trial or sentence (N (%)) | 37 (10.9) | 22 (10.4) | - | - | - | ns |
| Referred to treatment by criminal justice system (N (%)) | 26 (7.6) | 36 (17.1) | .98/0.31 | - | 2.7 | .002 |
| ASI composite scores | ||||||
| Cocaine (Mean ± SD) | .67 ± 0.22 | .66 ± 0.22 | - | - | - | ns |
| Other drags (Mean ± SD) | .05 ± 0.8 | .06 ± 0.08 | - | - | - | ns |
| Alcohol (Mean ± SD) | .11 ± 0.18 | .1 ± 0.17 | - | - | - | ns |
| Psychological (Mean ± SD) | .15 ± 0.18 | .16 ± 0.19 | - | - | - | ns |
| Medical (Mean ± SD) | .16 ± 0.28 | .13 ± 0.25 | - | - | - | ns |
| Employment (Mean ± SD) | .66 ± 0.29 | .56 ± 0.28 | −.09/0.02 | −3.4 | - | 0.001 |
| Family/social (Mean ± SD) | .16 ± 0.18 | .17 ± 0.18 | - | - | - | ns |
| Legal (Mean ± SD) | .09 ± 0.16 | .09 ± 0.17 | - | - | - | ns |
The unemployed group served as reference for the analyses
SE = Standard Error; OR = Odds Ratio; ns = statistically non-significant.
As observed in Table 3, baseline employment status was not associated with any cocaine use outcomes (collected within and after treatment). However, the odds of being employed at the end-of-treatment and at the one-, three- and six-month follow-ups were respectively 2.8, 2.38, 2.5 and 1.87 times higher among those initiating treatment employed than those unemployed (p = .001 for all).
Table 3:
Treatment outcomes by baseline employment status.
| Unemployed n = 342 (61.2%) |
Employed n= 211 (38.2%) |
Estimate/SE | t | OR | P. | |
|---|---|---|---|---|---|---|
| Clinical outcomes within treatment | ||||||
| Percentage of days abstinent during treatment (Mean ± SD) | 68.4 ± 28.1 | 71.3 ±25.3 | - | - | - | ns |
| Percentage cocaine-negative samples during treatment (Mean ± SD) | 29.7 ± 35.5 | 32.5 ± 37.6 | - | - | - | ns |
| Achieved 3 or more weeks of consecutive cocaine abstinence (N (%)) | 117 (35.6) | 62 (31.2) | - | - | - | ns |
| Longest duration abstinence during treatment (days) (Mean ± SD) | 18.7 ± 22.3 | 18.5 ± 21.1 | - | - | - | ns |
| Retained treatment and abstinence in last week of treatment (N (%)) | 83 (25.2) | 53 (27) | - | - | - | ns |
| Complete abstinence during treatment (N (%)) | 27 (8.4) | 25 (12.3) | - | - | - | ns |
| Employment at the end of treatment (N (%)) | 121 (41.7) | 122 (69.7) | 1.05/.21 | - | 2.8 | .001 |
| Clinical outcomes at follow-up | ||||||
| Percentage of days abstinent at follow-up month 1 (Mean ± SD) | 78.8 ± 28.7 | 81.7 ± 30 | - | - | - | ns |
| Employment at follow-up month 1 (N (%)) | 143 (50.4) | 134 (74) | .86/.22 | - | 2.38 | .001 |
| Percentage of days abstinent at follow-up month 3 (Mean ± SD) | 82 ± 28.3 | 82.8 ± 25.8 | - | - | - | ns |
| Employment at follow-up month 3 (N (%)) | 145 (51.4) | 136 (74.7) | .92/.21 | - | 2.5 | .001 |
| Percentage of days abstinent at follow-up month 6 (Mean ± SD) | 82.1 ± 28 | 81.1 ± 28.4 | - | - | - | ns |
| Employment at follow-up month 6 (N (%)) | 160 (57.6) | 131 (74) | .63/.22 | - | 1.87 | .001 |
The unemployed group served as reference for the analyses
SE = Standard Error; OR = Odds Ratio; ns = statistically non-significant.
3.2. Treatment outcomes by end-of-treatment employment status
As can be found in Table 4, within treatment cocaine use outcomes differed according to end-of-treatment employment status. Participants employed at the end-of-treatment reported a higher percentage of days abstinent from cocaine (t = 3.39; p = .001) and trends towards longer duration of abstinence (measured in days) (t = 2.19; p = .029) compared to those who were unemployed.
Table 4:
Treatment outcomes by end-of-treatment employment status.
| end-of-treatment Unemployed n = 223 (47.8%) |
end-of-treatment Employed n = 243 (52.2%) |
Estimate/SE | t | OR | P. | |
|---|---|---|---|---|---|---|
| Clinical outcomes within treatment | ||||||
| Percentage of days abstinent during treatment (Mean ± SD) | 64.4 ± 29.8 | 74.4 ± 23.3 | 8.69/2.56 | 3.39 | - | .001 |
| Percentage cocaine-negative samples during treatment (Mean ± SD) | 25.1 ± 33.4 | 33.9 ± 37.2 | - | - | - | ns |
| Achieved 3 or more weeks of consecutive cocaine abstinence (N (%)) | 68 (30.5) | 93 (38.3) | - | - | - | ns |
| Longest duration abstinence during treatment (days) (Mean ± SD) | 16.2 ± 19.9 | 22.6 ± 24.6 | 4.75/2.17 | 2.19 | - | .029 |
| Retained treatment and abstinence in last week of treatment (N (%)) | 55 (25.7) | 74 (31.6) | - | - | - | ns |
| Complete abstinence during treatment (N (%)) | 14 (6.5) | 26 (10.9) | - | - | - | ns |
| Clinical outcomes at follow-up | ||||||
| Percentage of days abstinent at follow-up month 1 (Mean ± SD) | 76.7 ± 29.8 | 82.3 ± 24.8 | 5.47/2.6 | 2.1 | - | .036 |
| Employment at follow-up month 1 (N (%)) | 43 (21.2) | 224 (93.3) | 3.86/31 | - | 47.68 | .001 |
| Percentage of days abstinent at follow-up month 3 (Mean ± SD) | 80.4 ± 30.2 | 83.3 ± 24.7 | - | - | - | ns |
| Employment at follow-up month 3 (N (%)) | 61 (30.3) | 207 (87.3) | 2.76 /.25 | - | 15.77 | .001 |
| Percentage of days abstinent at follow-up month 6 (Mean ± SD) | 80.9 ± 29.7 | 82.2 ± 26.5 | - | - | - | ns |
| Employment at follow-up month 6 (N (%)) | 90 (44.8) | 188 (83.2) | 1.75/.23 | - | 5.76 | .001 |
The unemployed group served as reference for the analyses
SE = Standard Error; OR = Odds Ratio; ns = statistically non-significant.
In terms of cocaine use outcomes during the follow-up period, participants who were employed at the end-of-treatment reported a trend toward higher percentage of days abstinent from cocaine at the one-month follow-up (t = 2.1; p = .036) compared to unemployed participants. However, there were no significant differences by end-of-treatment employment status for this outcome at the three- and six-month follow-up time points. Finally, the odds of being employed at the one-, three- and six-month follow-ups were respectively 47.68, 15.77, and 5.76 times higher among those employed at the end-of-treatment (p = .001 for all).
3.3. Treatment outcomes by change in employment status
The descriptives (means/percentages) in treatment outcomes according to the change in employment status from baseline to end-of-treatment are summarized in Table 5 while treatment outcome comparisons by change in employment status are presented in Table 6. In regard to the UU vs UE contrast, the UE group consistently had better treatment outcomes. Compared to participants in the UU, participants in the UE group reported a higher percentage of days abstinent from cocaine (t = 2.7; p = .007) and longer duration of continuous abstinence (t = 3.54; p = .001). Additionally, participants in the UE group showed trends towards greater odds of achieving 3 or more weeks of consecutive cocaine abstinence (OR = 1.86; p = .025) and a higher percentage of days abstinent from cocaine at the first-month follow-up (t = 2.23; p = .026). The odds ratio for being employed at the one-, three-and six-month follow-ups were respectively 42.4. ,14.2 and 6.77 in favor of the UE group (p = .001 for all).
Table 5:
Treatment outcomes descriptives by change in employment status.
| UU n = 169 (36.3%) |
EU n = 54 (11.6%) |
UE n = 121 (26%) |
EE n = 122 (26.2%) |
|
|---|---|---|---|---|
| Clinical outcomes within treatment | ||||
| Percentage of days abstinent during treatment (Mean ± SD) | 63 ± 29.6 | 68.4 ± 30.4 | 76.3 ± 24.3 | 72.7 ± 22.3 |
| Percentage cocaine-negative samples during treatment (Mean ± SD) | 21.3 ± 30.7 | 36.3 ± 38.6 | 38.2 ± 37.8 | 29.9 ± 36.3 |
| Achieved 3 or more weeks of consecutive cocaine abstinence (N (%)) | 48 (28.4) | 20 (37) | 56 (46.3) | 37 (30.3) |
| Longest duration abstinence during treatment (days) (Mean ± SD) | 14.6 ± 18.4 | 21.5 ± 23.5 | 27.1 ± 27.1 | 18.3 ± 21.1 |
| Retained and abstinence in last week of treatment (N (%)) | 37 (22.3) | 18 (37.5) | 42 (36.8) | 32 (26.7) |
| Complete abstinence during treatment (N (%)) | 7 (4.3) | 7 (13) | 14 (11.8) | 12 (10) |
| Clinical outcomes at follow-up | ||||
| Percentage of days abstinent at follow-up month 1 (Mean ± SD) | 75.2 ± 30.4 | 81.2 ± 27.6 | 83.1 ± 25.9 | 81.5 ± 23.7 |
| Employment at follow-up month 1 (N (%)) | 29 (18.8) | 14 (28.6) | 109 (91.6) | 115 (95) |
| Percentage of days abstinent at follow-up month 3 (Mean ± SD) | 79.2 ± 31.6 | 84 ± 25.3 | 85.3 ± 23.1 | 81.3 ± 26.1 |
| Employment at follow-up month 3 (N (%)) | 42 (27.6) | 19 (38.8) | 98 (84.5) | 109 (90.1) |
| Percentage of days abstinent at follow-up month 6 (Mean ± SD) | 80.2 ± 30.8 | 83.3 ± 26.4 | 83.5 ± 25.4 | 80.9 ± 27.6 |
| Employment at follow-up month 6 (N (%)) | 64 (41.6) | 26 (55.3) | 90 (83.3) | 98 (83.1) |
UU = unemployed at baseline and at the end-of-treatment; EU = employed at baseline and unemployed at the end-of-treatment;
UE = unemployed at baseline and employed the end-of-treatment; EE = and employed at baseline and at the end-of-treatment;
SE = Standard Error; OR = Odds Ratio.; ns = statistically non-significant.
Table 6:
Treatment outcomes by change in employment status.
| Contrast 1 UU vs UE | Contrast 2 EU vs EE | Contrast 3 EE vs UE | Contrast 4 EU vs UE | Contrast 5 UU vs EU | Contrast 6 UU vs EE | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| B/SE | OR/t | p. | B/SE. | OR/t | p. | B/SE. | OR/t | p. | B/SE. | OR/t | p. | B/SE | OR/t | p. | B/SE | OR/t | p. | |
| Clinical outcomes within treatment | ||||||||||||||||||
| Percentage of days abstinent during treatment (Mean ± SD) | 9.12/3.37 | 2.7 | .007 | - | - | - | - | - | ns | 8.7/4.4 | 2 | 0.48 | - | - | ns | 8.5/3.29 | 2.58 | .01 |
| Percentage cocaine-negative samples during treatment (Mean ± SD) | - | - | ns | - | - | - | - | - | ns | - | - | ns | - | - | ns | - | - | ns |
| Achieved 3 or more weeks of consecutive cocaine abstinence (N (%)) | .62/.27 | 1.86 | .025 | - | - | - | .65/.28 | 1.92 | .021 | - | - | ns | - | - | ns | - | - | ns |
| Longest duration abstinence during treatment (days) (Mean ± SD) | 9.94/2.81 | 3.54 | .001 | - | - | - | 7.63/2.91 | 2.62 | .009 | - | - | ns | - | - | ns | - | - | ns |
| Retained and abstinence in last week of treatment (N (%)) | - | - | ns | - | - | - | - | - | ns | - | - | ns | - | - | ns | - | - | ns |
| Complete abstinence during treatment (N (%)) | - | - | ns | - | - | - | - | - | ns | - | - | ns | - | - | ns | - | - | ns |
| Clinical outcomes at follow-up | ||||||||||||||||||
| Percentage of days abstinent at follow-up month 1 (Mean ± SD) | 7.51/3.36 | 2.23 | .026 | - | - | - | - | - | ns | - | - | ns | - | - | ns | - | - | ns |
| Employment at follow-up month 1 (N (%)) | 3.74/.39 | 42.4 | .001 | 3.89/.53 | 49.1 | .001 | - | - | ns | 3.32/.46 | 27.6 | .001 | - | - | ns | 4.32/.47 | 75.4 | .001 |
| Percentage of days abstinent at follow-up month 3 (Mean ± SD) | - | - | ns | - | - | ns | - | - | ns | - | - | - | - | - | ns | - | - | ns |
| Employment at follow-up month 3 (N (%)) | 2.65/.31 | 14.2 | .001 | 2.64/.42 | 14 | .001 | - | - | ns | 2.15/.39 | 8.59 | .001 | - | - | ns | 3.14/.35 | 23.1 | .001 |
| Percentage of days abstinent at follow-up month 6 (Mean ± SD) | - | - | ns | - | - | ns | - | - | ns | - | - | - | - | - | ns | - | - | ns |
| Employment at follow-up month 6 (N (%)) | 1.91/31 | 6.77 | .001 | 1.36/.38 | 3.89 | .001 | - | - | ns | 1.38/.39 | 3.98 | .001 | - | - | ns | 1.89/.3 | 6.63 | .001 |
UU = unemployed at baseline and at the end-of-treatment; EU = employed at baseline and unemployed at the end-of-treatment; UE = unemployed at baseline and employed the end-of-treatment; EE = and employed at baseline and at the end-of-treatment; SE = Standard Error; OR = Odds Ratio.; ns = statistically non-significant.
No significant group differences were observed for cocaine use outcomes collected within and following treatment for the EU vs EE. However, the odds of being employed at the one-, three- and six-month follow-ups were respectively 49.1, 14 and 3.89 times higher for the EE group compared to the EU (p = .001 for all).
In regard to the EE vs UE contrast, the UE group reported a longer duration of continuous abstinence (t = 2.62; p = .009) and a trend toward higher odds of achieving three or more weeks of consecutive abstinence during treatment (OR = 1.91; p = .021). No significant group differences were observed for the three follow-up employment status outcomes.
In regard to the EU vs UE contrast, we observed trends favoring the UE group towards a higher percentage of days abstinent from cocaine during treatment (t = 2; p = .048). Additionally, participants in the UE group had respectfully, 27.6, 8.59 and 3.9 higher odds of being employed at the first-, third- and sixth-month follow-ups (p. = .001 for all).
In regard to the UU vs EE contrast, the EE group presented a higher a percentage of days abstinent from cocaine during treatment (t = 2.58; p = .01). The odds of being employed at the one-, three- and six-month follow-ups were respectively 75.4.1, 23.1 and 6.63 times higher for the EE group compared to the UU (p = .001 for all). Finally, no significant outcome differences were observed for the UU vs EU contrast.
4. DISCUSSION
Unemployment is a major and persistent problem among treatment seeking substance users. This study extends previous research on the relationship between employment status and treatment outcomes for SUD by evaluating both static employment status and change in employment status during treatment as predictors of cocaine use treatment outcomes. Overall, the current study findings suggest that 1) cocaine use treatment outcomes may be suboptimal for individuals who remain unemployed during the course of treatment, 2) among individuals unemployed before treatment, acquiring a job during treatment may be a key factor in the recovery process, 3) employment status upon entering treatment is not a prognostic indicator of treatment outcome, and 4) end-of-treatment employment status and changes in employment status over time, particularly the transition from unemployment to employment, appears to be an important predictor of treatment outcomes.
In agreement with previous findings, we observed high rates of unemployment at treatment entry (61.2%) and at end-of-treatment (48.4%) (Platt, 1995; Sterling et al., 2001; Substance Abuse and Mental Health Services Administration, 2008). However, most baseline demographic characteristics and clinical variables, including gender, race, age, and education level, did not differ by baseline employment status. This was somewhat unexpected and diverges from previous studies that have found, among substance users, being female, of an ethnic minority, older, or having a lower education level to be associated with unemployment (A. B. Laudet, 2012; Oggins, Guydish, & Delucchi, 2001; Sigurdsson, Ring, O’Reilly, & Silverman, 2012). Likewise, the prevalence of comorbid psychiatric disorders was similar across employment groups in this study, differing from previous studies showing an association between psychiatric symptomatology and employment among individuals receiving treatment for SUD (A. B. Laudet, 2012; A.B. Laudet et al., 2002; Webster et al., 2007).
In addition, although we observed trends towards fewer years of cocaine use and a higher proportion of cocaine positive urine results at treatment intake among those employed at baseline, employment status at baseline was not significantly related to the pattern, frequency and severity of baseline cocaine use in this pooled dataset. These findings are consistent with recent studies that have found that employment was not related to the severity of substance use (Hogue, Dauber, Dasaro, & Morgenstern, 2010; A. B. Laudet, 2012).
In regard to history of legal problems, we found that participants who were unemployed at baseline had higher odds of having previously been arrested. This is consistent with previous evidence indicating unemployment as a major burden among individuals with a history of incarceration (Pager, Western, & Sugie, 2009; Sheely & Kneipp, 2015). In contrast however, we also found that the prevalence of individuals currently on probation/parole or referred to treatment by criminal justice system, was higher among those employed at baseline. This was unexpected, and while the reasons for this are unclear, we might speculate that for those on probation/parole, holding a working position may be mandatory to maintain a probation status. As such, these individuals could be particularly motivated to maintain employment. Furthermore, we observed a significant overlap between those currently on probation/parole and those referred to treatment by criminal justice system (data not shown).
In the current study, we observed a substantial level of change in employment status from baseline to end-of-treatment. For example, 41% of participants unemployed at baseline acquired a job by the end-of-treatment, while 30% of those who were employed at baseline were no longer employed by the end-of-treatment. This reflects how, in a relatively short period of time (an 8 or 12-week period), employment status can vary for this population.
In addition, those unemployed at baseline but employed at the end-of-treatment had better treatment outcomes than those who remained unemployed. Based on these findings, it is plausible to speculate that the reason treatment outcomes did not differ by baseline employment status is because of the positive outcomes achieved for the subsample of those unemployed at baseline but employed at end-of-treatment (i.e., treatment outcomes for this subsample contributed to the outcome values for the full sample of those unemployed at baseline). In any case, our findings do not support baseline employment status as a relevant predictor of treatment response.
In contrast, as expected, end-of-treatment employment status was consistently associated with better treatment outcomes. These findings are consistent with the existing literature (Arria, 2003; Jenner et al., 1998) and have two clear implications. First, it offers additional support of the association of end-of-treatment employment status on cocaine use outcomes. Second, and more importantly, it provides further evidence of the clinical relevance of end-of-treatment employment status supporting its adoption as a meaningful indicator of good treatment outcome.
Perhaps the most novel findings presented in this study involve the association between change in employment status and treatment outcomes. As expected, among those unemployed at baseline, acquiring a job during treatment was associated with better treatment outcomes. In fact, those who entered treatment unemployed but were able to secure a working position by the end-of-treatment not only achieved better treatment outcomes compared to those who remained unemployed throughout treatment, but also compared to those who were employed at baseline and end-of-treatment. On the other hand, different from our initial hypothesis, those who were employed during the course of treatment did not achieve better cocaine use outcomes compared to those who transitioned from unemployment to employment from baseline to end-of-treatment, or vice versa. Thus, it appears that gaining employment during treatment may be a more salient factor with respect to cocaine use outcomes than merely maintaining employment throughout treatment.
These findings have clear implications that should be considered by clinicians and health providers working and/or running SUD treatment programs. First, employment status should be included as a treatment outcome measure, being monitored through the entire length of treatment. More importantly, in order to provide more effective treatment (especially for substance users initiating treatment while unemployed), treatment programs for SUD should incorporate interventions/services that target unemployment.
4.1. Strengths and limitations
This study has several strengths. First, to our knowledge this was the first study to compare treatment outcomes by change in employment status during treatment. Secondly, all six RCTs used for the current analyses were conducted following rigorous methodological standards which included well recognized evidence-based psychosocial and pharmacological interventions and a range of validated and broadly used clinical outcomes. Third, the diversity of our sample, which included a substantial representation of women (37%) and ethnic minorities (45%) and the fact that these six trials were conducted decades apart and in different treatment settings may increase the generalizability of these findings. Finally, the high rates of end-of-treatment and follow-up assessment completion (ranging from 89% to 92%) increases the validity of these findings by reducing the risk of biased results due to missing data.
Some important limitations should also be noted. First, the pooled data used in these analyses was composed of treatment trials that included different psychosocial and pharmacological interventions. However, determining the impact of employment status on treatment outcomes for a specific intervention was not possible to study here, since looking at treatment outcomes for a specific intervention would lead to a substantial reduction of our sample size and, consequently our power to detect group differences. Therefore, it is important to consider that our current outcome findings comprise all of these interventions into one, and it is probable that the impact of employment status on treatment outcomes might differ depending on the treatment intervention. Second, regarding the methods used to create our cocaine use outcome measures, the percentage of cocaine-negative samples submitted during treatment outcome measure was created based exclusively on urine toxicological results. As so, the lack of association of this outcome and end-of-treatment employment status may be due to its vulnerability to missing data (Carroll, Kiluk, Nich, DeVito, et al., 2014; Donovan et al., 2012). In contrast, the associations seen for end-of-treatment employment status and the percentage of days abstinent during treatment and longest duration of abstinence during treatment outcomes (both created based exclusively on the Timeline Follow-back) may be due to possible bias related to the reliability of outcome bases solely on self-report (e.g., participants may have falsely denied cocaine use in the self-report). However, the low rates of discrepancy between self-report and toxicological results observed in the original studies (7-16% across trials) offer some support to the reliability of the self-reported cocaine use. Fourth, although our findings point to the association between employment status and change in employment status with treatment outcome responses, our analyses do not enable us to determine the nature of these associations. Thus, it remains unclear if acquiring a job promotes a better treatment response, if response to treatment promotes employment, or if both phenomena positively impact one another.
5. CONCLUSIONS
Despite the high rates of unemployment among treatment seeking substance users (Sterling et al., 2001; Substance Abuse and Mental Health Services Administration, 2008), the association of employment with other desirable treatment outcomes (Arria, 2003; Platt, 1995), the importance given to employment by in-treatment substance users (A.B. Laudet et al., 2002), and the clear clinical and economic benefits of facilitating patient’s return to the labor market (Shepard & Reif, 2004), research shows that most community treatment programs do not provide any form of employment counseling/service (Etheridge, Craddock, Dunteman, & Hubbard, 1995; Friedmann, D’Aunnuo, Jin, & Alexander, 2000), and many treatment providers believe that employment counseling should remain separate from SUD treatment programs (Young, 2000). While our data cannot resolve whether gaining employment in this sample was a factor in reducing their substance use, or whether reduced substance use helped them gain employment, the transition to employment may be an important marker of treatment response.
Our study offers additional evidence on the significant associations between employment and treatment response, which support the integration of interventions/services designed to promote employment into SUD treatment programs as a way to promote more effective treatment for SUD. Overall, further studies are warranted on the role of employment and transitions in employment in the addiction recovery process.
Highlights.
Unemployment is common among treatment seeking cocaine users.
Baseline employment status did not predict cocaine use outcomes in this sample.
End-of-treatment employment status predicted cocaine use outcomes.
Change in employment status during treatment predicted cocaine use outcomes.
4. Acknowledgments
We would like to acknowledge Tami Frankforter and Karen Hunkele for their invaluable help in assembling the dataset as well as participants, research staff and colleagues who participated in the original clinical trials.
1. Role of funding source
This study was supported by National Institute on Drug Abuse (NIDA) grants: R01 DA015969-09S1, P50-DA09241 and U10 DA015831 (Carroll, PI), R01DA030369-04S1 (Carroll/Paris, mPI), and R21DA041661 (Kiluk/Carroll, PI). Dr. Miguel and Dr. Mari were also supported by the Fundação de Amparo a Pesquisa de São Paulo (FAPESP) postdoctoral fellowship grants (2017/05371-8, 2017-22004-9) as part of the Research and Innovation grant for Prevention of Mental Disorders and abusive use of Alcohol and other Drugs, “Pesquisas e Inovações em Prevenção de Transtomos Mentais e Uso Abusivo de Álcool e Outras Drogas”, funded by the Division of Mental Health of the Brazilian Ministry of Health. NIDA and FAPESP had had no further role in study design; in the collection, analysis and interpretation of data; in the writing of the report; or in the decision to submit the paper for publication.
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 citable 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
This study included data from 2 CBT4CBT trials. Author Carroll is a member of CBT4CBT LLC, which makes CBT4CBT available to qualified clinical providers and organizations on a commercial basis. Dr. Carroll works with Yale University to manage any potential conflicts of interest. All other authors declare that they have no conflicts of interest.
6. REFERENCES
- Arria AM (2003). Drug treatment completion and post-discharge employment in the TOPPS-II Interstate Cooperative Study. Journal of Substance Abuse Treatment, 25, 9–18. [DOI] [PubMed] [Google Scholar]
- Baldwin ML, Marcus SS, & De Simone JD (2010). Job loss discrimination and former substance use disorders. Drug and Alcohol Dependence, 110, 1–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Carroll KM, Ball SA, Martino S, Nich C, Babuscio TA, Nuro KF, … Rounsaville BJ (2008). Computer-assisted delivery of cognitive-behavioral therapy for addiction: A randomized trial of CBT4CBT. American Journal of Psychiatry, 165(7), 881–888. doi: 10.1176/appi.ajp.2008.07111835 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Carroll KM, Kiluk BD, Nich C, DeVito EE, Decker S, LaPaglia D, … Ball SA, (2014). Toward empirical identification of a clinically meaningful indicator of treatment outcome: features of candidate indicators and evaluation of sensitivity to treatment effects and relationship to one year follow up cocaine use outcomes. Drug and Alcohol Dependence, 137, 3–19. doi: 10.4088/JCP.17m11669 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Carroll KM, Kiluk BD, Nich C, Gordon MA, Portnoy GA, Marino DR, & Ball SA (2014). Computer-Assisted Delivery of Cognitive-Behavioral Therapy: Efficacy and Durability of CBT4CBT Among Cocaine-Dependent Individuals Maintained on Methadone. American Journal of Psychiatry, 171, 436–444. doi: 10.1176/appi.ajp.2013.13070987 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Carroll KM, Nich C, DeVito EE, Shi JM, & Sofuoglu M (2018). Galantamine and Computerized Cognitive Behavioral Therapy for Cocaine Dependence: A Randomized Clinical Trial. Journal of Clinical Psychiatry, 79(1). doi: 10.4088/JCP.17m11669 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Carroll KM, Nich C, Petry NM, Eagan DA, Shi JM, & Ball SA (2016). A randomized factorial trial of disulfiram and contingency management to enhance cognitive behavioral therapy for cocaine dependence. Drug and Alcohol Dependence, 160, 135–142. doi: 10.1016/j.drugalcdep.2015.12.036 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Carroll KM, Nich C, Shi JM, Eagan D, & Ball SA (2012). Efficacy of disulfiram and Twelve Step Facilitation in cocaine-dependent individuals maintained on methadone: A randomized placebo-controlled trial. Drug and Alcohol Dependence, 126, 224–231. doi: 10.1016/j.drugalcdep.2012.05.019 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ciraulo DA, Piechniczek-Buczek J, & Iscan EN (2003). Outcome predictors in substance use disorders. Psychiatric Clinics of North America, 26, 381–409. doi: 10.1016/S0193-953x(02)00106-5 [DOI] [PubMed] [Google Scholar]
- Claussen B (1999). Alcohol disorders and re‐ employment in a 5‐ year follow‐ up of longterm unemployed. Addiction, 94, 133–138. [DOI] [PubMed] [Google Scholar]
- De Simone J (2002). Illegal drug use and employment. Journal of Labor Cconomics, 20, 952–977. [Google Scholar]
- Diagnostic and Statistical Manual of Mental Disorders, fourth edition (American Psychiatric Association, 1994). Washington, DC: APA Press. [Google Scholar]
- Donovan DM, Bigelow GF, Brigham GS, Carroll KM, Cohen AJ, Gardin JG, … & Marllat GA (2012). Primary outcome indices in illicit drug dependence treatment research: systematic approach to selection and measurement of drug use end-points in clinical trials. Addiction, 107, 694–708. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dooley D, Catalano R, & Hough R (1992). Unemployment and alcohol disorder in 1910 and 1990: Drift versus social causation. Journal of Occupational and Organizational Psychology, 65, 277–290. [Google Scholar]
- Etheridge RM, Craddock SG, Dunteman GH, & Hubbard RL (1995). Treatment services in two national studies of community-based drug abuse treatment programs. Journal of Substance Abuse Treatment, 7, 9–26. [DOI] [PubMed] [Google Scholar]
- Fergusson DM, Horwood LJ, & Lynskey MT (1997). The effects of unemployment on psychiatric illness during young adulthood. . Psychological Medicine, 27, 371–381. [DOI] [PubMed] [Google Scholar]
- First MB, Spitzer RL, Gibbon M, & Williams JB (1997). User’s guide for the Structured clinical interview for DSM-IV axis I disorders SCID-I: clinician version. Washington DC: American Psychiatric Press. [Google Scholar]
- Friedmann PD, D’Aunnuo TA, Jin L, & Alexander JA (2000). Medical and psychosocial services in drug abuse treatment: do stronger linkages promote client utilization? Health Service Research, 35, 443–465. [PMC free article] [PubMed] [Google Scholar]
- Green KM, Doherty EE, Reisinger HS, Chilcoat HD, & Ensminger M (2010). Social integration in young adulthood and the subsequent onset of substance use and disorders among a community population of urban African Americans. Addiction, 105, 484–493. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hammer T (1992). Unemployment and use of drug and alcohol among young people: a longitudinal study in the general population. British Journal of Addiction, 87, 1571–1581. [DOI] [PubMed] [Google Scholar]
- Hogue A, Dauber S, Dasaro C, & Morgenstern J (2010). Predictors of employment in substance-using male and female welfare recipients. Journal of Substance Abuse Treatment, 38, 108–118. doi: 10.1016/j.jsat.2009.09.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hubbard RL, Craddock SG, & Anderson J (2003). Overview of 5-year followup outcomes in the drug abuse treatment outcome studies (DATOS). Journal of Substance Abuse Treatment, 25, 125–134. [DOI] [PubMed] [Google Scholar]
- Jenner M, Lennox L, Hargrave R, Lennings CJ, & Andrew M (1998). Harm minimization outcomes for methadone recipients: role of employment. Journal of Substance Use, 3, 114–118. [Google Scholar]
- Kiluk BD, Carroll KM, Duhig A, Falk DE, Kampman K, Lai S, … Strain EC (2016). Measures of outcome for stimulant trials: ACTTION recommendations and research agenda. Drug and Alcohol Dependence, 158, 1–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Laudet AB (2012). Rate and predictors of employment among formerly polysubstance dependent urban individuals in recovery. Journal of Addictive Diseases, 31, 288–302. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Laudet AB, Magura S, Vogel HS, & Knight EL (2002). Interest in and obstacles to pursuing work among unemployed dually diagnosed individuals. Substance Use and Misuse, 37, 145–170. [DOI] [PMC free article] [PubMed] [Google Scholar]
- McLellan AT, Kushner H, Metzger D, Peters R, Smith I, Grissom G, … Argeriou M (1992). The Fifth Edition of the Addiction Severity Index. Journal of Substance Abuse Treatment, 9, 199–213. [DOI] [PubMed] [Google Scholar]
- McLellan AT, Lewis DC, O’Brien CP, & Kleber HD (2000). Drug dependence, a chronic medical illness: implications for treatment, insurance, and outcomes evaluation. Jama, 284, 1689–1695. [DOI] [PubMed] [Google Scholar]
- McLellan AT, Woody GE, Metzger D, McKay J, Durrell J, Alterman AI, & O’Brien CP (1996). Evaluating the effectiveness of addiction treatments: reasonable expectations, appropriate comparisons. Milbank Q. 1996; 74:51–85. Milbank Quarterly, 74, 51–85. [PubMed] [Google Scholar]
- Miguel AQ, Kiluk B, Babuscio T, Nich C, Mari JJ, & Carroll K (2019). Short and long-term improvements in psychiatric symptomatology to validate clinically meaningful treatment outcomes for cocaine use disorders. Drug and Alcohol Dependence. doi: 10.1016/j.drugalcdep.2019.01.046 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Oggins J, Guydish J, & Delucchi K (2001). Gender differences in income after substance abuse treatment. Journal of Substance Abuse Treatment, 20, 215–224. [DOI] [PubMed] [Google Scholar]
- Pager D, Western B, & Sugie N (2009). Sequencing Disadvantage: Barriers to Employment Facing Young Black and White Men with Criminal Records. Annals of American Academy of Political and Social Science, 623, 195–213. doi: 10.1177/0002716208330793 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Platt JJ (1995). Vocational rehabilitation of drug abusers. Psychoogical Bulletin, 117, 416–433. [DOI] [PubMed] [Google Scholar]
- Robinson SM, Sobell LC, Sobell MB, & Leo GI (2014). Reliability of the Timeline Followback for Cocaine, Cannabis, and Cigarette Use. Psychology of Addictive Behaviors, 28, 154–162. doi: 10.1037/a0030992 [DOI] [PubMed] [Google Scholar]
- Scorzelli JF (2007). Predicting sobriety from the employment status of dually diagnosed clients who are opiate dependent. Journal of Addictions & Offender Counseling, 28, 31–43. [Google Scholar]
- Sheely A, & Kneipp SM (2015). The Effects of Collateral Consequences of Criminal Involvement on Employment, Use of Temporary Assistance for Needy Families, and Health. Women Health, 55, 548–565. doi: 10.1080/03630242.2015.1022814 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shepard DS, & Reif S (2004). The value of vocational rehabilitation in substance user treatment: a cost-effectiveness framework. Substance Use and Misuse, 39, 2581–2609. [DOI] [PubMed] [Google Scholar]
- Sigurdsson SO, Ring BM, O’Reilly K, & Silverman K (2012). Barriers to employment among unemployed drug users: age predicts severity. The American Journal of Drug and Alcohol Abuse, 38, 580–587. doi: 10.3109/00952990.2011.643976 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sobell LC, & Sobell MB (1992). Timeline Follow-Back In: Litten RZ, Allen JP (eds) Measuring Alcohol Consumption. Totowa, NJ: Humana Press. [Google Scholar]
- Sterling RC, Gottheil E, Glassman SD, Weinstein SP, Serota RD, & Lundy A (2001). Correlates of employment: A cohort study. The American Journal of Drug and Alcohol Abuse, 27, 137–147. [DOI] [PubMed] [Google Scholar]
- SAMHSA. (2008). The DASIS Report: Employment Status and Substance Abuse Treatment Admissions 2006. Retrieved from Rockville, MD: [Google Scholar]
- Webster JM, Rosen PJ, McDonald HS, Staton-Tindall M, Garrity TF, & Leukefeld CG (2007). Mental health as a mediator of gender differences in employment barriers among drug abusers. The American Journal of Drug and Alcohol Abuse, 33, 259–265. [DOI] [PubMed] [Google Scholar]
- Young NK (2000). TIP 38: Integrating Substance Abuse Treatment and Vocational Services: Treatment Improvement Protocol (TIP) Series 38. US Department of Health and Human Services. [Google Scholar]