The Mediating Influence of Treatment Participation Rate on Post-Treatment Employment

Abstract

This study explored treatment participation rate as a mediator between individuals’ baseline alcohol use, drug use, and psychiatric issue levels and post-treatment employment status. The study sample included 106 unemployed or underemployed individuals with substance use disorders who were participants in an intensive drug abuse treatment program. Structural equation modeling was applied to examine relationships between study variables. The results showed that (a) the direct effect of individuals’ baseline alcohol use, drug use, and psychiatric issue severities reduced treatment participation rate; (b) the direct effect of treatment participation rate ameliorated post-treatment employment status; and (c) the indirect effect of individuals’ baseline alcohol use, drug use, and psychiatric issue severities on post-treatment employment status was mediated by treatment participation rate, which reduced the negative influence of baseline issue severity on employment. This significant mediating effect has both practical and theoretical importance in that it indicates the need for increased focus on treatment participation in practice and future research.

Keywords

treatment participation substance abuse employment career/vocational

For over three decades, researchers have sought to develop and evaluate comprehensive treatment services and strategies for individuals with substance use disorders (SUDs) to reduce their physical and psychological symptoms, improve employability, and increase independent living skills (Metsch & Pollack, 2005). In response to the Personal Responsibility and Work Opportunity Reconciliation Act of 1996, which imposed employment goals (e.g., work within 2 years) on those receiving public assistance, many states have expanded treatment services beyond their original SUD or psychiatric focus (De Alba, Samet, & Saitz, 2004; L. Richardson, Wood, Montaner, & Kerr, 2012). Studies related to diagnosis, screening, and treatment have indicated that these comprehensive services are beneficial not only in reducing consumers’ complex comorbid issues (e.g., alcohol and drug use and psychiatric issues) but also in increasing their employability (Finello & Poulsen, 2012; Halkitis, 2009; McLellan, Cacciola, Alterman, Rikoon, & Carise, 2006; Xie, Drake, McHugo, Xie, & Mohandas, 2010). According to the 2008 National Survey of Substance Abuse Treatment Services (N-SSATS), the number of rehabilitation centers providing comprehensive treatment increased 114% from 2005 to 2008, whereas the number of centers providing SUD or psychiatric treatment alone decreased approximately 5% and 15%, respectively (Substance Abuse and Psychiatric Services Administration, 2009).

Although this comprehensive treatment has been shown to be effective in improving consumers’ health, employability, and independent living skills (Barrowclough et al., 2010), such treatment requires long-term and intensive services. To explore the most strategic and efficient use of limited government funds, researchers have assessed the effectiveness of various services and determined the most important variables for measuring best treatment practices in the field (Humphreys & McLellan, 2011). Based on numerous studies, five critical variables (levels of alcohol use, drug use, and psychiatric issue severity; treatment participation rate; and post-treatment employment) have become benchmarks in the development of appropriate treatment goals and designing treatments for reducing consumers’ issues (Craig et al., 2008; Montgomery, Vaughn, Thompson, & Howard, 2013; Xie et al., 2010).

First, many treatment outcome assessments focus on changes in alcohol use, drug use, and psychiatric issue severities from baseline to a specified post-baseline time point (Connor, Pinquart, & Gamble, 2009; Curran et al., 2008). In exploring relationships between consumers’ health issues, researchers have found positive relationships between SUDs and psychiatric disorders, in that consumers with SUDs are likely to experience psychiatric issues (Feinstein, Richter, & Foster, 2012), and conversely, those with psychiatric issues are more likely to abuse substances (Connor et al., 2009). Therefore, individuals with severe levels of SUDs are at higher risk to be diagnosed with co-occurring disorders (Curran et al., 2008; Drake, O’Neal, & Wallach, 2008), and those with psychiatric issues are more likely to have SUDs (Addington & Addington, 2007; Sheidow, McCart, Zajac, & Davis, 2012).

Second, clinicians have observed a strong negative relationship between treatment participation rate and treatment drop-out rate, which has been shown to influence treatment outcomes (Hser, Evans, Huang, & Anglin, 2004; Neumann & Hare, 2008). Specifically, individuals who have poor treatment attendance are more likely to drop out of treatment than those with better attendance and less likely to have positive outcomes (Hatzenbuehler, Corbin, & Fromme, 2011). To achieve positive outcomes, a high rate of treatment participation is almost a prerequisite (Garnick, Lee, Horgan, Acevedo, & Washington Circle Public Sector Workgroup, 2009; Watson, 2005). To improve participation rate and positive treatment expectations, clinicians in substance use and psychiatric fields have developed and applied cognitive and behavioral strategies. For example, Jo and Kim (2013) suggested that reducing the consumers’ internal and external barriers and participating in treatment (e.g., low mobility, unstable housing, distrust of clinicians, and conflicts with employers) would help reduce a number of canceled, rescheduled, and missed appointments.

In spite of these initial efforts of clinicians, the treatment participation rate remains low for consumers with alcohol and substance use issues seeking employment. Because of this tendency, clinicians might become bias toward individuals with a high probability of dropping out of treatment due to their greater severity of symptoms (Neumann & Hare, 2008). This bias might also lead service providers to expect poor treatment participation and fewer positive outcomes for individuals with greater severity levels (Chisolm et al., 2013). Nevertheless, when strategies have been developed to improve participation rate, the consumers’ participation rates improve. This improved treatment outcome has been found across all issue severity levels. Specifically, research have found that strategies to increase treatment participation would influence consumer employment (Slesnick, Erdem, Collins, Bantchevska, & Katafizsz, 2011) and decrease of substance use (M. Richardson & Abraham, 2012) and psychiatric issue levels after treatment (M. Richardson & Abraham, 2012).

Third, individuals with higher levels of SUD and psychiatric issue severity have a tendency to be unemployed or underemployed to begin with (Biegel, Stevenson, Beimers, Ronis, & Boyle, 2010; Frounfelker, Wilkniss, Bond, Devitt, & Drake, 2011). Clearly, exploring employment status after treatment is a key indicator of treatment efficacy, and increasing the likelihood of employment is a primary goal of treatment for individuals with SUDs and psychiatric disorders (Frounfelker et al., 2011; Luchansky, Brown, Longhi, Stark, & Krupski, 2000). Researchers also have found that consumers who were employed after completing treatment showed more positive outcomes (e.g., improved health conditions, self-esteem, and quality of life) than those who were unemployed. Therefore, assessing employability or employment status is critical (McHugo, Drake, Xie, & Bond, 2012).

In conclusion, many studies exploring relationships between these five critical variables have shown evidence for (a) positive relationships between SUD and psychiatric issue severity levels (Latkin, Curry, Hua, & Davey, 2007), (b) positive relationship between treatment participation rate and treatment outcomes (Hatzenbuehler et al., 2011; Neumann & Hare, 2008), and (c) negative relationships between issue severity levels and employment status (Evans, Li, & Hser, 2009; Funn & Woodruff, 2011).

Hypotheses

Although researchers have examined relationships between these critical variables, they have not examined the mediating effect of treatment participation rate on employment status after treatment completion. Little is known about the comprehensive relationships between baseline levels of alcohol, drug use, and psychiatric issue severity; treatment participation rate; and post-treatment employment. When we consider that the average participation rate of those receiving substance use treatment is 40% (Tuten, Fitzsimons, Chisolm, Nuzzo, & Jones, 2012), the need to study the influence of treatment participation on outcomes is evident. The results of this study could increase the attention paid to treatment participation and aid researchers in understanding the comprehensive relationships between consumers’ issues, treatment participation, and outcomes. Thus, to develop effective treatment services and optimize treatment outcomes, studies that examine the complex relationships between these critical variables are required (Kemp, Harris, Vurel, & Sitharthan, 2007).

The aim of this study, then, was to determine whether treatment participation rate mediated the relationships between three baseline issue severities and post-treatment employment status, that is, whether some of the increased employment after treatment in consumers with higher baseline levels of alcohol use, drug use, and psychiatric issue severity could be attributed to a higher treatment participation rate. Support for this relationship came from prior studies of individuals with SUDs and psychiatric disorders, which suggested that (a) those with lower baseline substance use or psychiatric issue levels are more likely to show a higher treatment participation rate (Drapalski, Bennett, & Bellack, 2011; Evans et al., 2009; Neighbors, Barnett, Rohsenow, Colby, & Monti, 2010), (b) those with a higher treatment participation rate are more likely to be employed after treatment (Evans et al., 2009; Funn & Woodruff, 2011), and (c) those with lower baseline substance use or psychiatric issues are more likely to be employed (Bond, Drake, & Becker, 2012; Hogue, Dauber, Dasaro, & Morgenstern, 2010). Based on this research, then, we hypothesized that treatment participation rate would mediate the relationships between each of the three baseline issue severities and employment status at 210 days after intake into a long-term outpatient substance abuse program.

Method

Participants

This study used archival data from Project Working Recovery (PWR), a long-term substance abuse intensive outpatient program focused on vocational counseling services in a small city in a southern state. Between October 1, 2007, and July 30, 2010, a total of 313 consumers who met placement criteria for the American Society of Addiction Medicine’s intensive outpatient level of treatment enrolled in the program and consented to participate in evaluation activities. Eligibility criteria were (a) age of 18 years or older, (b) history of SUDs, and (c) lack of full-time employment.

At their entrance into the program, participants completed a demographic survey, a baseline PWR Evaluation Survey, and an intake interview with a member of the PWR staff (master’s and doctoral students of rehabilitation counseling program). The PWR Evaluation Survey, which consisted of an employment status question plus the 40 standard questions of the Addiction Severity Index, Fifth Edition (ASI-5), was administered at baseline and 210 days after baseline (see Atherton, 2011). PWR staff then developed and provided services based on individualized treatment plans that considered each participant’s internal and external characteristics and symptoms. Treatment plan goals were designed to decrease levels of severity of SUD and psychiatric issues and increase job readiness and employability. Based on their individualized plans, participants attended three 3-hr treatment blocks per week. All participants received treatment services integrating psychiatric, crisis contingency, disease management, relapse prevention, family counseling, and group support. In addition, they were provided with various vocational counseling interventions (e.g., assessment of employability, employment interest, job search strategies, and job interview preparation) designed to help them choose, get, and keep a job, with activities drawn from the workbook Working It Out (Thum, Briber, & Butler, 2000).

Of the 313 consumers who began the PWR, 106 completed the treatment and baseline and post-baseline PWR Evaluation Surveys, and were included in the study. People who did not complete the treatment tend to have significantly higher scores on alcohol use (difference in ASI–alcohol scores, t = 3.43, p = .04), drug use (difference in ASI–substance use scores, t = 3.53, p = .04), and psychiatric issue severities (difference in ASI–psychiatric issues scores, t = 4.63, p = .05). Of these 106 participants, 60 (56.6%) were male and 46 (43.4%) were female, ranging in age from 21 to 62 years (M = 41.1, SD = 11.8). Participants identified their ethnicity as follows: 55 (51.9%) African American, 45 (42.5%) Caucasian, and 6 (5.7%) Other. The distribution of the highest level of education completed was as follows: 66 (62.3%) high school degree or general education development (GED), 25 (23.6%) no high school degree, and 15 (14.1%) post-secondary degree. Of these last, 9 (8.5%) held an associate’s degree, 3 (2.8%) a bachelor’s degree, and 3 (2.8%) a graduate degree.

Measures

To examine the three hypotheses, we used the data from the PWR data set on the demographic characteristics of participants as well as on the five variables under study (alcohol use, drug use, psychiatric issue, treatment participation rate, employment status).

Clinical symptoms

As part of the PWR Evaluation Survey, participants completed the ASI-5, the most widely applied instrument for assessing various issue severities in the clinical research of SUDs (Cacciola, Alterman, Habing, & McLellan, 2011) and psychiatric disorders (McLellan et al., 2006). The ASI-5 measures severity levels of alcohol, drug, psychiatric, employment, family/social, legal, and medical issues over the previous 30-day period. At baseline (intake) and 210 days after baseline, participants in the PWR answered the ASI-5 questions on issue severity using a 5-point Likert-type scale of 0 (not at all) to 4 (extremely). The resulting raw scores were transformed into a composite score for each issue, ranging from 0 (not an issue) to 1 (severe issue). The ASI-5 has demonstrated Cronbach’s alpha coefficients from .87 to .92 (Cacciola et al., 2011). For the purposes of outcome measurement, the current study concerned itself only with the baseline and 210-day post-baseline ASI-5 composite scores for the domains of alcohol use (M = 0.23, SD = 0.09, Cronbach’s α = .87), drug use (M = 0.23, SD = 0.13, Cronbach’s α = .92), and psychiatric issues (M = 0.40, SD = 0.21, Cronbach’s α = .94).

Treatment participation

To calculate treatment participation rate, we tabulated each participant’s number of session attendances, cancellations, and no-shows. We calculated each person’s treatment participation rate, number of appointments kept / (number of appointment cancellations + number of no-shows + number of appointments kept).

Employment status

The PWR Evaluation Survey taken at baseline and post-treatment asked, “What is your employment status?” Study participants selected one of the following answers: (a) unemployed, (b) employed part-time (30 hr/week or less), or (c) employed full-time (more than 30 hr/week). For the analysis, answers were coded as 0 (unemployed) or 1 (employed either part-time or full-time).

Data Analysis

This study utilized a longitudinal design with archived data collected in PWR. To reduce a potential problem using the archived data (Whitley & Kite, 2013), we conducted various descriptive analyses to check coding and typing errors to ensure no unrealistic and impossible scores were included in the data analysis. With this procedure, we were able to increase reliability and validity of data analysis. The hypothesized model (see Figure 1) examined the relationships between the five observed variables (alcohol, drug use, psychiatric issue, treatment participation rate, and employment). Consistent with the literature, it was hypothesized that the five variables would be associated and correlations were specified between them. To explore multiple relationships between variables and confirm a particular model in collected data, a two-step approach was used. The first step was to identify associations between the variables, which were consistent with the literature. This study used Pearson correlation, and we found all variables are adequately correlated with each other.

Figure 1.

Standardized parameter estimates for the theoretical model.

The second step involved assessing the loadings of these variables within the hypothesized model, resulting in fitting all variables as the literature. Structural Equation Modeling (SEM) analyses were conducted using maximum likelihood parameter estimation, the AMOS 20.0. This model assumed that (a) consumer’ baseline issues influence treatment participation rate, (b) treatment participation rate influences employment status after completing treatment, and (c) consumer’ baseline issues influence the 210-day post-baseline employment status. Statistical model fit was assessed with (a) chi-square statistics, (b) the comparative fit index (CFI), (c) the normed fit index (NFI), and (d) the root mean squared error of approximation (RMSEA). Values less than 2 on the chi-square test indicate adequate model fit. The CFI and NFI indexes vary between 0 and 1 with values close to 1 indicating a well-fitting model. Values of RMSEA less than .05 are assumed to be a well-fitting model (Kline, 2004).

Results

Table 1 presents the ASI-5 correlations between the four study variables of alcohol, drug use, psychiatric issue, and treatment participation rate. The highest Spearman’s correlation was found between baseline drug use issues and treatment participation rate, −.58 (p < .01). Significant correlations were found between all three sets of variables. Results from the hypothesized model reveal that all path coefficients were significant (see Figure 1). The model represented an acceptable fit to the data (χ² = 2.22, df = 3, p = .53; χ²/df = .74; CFI = 1.00; NFI = .98; RMSEA = .01). The magnitude of standardized path coefficients from the three predictors (baseline levels of alcohol use, drug use, and psychiatric issue severity) and the mediating variable (treatment participation rate) to employment status ranged from −.31 to .45 (p < .001). In addition, this model showed incorporated pathways that examined reciprocal effects between consumers’ issues (alcohol and drug use, .48; alcohol use and psychiatric issue, .47; drug use and psychiatric issue, .28). Thus, each of the study’s three hypotheses was supported.

Table 1.

Correlations Between Consumer Issue Severities and Treatment Participation.

Variables	Baseline alcohol issues	Baseline drug issues	Baseline psychiatric issues	TPR
Baseline alcohol issues	—
Baseline drug issues	.48**	—
Baseline psychiatric issues	.28**	.47**	—
TPR	−.51**	−.58**	−.50**	—

Note. TPR = treatment participation rate.

p < .01.

Alcohol Use Severity

First, the mediating effect of treatment participation rate on the relationship between baseline alcohol use severity and employment status at 210 days was supported, p = .001. This finding derived from examinations of (a) the direct effect between baseline alcohol use and treatment participation rate, (b) the direct effect between treatment participation rate and 210-day employment status, and (c) the indirect effect between baseline alcohol use and 210-day employment status. First, baseline alcohol use influenced treatment participation negatively (standardized coefficient = −.28). That is, as baseline level of alcohol use increased, treatment participation rate decreased, which is consistent with several other studies (e.g., Hser et al., 2004; Lopez-Goni, Fernandez-Montalvo, & Arteaga, 2011; Schulte, Meier, Stirling, & Berry, 2010). Second, treatment participation rate influenced 210-day employment status positively (standardized coefficient = .45). As participants’ treatment participation rate increased, their likelihood of post-treatment employment also increased. This finding is also consistent with those of multiple studies (e.g., Evans et al., 2009; Jaffe, Du, Huang, & Hser, 2012). Third, baseline alcohol use issue severity influenced 210-day employment status (standardized coefficient = −.13), with treatment participation rate reducing the influence of baseline alcohol use severity on employment.

Drug Use Severity

The meditating effect of treatment participation rate on the relationship between baseline drug use severity and employment status at 210 days also was supported, p = .001. This finding derived from examination of (a) the direct effect between baseline drug use and treatment participation rate, (b) the direct effect between treatment participation rate and 210-day employment status, and (c) the indirect effect between baseline drug use and 210-day employment status. First, baseline drug use influenced treatment participation negatively (standardized coefficient = −.31). That is, as baseline level of drug use increased, treatment participation rate decreased, which is consistent with several other studies (e.g., Evans et al., 2009; Hser et al., 2004). Second, treatment participation rate influenced 210-day employment status positively (standardized coefficient = .45). As treatment participation rate increased, the likelihood of post-treatment employment also increased. This finding has also been found in other studies (e.g., Huebner & Cobbina, 2007; Jaffe et al., 2012). Third, baseline drug use issues influenced 210-day employment status indirectly (standardized coefficient = −.14), with treatment participation rate reducing the influence of baseline drug use severity on employment.

Psychiatric Issue Severity

Finally, the treatment participation rate was found to mediate the relationship between baseline psychiatric issue severity and employment status at 210 days at a significance level of .001. This finding derived from examination of (a) the direct effect between baseline psychiatric issue severity and treatment participation rate, (b) the direct effect between treatment participation rate and 210-day employment status, and (c) the indirect effect between baseline psychiatric issue severity and 210-day employment status. First, baseline psychiatric issues influenced treatment participation rate negatively (standardized coefficient = −.28). That is, as baseline psychiatric issue severity increased, treatment participation rate decreased, which is consistent with several other studies (Angelo et al., 2013; Hiller, Knight, & Simpson, 1999; Tsang, Fung, & Chung, 2010). Second, treatment participation rate influenced 210-day employment status positively (standardized coefficient = .45). As treatment participation rate increased, participants’ likelihood of post-treatment employment also increased. This result has also been reported in several studies (Evans et al., 2009; Huebner & Cobbina, 2007; Jaffe et al., 2012). Third, baseline psychiatric issues influenced 210-day employment status indirectly (standardized coefficient = −.12), with treatment participation rate reducing the influence of baseline psychiatric issue severity on employment.

Discussion

With the Project Work and Recovery clinic data, we developed a SEM to examine the relationship between alcohol use, drug use, and psychiatric issues and the mediating influence of treatment participation rate on the relationship between individuals’ baseline levels on each of three issue severities and post-treatment employment. As in previous studies by Biegel et al. (2010) and Rogers, Anthony, Cohen, and Davies (1997), we found the direct effect of psychiatric issue severity for people with SUDs to be a significant predictor of treatment participation and employment. Second, we found support for prior findings that treatment participation rate is associated with baseline alcohol use severity (Lopez-Goni et al., 2011; Odenwald & Semrau, 2013), drug use severity (Evans et al., 2009; Hser et al., 2004), and psychiatric issue severity (Angelo et al., 2013; Tsang et al., 2010). Third, our findings mirror those of previous studies showing that a lower treatment participation rate (i.e., frequent appointment cancellations and a high no-show rate) directly influenced post-treatment employment (Evans et al., 2009; McKay, 2009). Fourth, our hypothesized model contributes to existing research in its determination that treatment participation rate mediates the relationship between the baseline issue severity and post-treatment employment status (Huang, Evans, Hara, Weiss, & Hser, 2011; Jang, Wang, & Lin, 2014).

In the model, consumers’ alcohol use, drug use, and psychiatric issue severities influenced treatment participation and employment. In particular, study participants with diagnosed SUDs showed higher psychiatric issue levels on the ASI-5 compared with alcohol use and drug use issue levels, though no significant difference was found. We also found a positive relationship between alcohol use (p < .01), drug use (p < .01), and psychiatric issue levels (p < .01) in this study (see Table 1). This finding can be attributed to the tendency of individuals with SUDs to be at higher risk for co-occurring disorders (Drake et al., 2008; Feinstein et al., 2012) and the active screening for a history of SUDs among study participants. This finding also is consistent with earlier studies indicating that people with alcohol or drug issues are more likely to show higher co-occurring psychiatric issues due to accumulated exposures in their lives (e.g., Hasin, Fenton, Beseler, Park, & Wall, 2012; Redonnet, Chollet, Fombonne, Bowes, & Melchior, 2012; Thoma et al., 2011). Because of this greater possibility of co-occurring disorders, providers would do well to consider the SUDs of consumers as well as their psychiatric issues when developing and providing treatment services (Connor et al., 2009; Feinstein et al., 2012).

The model developed for this study illustrates a three-step cycle: First, the more severe an individual’s baseline issues, the lower his or her treatment participation rate is likely to be. Second, the lower an individual’s treatment participation rate, the less likely the individual has acquired appropriate skills for confronting his or her issues and increasing his or her employability. Third, when individuals have high severity levels on issues related to substance use and/or psychiatric issues and low work-related skills, their opportunities to enter the workforce decrease. Thus, high baseline issue severities can create a vicious cycle in the employment process that prevents individuals from entering the workforce. When treatment participation rate is addressed as a mediating factor that can be improved, however, even individuals with high severity levels for alcohol use, drug use, and/or psychiatric issue at baseline can have a greater likelihood of decreasing their issues and gaining employment.

In describing the three-step model constructed in this study, some important applications became apparent based on the model’s integration of two streams of previous research. The first includes studies showing that individuals’ baseline severities of alcohol use (Lopez-Goni et al., 2011), drug use (Neumann & Hare, 2008), and psychiatric issues (Kawakami et al., 2013) significantly predict lower treatment participation rates. The second involves findings that individuals who are less likely to participate in treatment regularly (e.g., those with more higher alcohol use severity at baseline, as studied by Campbell, Bond, & Drake, 2011; Rohde, Stice, & Gau, 2012) are more likely to have poor treatment outcomes in terms of high alcohol use severity and lower employment rates. Our results indicated that treatment participation rate acted as a powerful mediating factor in the analysis, making the model more distinct and improving our understanding of the relationships between the baseline issue severities of consumers and their treatment outcomes.

To improve treatment participation rate, cognitive and behavioral strategies have been developed and applied in the counseling setting (Sivikis, Silverman, Hang, Stitzer, & Keyser-Marcus, 2007). Although these strategies have been accepted, Watson (2005) suggested that more concrete guidelines should be developed by clinicians and agencies. For example, the clinicians should strive to build a trust relationship between clinicians and consumers to engage and retain consumers in a treatment program. Although it is difficult to define each feature of this counseling relationship, clinicians should be trained under common themes of unbiased trust, a firm partnership, and unconditional support. Watson emphasized that using verbal encouragement including the family in decision-making, learning various communication styles, and facilitating ease of access to services are critical to build counseling relationship.

Furthermore, agencies need to develop a more consumer-oriented system, such as a decrease in eligibility criteria and increase the amount of contact clinicians have per consumers. Dawson and Berry (2002) indicated that removing eligibility criteria increases the rate of service participation because the criteria may result in consumers reporting exaggerated problems. Also, agencies would include information about response quickly and frequent maintenance of contact in the service manual. Watson (2005) suggested that long delays mean consumers would miss out on services so clinicians would contact them within 24 to 48 hr of being referred. Katz et al. (2001) also indicated that weekly based contact is likely to show higher participation rates than monthly based contact. When consumers fail to keep an appointment or do not return a call, clinicians should attempt continuously at least 3 times to reestablish contact.

In addition, researchers and clinicians may wish to explore the influence of the consumers’ personality characteristics on treatment participation rate. For example, hardiness (e.g., Kobasa, 1979) is the composite of the attitudes of commitment, control, and challenge that provide the motivation to convert potentially negative circumstances into growth opportunities (cf. Maddi, Harvey, Khoshaba, Fazel, & Resurreccion, 2009). Researchers (e.g., Bartone, Hystad, Eid, & Brevik, 2012; Kobasa, 1979) have found that hardy and less hardy consumers have produced different outcomes despite confronting the same stressors. It seems likely that consumers strong in commitment would rely on themselves to find ways of turning treatment into activities that would be interesting and meaningful. This type of consumer would also choose to become involved and engaged rather than feel alienated. Those consumers high in control would maintain that, through personal effort and persistence, they could influence treatment outcomes rather than passively seeing themselves as victims of bad luck or unfavorable conditions. Finally, those consumers strong in challenge would believe that personal fulfillment is to be encountered through persistent growth in wisdom, knowledge, and skills learned through personal experience rather than passively choosing an approach seeking comfort, security, and following a monotonous program.

When addressing substance, alcohol use, and psychological issues, successful consumers’ treatment participation might have been motivated by hardiness (i.e., thoughts and emotions characterized by commitment, control, and challenge; cf. Bartone et al., 2012; Kobasa, 1979; Maddi, Wadhwa, & Haier, 1996). After treatment, these consumers may have been motivated by these same cognitive states to address employment issues. In our view, participation in treatment had two benefits. First, during treatment, consumers were given feedback and were able to evaluate their substance, alcohol use, and psychological issues to gain perspective on and understanding of these stressors. Using this information, these consumers improved their skills to create and execute decisive problem-solving action plans. Second, consumers who used the feedback obtained during treatment were more likely to engage employment-seeking strategies by emphasizing motivational self-perceptions of commitment, control, and challenge.

Specific to employment variables in our study, there are numerous links between hardiness and a range of important work factors including job performance (Maddi, 2006; Westman, 1990), job satisfaction (Luszczynska & Cieslak, 2005; McCalister, Dolbier, Webster, Mallon, & Steinhardt, 2006), leadership (Bartone, Eid, Johnsen, & Laberg, 2009; Johnsen, Bartone, & Nissestad, 2009), and intentions to leave (Law, 2005). More specifically, persisting consumers may have had a stronger future orientation than those who dropped out of treatment. Moreover, because of their tendency to look optimistically to the future while learning from the past, some may have been more focused on future opportunities and challenges than previous difficulties and temporary setbacks. Furthermore, when encountering new experiences, the hardy-resilient consumer would be tenacious because he or she was action-oriented, competent, and confident compared with those who had a low treatment participation rate.

Limitations

The design of this study presents some limitations. The first concerns the unknown reliability and validity of archival data (Whitley & Kite, 2013), which may include coding and typing errors. Specifically, with regard to more than 60% attrition rate, it represents that more than 200 people began but did not complete the treatment. Because of these issues, the findings of this study should be interpreted carefully. Second, the generalizability of study findings may be limited to some extent due to a quasi-experimental, one group pre–post research design used in gathering the original data. Although a randomized controlled trial has been the gold standard of clinical research, quasi-experimental methods are used increasingly to explore new ideas and serve as a preliminary step for later clinical trials (Kline, 2004). Third, adequate control of extraneous variables, which improves the internal validity and examines the influence of observed variables, is a key issue in developing appropriate models. Unless appropriate controls for the extraneous variables were introduced, the relationships between observed variables would be interpreted carefully because of multiple levels of unknown variables to influence observed variables (Street, 1995). Specifically, because this study was conducted in a working clinician environment, many of the extraneous variables were difficult to control. Therefore, conclusions and applications of these findings about consumers’ employment, the extent of effects of consumers’ issues, and participation rates on employment should be interpreted carefully. Fourth, although self-report instruments such as the ASI-5 are commonly used in research because of their ease of use, study participants may positively overstate treatment effectiveness (Wand, Chiffelle, O’Connell, McAuley, & Desouza, 2010). To reduce this potential concern, a greater variety of scales should be used to examine consumers’ issue levels from both the consumers’ and clinicians’ perspectives.

Conclusion

Despite possible limitations, the results of this study provide significant evidence supporting the importance of increased treatment participation to improve the likelihood of post-treatment employment. To mediate the effect of consumers’ high baseline issue severity on post-treatment employability, treatments should use strategies designed to increase individuals’ motivation to consistently participate in treatment (Jang et al., 2014; Rinaldi, Montibeller, & Perkins, 2011). Although we must readily accept the issue severities with which consumers enter treatment, we also must support and encourage treatment participation to mediate the adverse effect of baseline issue severities on employment outcomes.

Footnotes

Acknowledgements

The authors express their deepest gratitude to all of the consumers who participated in this study.

Authors’ Note

Project Working Recovery staff designed the study and data collection tools and coordinated data collection. Min Kim carried out the literature research and wrote the first draft of the manuscript. All authors have read and approved of submission of this article to Drug and Alcohol Dependence.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Funding for Project Working Recovery was provided by Kate B. Reynolds Charitable Trust grant; and the funding agency 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.

References

Addington

(2007). Patterns, predictors and impact of substance use in early psychosis: A longitudinal study. Acta Psychiatrica Scandinavica, 115, 304–309. doi:10.1111/j.1600-0447.2006.00900.x

Angelo

F. N.

McDonell

M. G.

Lewin

M. R.

Srebnik

Lowe

Roll

Ries

(2013). Predictors of stimulant abuse treatment outcomes in severely mentally ill outpatients. Drug and Alcohol Dependence, 131, 162–165. doi:10.1016/j.drugalcdep.2012.11.017

Atherton

W. L.

(2011). The impact of a vocational counseling based substance abuse intensive outpatient program upon work and well-being: A pilot study (Unpublished doctoral dissertation). East Carolina University, Greenville, NC.

Barrowclough

Haddock

Wykes

Beardmore

Conrod

Craig

. . . Tarrier

(2010). Integrated motivational interviewing and cognitive behavioural therapy for people with psychosis and comorbid substance misuse: Randomized controlled trial. British Medical Journal, 341, 1–12. doi:10.1136/bmj.c6325

Bartone

P. T.

Eid

Johnsen

B. H.

Laberg

J. C.

(2009). Big Five personality factors, hardiness, and social judgment as predictors of leader performance. Leadership & Organization Development Journal, 30, 498–521.

Bartone

P. T.

Hystad

S. W.

Eid

Brevik

J. I.

(2012). Psychological hardiness and coping style as risk/resilience factors for alcohol abuse. Military Medicine, 177, 517–522.

Biegel

D. E.

Stevenson

L. D.

Beimers

Ronis

R. J.

Boyle

(2010). Predictors of competitive employment among consumers with co-occurring mental and substance use disorders. Research on Social Work Practice, 20, 191–201.

Bond

G. R.

Drake

R. E.

Becker

D. R.

(2012). Generalizability of the individual placement and support model of supported employment outside the US. World Psychiatry, 11, 32–39. doi:10.1016/j.wpsyc.2012.01.005

Cacciola

J. S.

Alterman

A. I.

Habing

McLellan

(2011). Recent status scores for version 6 of the Addiction Severity Index (ASI-6). Addiction, 106, 1588–1602. doi:10.1111/j.1360-0443.2011.03482.x

10.

Campbell

Bond

G. R.

Drake

R. E.

(2011). Who benefits from supported employment: A meta-analytic study. Schizophrenia Bulletin, 37, 370–380. doi:10.1093/schbul/sbp066

11.

Chisolm

M. S.

Fitzsimons

Leoutsakos

Acquavita

S. P.

Heil

S. H.

Wilson-Murphy

. . . Hones

H. E.

(2013). A comparison of cigarette smoking profiles in opioid-dependent pregnant patients receiving methadone or buprenorphine. Nicotine & Tobacco Research, 15, 1297–1304. doi:10.1093/ntr/nts274

12.

Connor

K. R.

Pinquart

Gamble

S. A.

(2009). Meta-analysis of depression and substance use among individuals with alcohol use disorders. Journal of Substance Abuse Treatment, 37, 127–137. doi:10.1016/j.jsat.2008.11.007

13.

Craig

T. K.

Johnson

McCrone

Afuwape

Hughes

Gournay

. . . Thronicroft

(2008). Integrated care for co-occurring disorders: Psychiatric symptoms, social functioning, and service costs at 18 months. Psychiatric Services, 59, 276–282.

14.

Curran

G. M.

Sullivan

Williams

Han

Allee

Kotria

K. J.

(2008). The association of psychiatric comorbidity and use of the emergency department among persons with substance use disorders: An observational cohort study. BMC Emergency Medicine, 8, Article 17. doi:10.1186/1471-227X-8-17

15.

Dawson

Berry

(2002). Engaging Families in child welfare services: An evidence-based approach to best practice. Child Welfare, 81, 293–317.

16.

De Alba

Samet

J. H.

Saitz

. (2004). Burden of medical illness in drug- and alcohol-dependent persons without primary care. American Journal on Addictions, 13, 33–45. doi:10:1080/10550490490265307

17.

Drake

R. E.

O’Neal

E. L.

Wallach

M. A.

(2008). A systematic review of psychosocial research on psychosocial interventions for people with co-occurring severe mental and substance use disorders. Journal of Substance Abuse Treatment, 34, 123–138. doi:10.1016/j.jsat.2007.01.011

18.

Drapalski

Bennett

Bellack

(2011). Gender differences in substance use, consequences, motivation to change, and treatment seeking in people with serious mental illness. Substance Use & Misuse, 46, 808–818. doi:10.3109/10826084.2010.538460

19.

Evans

Hser

Y. I.

(2009). Client and program factors associated with dropout from court mandated drug treatment. Evaluation and Program Planning, 32, 204–212. doi:10.1016/j.evalprogplan.2008.12.003

20.

Feinstein

E. C.

Richter

Foster

S. E.

(2012). Addressing the critical health problem of adolescent substance use through health care, research, and public policy. Journal of Adolescent Health, 50, 431–436. doi:10.1016/j.jadohealth.2011.12.033

21.

Finello

K. M.

Poulsen

M. K.

(2012). Unique system of care issues and challenges in serving children under age 3 and their families. American Journal of Community Psychology, 49, 417–429. doi:10.1007/s10464-011-9458-6

22.

Frounfelker

R. L.

Wilkniss

S. M.

Bond

G. R.

Devitt

T. S.

Drake

R. E.

(2011). Enrollment in supported employment services for clients with a co-occurring disorder. Psychiatric Services, 62, 545–547. doi:10.1176/appi.ps.62.5.545

23.

Funn

Woodruff

S. I.

(2011). Prevalence and sociodemographic correlates of mental health problems among emergency department patients with high-risk and dependent alcohol use patterns. Journal of Dual Diagnosis, 7, 14–25. doi:10.1080/15504263.2011.569192

24.

Garnick

D. W.

Lee

M. T.

Horgan

C. M.

Acevedo

, Washington Circle Public Sector Workgroup. (2009). Adapting Washington circle performance measures for public sector substance abuse treatment systems. Journal of Substance Abuse Treatment, 36, 265–277. doi:10.1016/j.jsat.2008.06.008

25.

Halkitis

P. N.

(2009). Methamphetamine addiction: Biological foundations, psychological factors, and social consequences. Washington, DC: American Psychological Association. doi:10.1037/11883-000

26.

Hasin

D. S.

Fenton

M. C.

Beseler

Park

J. Y.

Wall

M. M.

(2012). Analyses related to the development of DSM-5 criteria for substance use related disorders: 2. Proposed DSM-5 criteria for alcohol, cannabis, cocaine and heroin disorders in 663 substance abuse patients. Drug and Alcohol Dependence, 122, 28–37.

27.

Hatzenbuehler

M. L.

Corbin

W. R.

Fromme

(2011). Discrimination and alcohol-related problems among college students: A prospective examination of mediating effects. Drug and Alcohol Dependence, 115, 213–220. doi:10.1016/j.drugalcdep.2010.11.002

28.

Hiller

M. L.

Knight

Simpson

D. D.

(1999). Risk factors that predict dropout from corrections-based treatment for drug abuse. The Prison Journal, 79, 411–430. doi:10.1177/0032885599079004004

29.

Hogue

Dauber

Dasaro

Morgenstern

(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

30.

Hser

Y. I.

Evans

Huang

Anglin

D. M.

(2004). Relationship between drug treatment services, retention, and outcomes. Psychiatric Services, 55, 767–774. doi:10.1176/appi.ps.55.7.767

31.

Huang

D. Y. C.

Evans

Hara

Weiss

R. E.

Hser

Y. I.

(2011). Employment trajectories: Exploring gender differences and impacts of drug use. Journal of Vocational Behavior, 79, 277–289. doi:10.1016/j.jvb.2010.12.001

32.

Huebner

B. M.

Cobbina

(2007). The effect of drug use, drug treatment participation, and treatment completion on probationer recidivism. Journal of Drug Issues, 37, 619–641. doi:10.1177/002204260703700307

33.

Humphreys

McLellan

A. T.

(2011). A policy-oriented review of strategies for improving the outcomes of services for substance use disorder patients. Addiction, 106, 2058–2066. doi:10.1111/j.1360-0443.2011.03464.x

34.

Jaffe

Huang

Hser

(2012). Drug-abusing offenders with comorbid mental disorders: Problem severity, treatment participation, and recidivism. Journal of Substance Abuse Treatment, 43, 244–250. doi:10.1016/j.jsat.2011.12.002

35.

Jang

Wang

Lin

(2014). Factors affecting employment outcomes for people with disabilities who received disability employment services in Taiwan. Journal of Occupational Rehabilitation, 24, 11–21. doi:10.1007/s10926-013-9433-1

36.

S. J.

Kim

(2013). An examination of major models and problems of research in return-to-work: Review of literature. Journal of Rehabilitation Research, 17, 119–134.

37.

Johnsen

B. H.

Bartone

Nissestad

O. A.

(2009). Predicting transformational leadership in naval cadets: Effects of personality hardiness and training. Journal of Applied Social Psychology, 39, 2213–2235.

38.

Katz

K. S.

El-Mohandes

Johnson

D. M.

Jarrett

Rose

Cober

(2001). Retention of low income mothers in a parenting intervention study. Journal of Community Health, 26, 203–218.

39.

Kawakami

Chiu

W. T.

Chatterji

deGraaf

Marschinger

LeBlanc

. . . Kessler

R. C.

(2013). Early-onset mental disorders and adult household income. In Alsoso

Chatterji

(Eds.), The burden of mental disorders (pp. 110–134). Cambridge, UK: Cambridge University Press.

40.

Kemp

Harris

Vurel

Sitharthan

(2007). Stop using stuff: Trial of a drug and alcohol intervention for young people with comorbid mental illness and drug and alcohol problems. Australasian Psychiatry, 15, 490–493. doi:10.1080/10398560701439665

41.

Kline

R. B.

(2004). Principles and practice of structural equation modeling (2nd ed.). New York, NY: Guilford Press.

42.

Kobasa

S. C.

(1979). Stressful life events, personality, and health: An inquiry into hardiness. Journal of Personality and Social Psychology, 37, 1–11.

43.

Latkin

C. A.

Curry

A. D.

Hua

Davey

M. A.

(2007). Direct and indirect associations of neighborhood disorder with drug use and high-risk sexual partners. American Journal of Preventive Medicine, 32, 234–241. doi:10.1016/j.amepre.2007.02.023

44.

Law

D. W.

(2005). Interactive organizational commitment and hardiness in public accountants’ turnover. Managerial Auditing Journal, 20, 383–393.

45.

Lopez-Goni

J. J.

Fernandez-Montalvo

Arteaga

(2011). Addiction treatment dropout: Exploring patients’ characteristics. The American Journal on Addictions, 21, 78–85. doi:10.1111/j.1521-0r391.2011.00188.x

46.

Luchansky

Brown

Longhi

Stark

Krupski

(2000). Chemical dependency treatment and employment outcomes: Results from the “ADATSA” program in Washington State. Drug and Alcohol Dependence, 60, 151–159. doi:10.1016/S0376-8716(99)00153-2

47.

Luszczynska

Cieslak

(2005). Protective, promotive, and buffering effects of perceived social support in managerial stress: The moderating role of personality. Anxiety, Stress, & Coping, 18, 227–244.

48.

Maddi

S. R.

(2006). Hardiness: The courage to grow from stresses. The Journal of Positive Psychology, 54, 160–168.

49.

Maddi

S. R.

Harvey

R. H.

Khoshaba

D. M.

Fazel

Resurreccion

(2009). The personality construct of hardiness, IV: Expressed in positive cognitions and emotions concerning oneself and developmentally relevant activities. Journal of Humanistic Psychology, 49, 292–305.

50.

Maddi

S. R.

Wadhwa

Haier

R. J.

(1996). Relationship of hardiness to alcohol and drug use in adolescents. American Journal of Drug and Alcohol Abuse, 22, 247–257.

51.

McCalister

K. T.

Dolbier

C. L.

Webster

J. A.

Mallon

M. W.

Steinhardt

M. A.

(2006). Hardiness and support at work as predictors of work stress and job satisfaction. American Journal of Health Promotion, 20, 183–191.

52.

McHugo

G. J.

Drake

R. E.

Xie

Bond

G. R.

(2012). A 10-year study of steady employment and non-vocational outcomes among people with serious mental illness and co-occurring substance use disorders. Schizophrenia Research, 138, 233–239. doi:10.1016/j.schres.2012.04.007

53.

McKay

J. R.

(2009). Treating substance use disorders with adaptive continuing care. Washington, DC: American Psychological Association. doi:10.1037/11888-000

54.

McLellan

A. T.

Cacciola

J. S.

Alterman

A. I.

Rikoon

S. H.

Carise

(2006). The Addiction Severity Index at 25: Origins, contributions and transitions. American Journal of Addiction, 15, 113–124. doi:10.1080/10550490500528316

55.

Metsch

L. R.

Pollack

H. A.

(2005). Welfare reform and substance abuse. Milbank Quarterly, 83, 65–99. doi:10.1111/j.0887-378X.2005.00336.x

56.

Montgomery

K. L.

Vaughn

M. G.

Thompson

S. J.

Howard

M. O.

(2013). Heterogeneity in drug abuse among juvenile offenders: Is mixture regression more informative than standard regression? International Journal of Offender Therapy and Comparative Criminology, 57, 1326–1346. doi:10.1177/0306624X12459185

57.

Neighbors

C. J.

Barnett

N. P.

Rohsenow

D. J.

Colby

S. M.

Monti

P. M.

(2010). Cost-effectiveness of a motivational intervention for alcohol-involved youth in a hospital emergency department. Journal of Studies on Alcohol and Drugs, 71, 384–394.

58.

Neumann

C. S.

Hare

R. D.

(2008). Psychopathic traits in a large community sample: Links to violence, alcohol use, and intelligence. Journal of Consulting and Clinical Psychology, 76, 893–899. doi:10.10/0022-006X.76.5.893

59.

Odenwald

Semrau

(2013). Dropout among patients in qualified alcohol detoxification treatment: The effect of treatment motivation is moderated by trauma load. Substance Abuse Treatment, Prevention, and Policy, 8(13), 1–21. doi:10.1186/1747-597X-8-14

60.

Redonnet

Chollet

Fombonne

Bowes

Melchior

(2012). Tobacco, alcohol, cannabis and other illegal drug use among young adults: The socioeconomic context. Drug and Alcohol Dependence, 121, 231–239. doi:10.1016/j.drugalcdep.2011.09.002

61.

Richardson

Wood

Montaner

Kerr

(2012). Addiction treatment-related employment barriers: The impact of methadone maintenance. Journal of Substance Abuse Treatment, 43, 276–284. doi:10.1016/j.jsat.2011.12.008

62.

Richardson

Abraham

(2012). Psychological correlates of university students’ academic performance: A systematic review and meta-analysis. Psychological Bulletin, 138, 353–387. doi:10.1037/a0026838

63.

Rinaldi

Montibeller

Perkins

(2011). Increasing the employment rate for people with longer-term mental health problems. The Psychiatrist, 35, 339–343.

64.

Rogers

E. S.

Anthony

W. A.

Cohen

Davies

R. R.

(1997). Prediction of vocational outcome based on clinical and demographic indicators among vocationally ready clients. Community Mental Health Journal, 33, 99–112.

65.

Rohde

Stice

Gau

J. M.

(2012). Effects of three depression prevention interventions on risk for depressive disorder onset in the context of depression risk factors. Prevention Science, 13, 584–593. doi:10.1007/s11121-012-0284-3

66.

Schulte

S. J.

Meier

P. S.

Stirling

Berry

(2010). Unrecognized dual diagnosis—A risk factor for dropout of addiction treatment. Mental Health and Substance Use, 3, 94–109. doi:10.1080/17523281003705199

67.

Sheidow

A. J.

McCart

Zajac

Davis

(2012). Prevalence and impact of substance use among emerging adults with serious mental health conditions. Psychiatric Rehabilitation Journal, 35, 235–243. doi:10.2975/35.3.2012.235.243

68.

Sivikis

D. S.

Silverman

Hang

N. A.

Stitzer

Keyser-Marcus

(2007). Behavioral strategies to improve treatment participation and retention by pregnant drug-dependent women. Substance Use & Misuse, 42, 1527–1735.

69.

Slesnick

Erdem

Collins

Bantchevska

Katafizsz

(2011). Predictors of treatment attendance among adolescent substance abusing runaways: A comparison of family and individual therapy modalities. Journal of Family Therapy, 33, 66–84. doi:10.1111/j.1467-6427.2010.00530.x

70.

Street

D. L.

(1995). Controlling extraneous variables in experimental research: A research note. Accounting Education, 4, 169–188.

71.

Substance Abuse and Psychiatric Services Administration. (2009). National Survey of Substance Abuse Treatment Services (N-SSATS): 2008. Data on substance abuse treatment facilities (DAIS Series: S-49, DHHS Publication No. SMA 09-4451). Rockville, MD: Substance Abuse and Mental Health Services Administration Office of Applied Studies.

72.

Thoma

R. J.

Monnig

M. A.

Lysne

P. A.

Ruhl

D. A.

Pommy

J. A.

Bogenschutz

. . . Yeo

R. A.

(2011). Adolescent substance abuse: The effects of alcohol and marijuana on neuropsychological performance. Alcoholism: Clinical & Experimental Research, 35, 39–46. doi:10.1111/j.1530-0277.2010.01320.x

73.

Thum

C. C.

Briber

Butler

S. F.

(2000). Working it out: A vocational education program. Newton, MA: Inflexxion.

74.

Tsang

H. W.

Fung

K. M.

Chung

R. C.

(2010). Self-stigma and stages of change as predictors of treatment adherence of individuals with schizophrenia. Psychiatric Research, 180, 10–15. doi:10.1016/j.psychres.2009.09.001

75.

Tuten

Fitzsimons

Chisolm

M. S.

Nuzzo

P. A.

Jones

H. E.

(2012). Contingent incentives reduce cigarette smoking among pregnant, methadone-maintained women: Results of an initial feasibility and efficacy randomized clinical trial. Addiction, 107, 1868–1877. doi:10.1111/j.1360-0443.2012.03923.x

76.

Wand

B. M.

Chiffelle

L. A.

O’Connell

N. E.

McAuley

J. H.

DeSouza

L. H.

(2010). Self-reported assessment of disability and performance-based assessment of disability are influenced by different patient characteristics in acute low back pain. European Spine Journal, 19, 633–640. doi:10.1007/s00586-009-1180-9

77.

Watson

(2005). Active engagement: Strategies to increase service participation by vulnerable families. Ashfield, Australia: Center for Parenting & Research.

78.

Westman

(1990). The relationship between stress and performance: The moderating effect of hardiness. Human Performance, 3, 141–155.

79.

Whitley

B. E.

Kite

M. E.

(2013). Principles of research in behavioral science (3rd ed.). New York, NY: Routledge.

80.

Xie

Drake

R. E.

McHugo

G. J.

Xie

Mohandas

(2010). The 10-year course of remission, abstinence, and recovery in dual diagnosis. Journal of Substance Abuse Treatment, 39, 132–140. doi:10.1016/j.jsat.2010.05.011