The Assessment of Recovery Capital (ARC) predicts substance abuse treatment completion

Journal Pre-proofs The Assessment of Recovery Capital (ARC) Predicts Substance Abuse Treatment Completion Jennifer Sánchez, Ethan Sahker, Stephan Arndt PII: DOI: Reference:

S0306-4603(19)30361-2 https://doi.org/10.1016/j.addbeh.2019.106189 AB 106189

To appear in:

Addictive Behaviors Addictive Behaviors

Received Date: Revised Date: Accepted Date:

31 March 2019 13 September 2019 21 October 2019

Please cite this article as: J. Sánchez, E. Sahker, S. Arndt, The Assessment of Recovery Capital (ARC) Predicts Substance Abuse Treatment Completion, Addictive Behaviors Addictive Behaviors (2019), doi: https://doi.org/ 10.1016/j.addbeh.2019.106189

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. 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.

© 2019 Published by Elsevier Ltd.

The Assessment of Recovery Capital (ARC) Predicts Substance Abuse Treatment Completion

Jennifer Sánchez, PhD a,b,c*† Ethan Sahker, MA b,d Stephan Arndt, PhD b,e,f

a

Department of Rehabilitation and Counselor Education, College of Education, The University of Iowa, N346 Lindquist Center, Iowa City, IA 52242

b

Iowa Consortium for Substance Abuse Research and Evaluation, The University of Iowa, 2662 Crosspark Rd., Coralville, IA 52241

c

I-SERVE (Iowa-Support, Education, and Resources for Veterans and Enlisted), The University

of Iowa, N122 Lindquist Center, Iowa City, IA 52242 d

Department of Psychological and Quantitative Foundations, Counseling Psychology Program, College of Education, The University of Iowa, 361 Lindquist Center, Iowa City, IA 52242

e

Department of Psychiatry, Carver College of Medicine, The University of Iowa, 451 Newton Road 200, Medicine Administration Building, Iowa City, IA 52242

f

Department of Biostatistics, College of Public Health, The University of Iowa, 145 N. Riverside Drive 100 CPHB, Iowa City, IA 52242

* Correspondence concerning this article should be addressed to Jennifer Sanchez, Ph.D., CRC, LMHC, Department of Rehabilitation and Counselor Education, The University of Iowa, N346 Lindquist Center, Iowa City, IA 52242. Contact: (319) 335-5985 or [email protected] † ORCID: 0000-0001-8448-0678

Abstract Recovery from addiction requires various personal and environmental resources. The purposes of this study were to determine if the Assessment of Recovery Capital (ARC) scores measured at admission could predict substance abuse treatment (SAT) completion and to identify personal and environmental factors associated with ARC scores. Participants (N=2,265) comprised clients entering a Midwestern SAT facility (August 2015–June 2017). Logistic regression was used to predict SAT completion using ARC scores. Nonparametric group comparisons were used for personal and environmental covariates. ARC scores significantly predicted successful SAT completion (OR = 1.05, 95% CI = 1.04, 1.05, Wald z = 12.9, p < 0.001). Employment had a positive relationship with ARC scores (Kruskal-Wallis χ2 = 215.96, df = 8, p < 0.001). ARC scores varied according to primary substance (Kruskal-Wallis χ2 = 101.10, df = 6, p < 0.001); alcohol and marijuana showed the highest scores and heroin the lowest. ARC scores decreased as number of problem substances increased (Kruskal-Wallis χ2 = 70.57, df = 2, p < 0.001, rS = – 0.163, p < 0.001). Living arrangement was also significant (Kruskal-Wallis χ2 = 146.36, df = 8, p < 0.001); clients who were homeless had the lowest ARC scores. A number of personal and environmental covariates were associated with the ARC scores and potentially with the outcome. After adjustment, the ARC remained a strong predictor of SAT completion. The ARC should be used in SAT facilities to guide treatment decisions and to create individualized treatment plans for clients.

Keywords: Recovery capital, substance abuse treatment completion, personal and environmental contextual factors, marijuana, alcohol, heroin

The Assessment of Recovery Capital (ARC) Predicts Substance Abuse Treatment Completion 1. Introduction Assessing substance use recovery is a difficult task with multiple attempts at developing useful instruments. Recovery from addiction requires use of a) personal, and b) environmental resources to reduce substance use (Laudet, 2007). The interaction between a person and their environment influences disability adjustment (Wright, 1983), or recovery from a substance use disorder (SUD). Thus, personal and environmental resources are crucial for maintaining recovery. For example, self-efficacy to resist use (Moos & Moos, 2007), positive social networks (e.g., family, work, peers in recovery; Moos & Moos, 2005), and neighborhood safety (Evans, Li, Buoncristiani, & Hser, 2014) are associated with abstinence years after completing substance abuse treatment (SAT). These personal and environmental resources serve as assets associated with recovery and are collectively known as “recovery capital” (White & Cloud, 2008). Deficit-based instruments that focus solely on substance use present inherent challenges to measuring recovery capital (Laudet, 2009; Simpson & Sells, 1990). Rather, important strengths-based components conceptualized in recovery capital include physical and mental health, social functioning, social networks, recovery group participation, meaning of recovery, motivation for initiating recovery, and mutual-aid group support use (Best et al., 2012; Best, Gow, Taylor, Know, & White, 2011). While recovery capital was initially developed to understand those who overcome addiction without treatment (Granfield & Cloud, 1999), insufficient research has been done to determine if recovery capital is useful in predicting treatment success. Recovery assessment instruments incorporate social and recovery networks (Brown, O’Grady, Battjes, & Katz, 2004), physical and mental health, employment/self-support, family

relations, and spirituality (Dodge, Krantz, & Kenny, 2010). Concerted efforts have been made to assess recovery potential in order to improve SAT outcomes. For example, the Global Appraisal of Individual Needs (GAIN; Dennis, White, Titus, & Unsicker, 2008), succeeded in gathering multiple social and environmental resources. However, it required extensive time to administer. Another measure of recovery capital excluded many by only focusing on alcohol use disorder (Sterling, Slusher, & Weinstein, 2008). Additionally, the Recovery Capital Scale (White, n.d.), succeeds in identifying multiple components of recovery capital, but is not yet validated and has an undetermined factor structure. Most recently, the Assessment of Recovery Capital (ARC) is a valid and reliable measure developed with input from individuals in recovery, rehabilitation professionals, and the extensive recovery literature (Groshkova, Best, & White, 2013). The ARC assesses “addiction recovery strengths” (Groshkova et al., 2013, p. 187) across 10 domains: 1. Substance Use and Sobriety (SUS; e.g., achieving abstinence, staying sober); 2. Global Health-Psychological (GH-Psy; e.g., confidence, self-efficacy); 3. Global Health-Physical (GH-Ph; e.g., energy, sleep hygiene); 4. Citizenship/Community Involvement (CCI; e.g., sense of belonging, social contribution); 5. Social Support (SS; e.g., positive relationships, adequate assistance); 6. Meaningful Activities (MA; e.g., recreation, self-improvement); 7. Housing and Safety (HS; e.g., independent, feeling secure); 8. Risk Taking (RT; e.g., prosociality, accountability); 9. Coping and Life Functioning (CLF; e.g., responsibility, self-care); and 10. Recovery Experience (RE; e.g., life purpose, optimism). Predicting successful SAT completion could be extremely valuable. Reviews of factors that complicate the course of SUDs underscore the importance of recovery capital; it may prove

to be a useful framework in guiding treatment and assessing recovery potential (White & Cloud, 2008). The ARC is already validated for predicting recovery. However, there is a need for a relatively short and simple measure to predict SAT completion, identify potential barriers to recovery, and assist rehabilitation professionals in providing appropriate services. The purposes of this study were to: (1) determine the utility of the ARC in predicting SAT completion, and (2) identify personal and environmental factors associated with ARC scores. Based on the aforementioned research, we hypothesized that recovery capital (represented by ARC scores) at admission will predict SAT completion. Personal and environmental factors (e.g., age, sex/gender, marital status, education, employment) were included for exploratory analyses. 2. Methods 2.1. Participants Participants (N=2,265) comprised clients entering a Midwestern SAT facility (August 2015–June 2017). Clients were in extended (72%) or intensive (10%) outpatient, high-intensity residential care (17%), or other (partial hospitalization, families in focus, continuing care, none; <1%) treatment. Treatment plans were individualized, with outpatient programming including medication management, urine screening, professional (individual, couples/family, group) counseling, and peer mentoring. Residential programming included primary health and psychiatric care, gender-specific individual and group counseling (e.g., dialectical behavior therapy, relapse prevention), family-based intervention, and nutrition education. All data were collected as part of an agency-based evaluation. For this study, personal health information was removed, and all data were de-identified (e.g., ARC scores, demographics, substance use history). The study adhered to Helsinki ethical guidelines: The University of Iowa institutional

review board exempted this study from review, confidentiality relating to medical (HIPPA) and SUD (42 CFR Part 2) patient records did not apply, and informed consent was not required. 2.2. Measures Admission data included standard clinical assessments (e.g., ARC, problem substances used), and sociodemographic information (e.g., age, sex/gender, race/ethnicity, income source). The ARC score, primary problem substance, and total number of problem substances were used in analyses. Due to low frequencies, race was recoded (i.e., as White, Black, and Other). The primary outcome was SAT completion. Treatment-related information, entered at discharge by treatment staff, was dichotomized as “successful” (Completed Treatment/Treatment Plan Completed or Completed Treatment/Treatment Plan Substantially Completed) or “unsuccessful” (all others—transferred, left against professional advice, terminated by facility, incarcerated, other) SAT completion. The ARC (Groshkova et al., 2013), a 50-item self-report scale, was used to assess recovery capital. Sample items include: “I am currently completely sober” (SUS), “I am coping with the stresses in my life” (GH-Psy), “I feel physically well enough to work” (GH-Ph), “It is important for me that I make a contribution to society” (CCI), “I get lots of support from friends” (SS), “I engage in activities that I find enjoyable and fulfilling” (MA), “I feel safe and protected where I live” (HS), “I have the privacy I need” (RT), “I meet all of my obligations promptly” (CLF), and “When I think of the future I feel optimistic” (RE). Respondents check each statement they agree with, checked items are summed to create ARC score (0–50), higher scores indicate greater recovery capital. The ARC is a unidimensional measure of recovery capital (Arndt, Sahker, & Hedden, 2017; Groshkova et al., 2013) with excellent internal consistency reliability (alpha = 0.92; Arndt et al., 2017), modest test-retest reliability (r = 0.61), and good

concurrent validity based on correlations with the World Health Organization’s physical (r = 0.83), psychological (r = 0.84), social (r = 0.69), and environmental (r = 0.82) quality of life domains, and the Treatment Outcome Profile’s physical health (r = 0.35), psychological health (r = 0.39), and quality of life (r = 0.40) subscales (Groshkova et al., 2013). 2.3. Statistical Analysis We used the Statistical Analysis Software (SAS) 9.4 for Windows to manage raw data and perform all analyses. Missing data were gender (n = 2), marital status (n = 5), and race (n = 34 who identified as Hispanic/Latino—ethnicity); we deleted missing cases when running relevant analyses. To predict SAT completion, we used logistic regression. Due to the large sample size, we set significance to p < 0.01 to focus on the more meaningfully-sized effects, and assist with Type I error rate inflation from multiple tests. Effect sizes (Cohen’s d, adjusted/odds ratios [OR/AOR], and area under the curve [AUC]) are provided for major findings. To assess the relationship of the ARC scores to personal and environmental covariates, we used nonparametric tests. ARC scores were markedly skewed (upper 25% = 46–50, lower 25% = 0– 27), so we used Kruskal-Wallis tests to compare demographic groups and Spearman correlations for continuous variables. 3. Results 3.1. ARC Predictive Utility A simple logistic regression predicting successful SAT completion from ARC scores yielded a significant effect (OR = 1.05, 95% CI = 1.04, 1.05, Wald z = 12.90, p < 0.001), with AUC being 0.67. Repeating this analysis controlling for all personal and environmental variables only slightly attenuated the odds ratio for the ARC (AOR = 1.04, 95% CI = 1.02, 1.05, Wald z = 6.99, p < 0.001), with AUC being 0.74, indicating the effect was robust and provided

independent information. All covariates only produced an AUC of 0.72. Clients were grouped in quartiles by ARC score and percentage of clients who completed SAT (see Figure 1 for illustration). <<>> 3.2. Personal Factors Clients median age was 32 (M = 34.2, SD = 12.2), most were White, nonHispanic/Latino, and about two-thirds were male. The median number of years of education was 12 (range: 0-20), with less than one-fourth reporting less than 12 years. Almost half (44.5%) were employed full- or part-time. Mean monthly income from wages was $1,147 (SD = $2,110); more than one-fourth (27.6%) reported no income. The most frequently reported primary problem substance was alcohol. (See Table 1 for complete personal factor data.) <<>> Hispanics/Latinos had significantly higher ARC scores than non-Hispanics/Latinos (Kruskal-Wallis χ2 = 10.15, df = 1, p < 0.001); however, differences among race groups were insignificant. There was a small negative correlation between ARC scores and age (rs = –0.13, p < 0.001), and males had slightly higher ARC scores than females (Kruskal-Wallis χ2 = 28.15, df = 1, p < 0.001). While there were no significant correlations with education (rs = –0.02, p = 0.29), higher ARC scores were significantly related to employment (Kruskal-Wallis χ2 = 215.96, df = 8, p < 0.001) and directly correlated with monthly gross income (rs = 0.24, p < 0.001). There was an inverse relationship with ARC scores and number of problem substances reported (Kruskal-Wallis χ2 = 70.57, df = 2, p < 0.001, rS = –0.163, p < 0.001) and primary substance was significantly related to ARC scores (Kruskal-Wallis χ2 = 101.10, df = 6, p < 0.001), with alcohol and marijuana showing the highest scores and heroin showing markedly low scores.

3.3. Environmental Factors Most frequently cited sources of referral were probation (21.5%) and self-referral (22.3%). Less than half (41.6%) of the clients reported wages/salary as their primary income source. Few (3.6%) reported being homeless. (See Table 2 for full environmental factor data.) <<>> ARC scores varied by referral source (Kruskal-Wallis χ2 = 400.88, df = 10, p < 0.001); referrals from Employer/EAP/School, for driving while intoxicated (DUI/OWI), and probation had the highest scores, while those from drug courts, health care providers, and self-referrals had the lowest. Clients who reported wages/salary as their main income source had higher ARC scores than those through other sources (Kruskal-Wallis χ2 = 166.10, df = 3, p < 0.001). Living arrangement was also a significant factor (Kruskal-Wallis χ2 = 146.36, df = 8, p < 0.001); clients who were homeless had the lowest ARC scores of any group. There was a significant difference in ARC scores between those who successfully completed SAT and those who did not (KruskalWallis χ2 = 192.58, df = 1, p < 0.001), representing a medium-large effect size (Cohen’s d = 0.61; Cohen, 1992). 4. Discussion The main purpose of this study was to provide additional evidence for the predictive utility of the ARC. We demonstrated that the ARC could predict SAT completion. Our hypotheses were that higher ARC scores (i.e., greater recovery capital) at admission would predict successful SAT completion, and this was supported. The second purpose was to examine the associations between ARC scores and personal and environmental variables. We found that being married and employed, and having higher levels of education and secure housing, were positively related to ARC scores, while having co-occurring mental health issues and SUD

severity were negatively related. The main hypothesis was supported, with additional findings demonstrating some personal and environmental factors were associated with SAT completion. 4.1. ARC Predictive Utility Recovery capital (ARC scores) predicted successful SAT completion in an almost linear fashion. The likelihood of successful SAT completion increased 1.04 times for every 1-point ARC score increase, even when controlling for all personal and environmental variables. This suggests the ARC is a clinically useful assessment tool for admission, treatment, and discharge planning. Rehabilitation professionals working in SAT facilities can use the ARC to assess clients’ recovery capital, and inform development and implementation of individualized treatment plans, capitalizing on recovery capital. Practitioners could use the ARC to identify SAT barriers, provide interventions to increase recovery capital (Best et al., 2012; Hibbert & Best, 2011; Moos & Moos, 2005), and improve clients’ chances of successfully completing SAT. 4.2. Personal Differences Race was not associated with recovery capital; however, ethnicity was. Previous researchers found that race (Arndt, Acion, & White, 2013; Sahker, Toussaint, Ramirez, Ali, & Arndt, 2015; Saloner, & Lê Cook, 2013) and ethnicity (Arndt et al., 2013; Sahker et al., 2015) were associated with SAT completion, favoring Whites/Caucasians over Blacks/African Americans and Hispanics/Latinos. Our finding that being Hispanic/Latino was associated with SAT completion is both novel and important. One potential reason for our divergent finding may be related to familismo—a social structure in which the family’s needs take precedence over any one family member’s needs. Among Mexican Americans, family support served to protect against psychological distress and environmental crises (Umaña-Taylor, Updegraff, & Gonzales-

Backen, 2011). Thus, the social capital associated with more collective cultures may add a significant source of recovery capital for individuals who are entering SAT. Employment was directly related to recovery capital. Our finding aligns with other researchers who found work to be an important factor of recovery (Best et al., 2012; CritsChristoph et al., 2015; Haaga, Hall, & Haas, 2005; White & Cloud, 2008). Work-related benefits such as structure, self-worth, and social support aid recovery (Platt, 1995), while unemploymentrelated issues including higher substance use (Compton, Gfroerer, Conway, & Finger, 2014) and incarceration (Platt, 1986) may further impede recovery. Contrary to previous findings (Knight, Logan, & Simpson, 2001), education level was not associated with recovery capital in our study; however, being a student was. Having a higher level of education in the absence of employment may not meet goals and may explain why education level alone revealed no association. SAT facilities should collaborate with state Vocational Rehabilitation (VR) agencies to incorporate employment-related (e.g., job placement/support) services to increase recovery capital. Mental health issues were negatively associated with recovery capital. Research on psychiatric symptoms’ impact on SAT completion is conflicted. Some researchers found lower levels of depression among clients in recovery (Hibbert & Best, 2011), while others found psychopathology was unrelated to SAT completion (Darke, Campbell, & Popple, 2012). The benefits of treating psychiatric and SUDs concurrently versus sequentially (Sánchez et al., 2017) are important to consider. Number of substances used and primary problem substance were inversely associated with recovery capital, and may represent SUD severity. These findings support existing research demonstrating poor SAT outcomes associated with use frequency (Allen & Olson, 2015), as well as number (Wickizer et al., 1994) and type (Roll, Prendergast,

Richardson, Burdon, & Ramirez, 2005) of problem substances used. The health capital associated with having a SUD that is less “severe,” may impact recovery capital. 4.3. Environmental Differences Referral source and recovery capital mirror known aspects about referral source and SAT completion. Employee assistance program (EAP) referrals are associated with some of the highest SAT completion rates (Arndt et al., 2013; Sahker et al., 2015), and had some of the highest ARC scores in our study. Clients referred by their employers may feel a sense of accountability to complete their program, and fear losing their job and benefits (e.g., financial stability) if they do not. Conversely, health care provider and self-referrals have been associated with the lowest rates of SAT completion (Arndt et al., 2013; Sahker et al., 2015), and with the lowest ARC scores in our study. For those who successfully complete SAT, self-referral is associated with after-care treatment engagement (Duffy & Baldwin, 2013). Rehabilitation professionals should consider the positive effect referral sources and motivation for entering SAT can have on successful SAT completion (external) and long-term recovery (internal). Living arrangement was related to recovery capital. Homelessness had a strong negative effect on SAT completion, which may be for various reasons. This finding provides further support for the association between unsuccessful SAT completion and homelessness (Didenko, & Pankratz, 2007; Lennings, Kenny, & Nelson, 2006) due to lack of access to health care treatment or health insurance (Palepu et al., 2013), fear of social estrangement from their community (Freund & Hawkins, 2004), or peer pressure (Rhoades et al., 2011). In our sample, over one-fourth reported having no income. Thus, the status capital associated with financial security and secure housing may translate to recovery capital for clients. 4.4. Limitations

The following limitations should be considered when interpreting our results. Many admission variables (e.g., type and frequency of substances used) were gathered via self-report, which may be inaccurate due to cognitive impairments resulting from chronic substance use (Brown, Kranzler, & Del Boca, 1992). However, self-report data, particularly as it pertains to substance use, was found to be reliable and valid (Bell, Williams, Senier, Strowman, & Amoroso, 2003; Napper, Fisher, Johnson, & Wood, 2010). The majority of our sample identified as White, graduated high school, reported some income, and had secure housing (i.e., were not homeless); therefore, our results may not generalize to more diverse populations or those of lower socioeconomic status. Our data were obtained from a Midwestern state where residents are expected to be primarily White and of higher socioeconomic status (Arndt et al., 2017), which may explain the low level of homelessness in our sample. Finally, data were gathered from one SAT facility and may not generalize to others. However, a clear advantage of collecting data from only one facility is consensus defining “treatment completion,” which may be lacking when attempting to compile data from numerous facilities. 4.5. Conclusion In conclusion, our findings mostly support our hypotheses. ARC scores predicted successful SAT completion and may be a useful aid in treatment and assessment planning. Personal (e.g., employment) and environmental (e.g., referral source) factors influenced ARC scores. These findings add further nuance to the predictive validity of the ARC. Our findings highlight the need for integrated and wrap-around services for clients entering SAT to bolster recovery capital and successful completion, especially when treating opioid use disorders.

References Allen, R. S., & Olson, B. D. (2015). Predicting attrition in the treatment of substance use disorders. Journal of Addiction Research & Therapy, 6, 238. doi:10.4172/21556105.1000238 Arndt, S., Acion, L., & White, K. (2013). How the states stack up: Disparities in substance abuse outpatient treatment completion rates for minorities. Drug and Alcohol Dependence, 132, 547–554. doi:10.1016/j.drugalcdep.2013.03.015 Arndt, S., Sahker, E., & Hedden, S. (2017). Does the Assessment of Recovery Capital scale reflect a single or multiple domains? Substance Abuse and Rehabilitation, 8, 39–43. doi:10.2147/SAR.S138148 Bell, N. S., Williams, J. O., Senier, L., Strowman, S. R., & Amoroso, P. J. (2003). The reliability and validity of the self-reported drinking measures in the Army's Health Risk Appraisal survey. Alcoholism: Clinical and Experimental Research, 27, 826–834. doi:10.1097/01.ALC.0000067978.27660.73 Best, D., Gow, J., Knox, T., Taylor, A., Groshkova, T., & White, W. (2012). Mapping the recovery stories of drinkers and drug users in Glasgow: Quality of life and its associations with measures of recovery capital. Drug and Alcohol Review, 31, 334–341. doi:10.1111/j.1465-3362.2011.00321.x Best, D., Gow, J., Taylor, A., Knox, A., & White, W. (2011). Recovery from heroin or alcohol dependence: A qualitative account of the recovery experience in Glasgow. Journal of Drug Issues, 41, 359–377. doi:10.1177/002204261104100303

Brown, B. S., O’Grady, K. E., Battjes, R. J., & Katz, E. C. (2004). The Community Assessment Inventory—Client views of supports to drug abuse treatment. Journal of Substance Abuse Treatment, 27, 241–251. doi:10.1016/j.jsat.2004.08.002 Brown, J., Kranzler, H. R., & Del Boca, F. K. (1992). Self-reports by alcohol and drug abuse inpatients: Factors affecting reliability and validity. British Journal of Addiction, 87, 1013–1024. doi:10.1111/j.1360-0443.1992.tb03118.x Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155–159. Compton, W. M., Gfroerer, J., Conway, K. P., & Finger, M. S. (2014). Unemployment and substance outcomes in the United States 2002–2010. Drug and Alcohol Dependence, 142, 350–353. doi:10.1016/j.drugalcdep.2014.06.012 Crits-Christoph, P., Markell, H. M., Gallop, R., Gibbons, M. B. C., McClure, B., & Rotrosen, J. (2015). Predicting outcome of substance abuse treatment in a feedback study: Can recovery curves be improved upon? Psychotherapy Research, 25, 694–704. doi:10.1080/10503307.2014.994146 Darke, S., Campbell, G., & Popple, G. (2012). Retention, early dropout and treatment completion among therapeutic community admissions. Drug and Alcohol Review, 31, 64–71. doi:10.1111/j.1465-3362.2011.00298.x Dennis, M. L., White, M., Titus, J. C., & Unsicker, J. (2008). Global Appraisal of Individual Needs: Administration guide for the GAIN and related measures (Version 5). Normal, IL: Chestnut Health Systems. Didenko, E., & Pankratz, N. (2007). “Substance use: Pathways to homelessness? Or a way of adapting to street life?” Visions: BC’s Mental Health and Addictions Journal, 4(1), 9–10. Retrieved from https://www.heretohelp.bc.ca/visions

Dodge, K., Krantz, B., & Kenny, P. J. (2010). How can we begin to measure recovery? Substance Abuse Treatment, Prevention, and Policy, 5, 31. doi:10.1186/1747-597X-5-31 Duffy, P., & Baldwin, H. (2013). Recovery post treatment: Plans, barriers and motivators. Substance Abuse Treatment, Prevention, and Policy, 8, 6. doi:10.1186/1747-597X-8-6 Evans, E., Li, L., Buoncristiani, S., & Hser, Y.-I. (2014). Perceived neighborhood safety, recovery capital, and successful outcomes among mothers 10 years after substance abuse treatment. Substance Use & Misuse, 49, 1491–1503. doi:10.3109/10826084.2014.913631 Freund, P. D., & Hawkins, D. W. (2004). What street people reported about service access and drug treatment. Journal of Health & Social Policy, 18(3), 87–93. doi:10.1300/J045v18n03_05 Granfield, R., & Cloud, W. (1999). Coming clean: Overcoming addiction without treatment. New York, NY: NYU Press. Groshkova, T., Best, D., & White, W. (2013). The assessment of recovery capital: Properties and psychometrics of a measure of addiction recovery strengths. Drug and Alcohol Review, 32, 187–194. doi:10.1111/j.1465-3362.2012.00489.x Haaga, D. A. F., Hall, S. M., & Haas, A. (2005). Participant factors in treating substance use disorders. In L. Castonguay & L. Beutler (Eds.), Principle of therapeutic change that work (pp. 275–292). New York, NY: Oxford University Press. Hibbert, L. J., & Best, D. W. (2011). Assessing recovery and functioning in former problem drinkers at different stages of their recovery journeys. Drug and Alcohol Review, 30, 12– 20. doi:10.1111/j.1465-3362.2010.00190.x

Knight, D. K., Logan, S. M., & Simpson, D. (2001). Predictors of program completion for women in residential substance abuse treatment. The American Journal of Drug and Alcohol Abuse, 27, 1–18. doi:10.1081/ADA-100103116 Laudet, A. B. (2009). Environmental scan of measures of recovery. Rockville, MD: Substance Abuse and Mental Health Services Administration. Retrieved from http://www.williamwhitepapers.com/pr/Recovery%20Measures%20Laudet%202009.pdf Laudet, A. B. (2007). What does recovery mean to you? Lessons from research experience for research and practice. Journal of Substance Abuse Treatment, 33, 243–256. doi:10.1016/j.jsat.2007.04.014 Lennings, C., Kenny, D. T., & Nelson, P. (2006). Substance use and treatment seeking in young offenders on community orders. Journal of Substance Abuse Treatment, 31, 425–432. doi:10.1016/j.jsat.2006.05.017 Moos, R. H., & Moos, B. S. (2005). Sixteen-year changes and stable remission among treated and untreated individuals with alcohol use disorders. Drug and Alcohol Dependence, 80, 337–347. doi:10.1016/j.drugalcdep.2005.05.001 Moos, R. H., & Moos, B. S. (2007). Protective resources and long-term recovery from alcohol use disorders. Drug and Alcohol Dependence, 86, 46–54. doi:10.1016/j.drugalcdep.2006.04.015 Napper, L. E., Fisher, D. G., Johnson, M. E., & Wood, M. M. (2010). The reliability and validity of drug users' self reports of amphetamine use among primarily heroin and cocaine users. Addictive Behavaviors, 35, 350–354. doi:10.1016/j.addbeh.2009.12.006 Palepu, A., Gadermann, A., Hubley, A. M., Farrell, S., Gogosis, E., Aubry, T., & Hwang, S. W. (2013). Substance use and access to health care and addiction treatment among homeless

and vulnerably housed persons in three Canadian cities. PLoS ONE, 8(10): e75133. doi:10.1371/journal.pone.0075133 Platt, J. J. (1986). Heroin addiction: Theory, research and treatment. (Vol. 1, 2nd ed.). Melbourne, FL: Krieger. Platt, J. J. (1995). Vocational rehabilitation of drug abusers. Psychological Bulletin, 117, 416– 433. doi:10.1037/0033-2909.117.3.416 Rhoades, H., Wenzel, S. L., Golinelli, D., Tucker, J. S., Kennedy, D. P., Green, H. D., & Zhou, A. (2011). The social context of homeless men’s substance use. Drug and Alcohol Dependence, 118, 320–325. doi:10.1016/j.drugalcdep.2011.04.011 Roll, J. M., Prendergast, M., Richardson, K., Burdon, W., & Ramirez, A. (2005). Identifying predictors of treatment outcome in a drug court program. The American Journal of Drug and Alcohol Abuse, 31, 641–656. doi:10.1081/ADA-200068428 Sahker, E., Toussaint, M. N., Ramirez, M., Ali, S. R., & Arndt, S. (2015). Evaluating racial disparity in referral source and successful completion of substance abuse treatment. Addictive Behaviors, 48, 25–29. doi:10.1016/j.addbeh.2015.04.006 Saloner, B., & Lê Cook, B. (2013). Blacks and Hispanics are less likely than Whites to complete addiction treatment, largely due to socioeconomic factors. Health Affairs, 32, 135–145. doi:10.1377/hlthaff.2011.0983 Sánchez, J., Muller, V., Garcia, M. E., Martinez, S. N., Cool, S. T., & Gandarilla, E. (2017). Psychiatric rehabilitation outcomes among Hispanics with co-occurring serious mental illness and substance use disorders: A systematic review. Journal of Applied Rehabilitation Counseling, 48(1), 40–49.

Simpson, D. D., & Sells, S. B. (1990). Opioid addiction and treatment: A 12-year follow-up. Malabar, FL: Robert E. Krieger Publishing Company. Sterling, R., Slusher, C., & Weinstein, S. (2008). Measuring recovery capital and determining its relationship to outcome in an alcohol dependent sample. The American Journal of Drug and Alcohol Abuse, 34, 603–610. doi:10.1080/00952990802308114 Umaña-Taylor, A. J., Updegraff, K. A., & Gonzales-Backen, M. A. (2011). Mexican-origin adolescent mothers' stressors and psychosocial functioning: Examining ethnic identity affirmation and familism as moderators. Journal of Youth Adolescence, 40, 140–157. doi:10.1007/s10964-010-9511-z White, W. (n.d.). Recovery Capital Scale. Retrieved from http://www.williamwhitepapers.com/pr/Recovery%20Capital%20Scale.pdf White, W., & Cloud, W. (2008). Recovery capital: A primer for addictions professionals. Counselor, 9(5), 22–27. Retrieved from http://www.williamwhitepapers.com/pr/2008RecoveryCapitalPrimer.pdf Wickizer, T., Maynard, C., Atherly, A., Frederick, M., Koepsell, T., Krupski, A., & Stark, K. (1994). Completion rates of clients discharged from drug and alcohol treatment programs in Washington State. American Journal of Public Health, 84, 215–221. doi:10.2105/AJPH.84.2.215 Wright, B. A. (1983). Physical disability: A psychosocial approach (2nd ed.). New York, NY: HarperCollins.

Table 1 Sample Personal Contextual Factors and Differences in ARC Scores (N = 2,265) Personal Factors

Sample Percent (n)

Mean ARC (sd)1

Sex*** Male 67.6% (1529) 36.08 (12.19) Female 32.4% (734) 33.25 (12.67) Race (n.s.) White 94.4% (2107) 35.04 (12.49) Black/African American 4.6% (103) 37.66 (11.15) Other 0.9% (21) 31.48 (10.85) Ethnicity** Hispanic/Latino 4.1% (92) 38.64 (11.78) Not Hispanic/Latino 95.9% (2173) 35.02 (12.41) Marital Status*** Single, Never Married 54.2% (1224) 35.90 (12.43) Cohabitating 5.0% (114) 35.62 (13.28) Married 15.4% (347) 35.60 (11.26) Separated 6.4% (145) 31.34 (13.52) Divorced 18.1% (408) 33.91 (12.48) Widowed 1.0% (22) 33.82 (10.06) Employment Status*** Employed, Full-Time 35.0% (792) 39.35 (10.54) Employed, Part-Time 9.5% (216) 37.74 (10.90) Student 2.6% (59) 39.12 (10.38) Homemaker 0.5% (12) 35.00 (11.83) Unemployed, Looking for Work 32.1% (728) 32.76 (12.55) Unemployed, not Looking 12.8% (290) 29.86 (13.53) Unemployed, due to disability 5.2% (117) 30.02 (13.45) Unemployed, Inmate 0.8% (17) 24.59 (11.24) Retired 1.5% (34) 34.56 (9.86) Mental Health Issues*** Yes 56.1% (1270) 31.54 (12.65) No 43.9% (995) 39.80 (10.40) Primary Problem Substance*** Alcohol 42.2% (955) 36.45 (11.64) Marijuana/Hashish 23.5% (533) 38.06 (11.21) Cocaine/Crack 0.4% (10) 33.10 (13.80) Methamphetamine 28.3% (641) 31.84 (13.49) Opiates/Synthetics 3.3% (74) 31.24 (11.41) Heroin 1.2% (27) 26.93 (13.64) Other 1.1% (25) 31.08 (11.82) Number of Problem Substances*** 1 38.7% (876) 37.11 (11.44) 2 35.8% (811) 35.57 (12.64) 3 25.5% (578) 31.65 (12.76) 1 Differences among groups were tested using Kruskal-Wallis one-way ANOVA. Note. ARC = Assessment of Recovery Capital. ** p < 0.01; *** p < 0.001; n.s. = not significant

Table 2 Sample Environmental Contextual Factors and Differences in ARC Scores (N = 2,265) Environmental Factors

Sample Percent (n)

Mean ARC (sd)1

Referral Source*** Self 22.3% (506) 29.05 (13.28) Health Care Provider 11.8% (267) 28.78 (11.27) Alcohol/Drug Abuse Provider 0.5% (12) 35.50 (11.49) Employer/EAP/School 1.5% (35) 39.97 (8.52) DHS/Other 10.6% (239) 34.87 (11.45) Other Community 0.4% (10) 36.40 (7.92) Civil Commitment 3.0% (69) 30.35 (12.25) Drug Court 0.8% (18) 28.67 (13.69) DUI/OWI 14.7% (332) 41.57 (9.28) Other Criminal Justice/Court 12.8% (290) 37.90 (11.68) Probation 21.5% (487) 39.73 (10.11) Living Arrangement*** Alone 20.4% (462) 34.74 (12.55) With Parents 21.0% (476) 36.07 (12.32) With significant other only 9.9% (224) 35.08 (11.63) With significant other & child(ren) 11.2% (254) 36.54 (11.09) With child(ren) alone 4.1% (93) 35.29 (12.48) With another adult 15.7% (356) 33.72 (13.12) With another adult & child(ren) 5.9% (133) 36.35 (10.93) Supervised Living 8.2% (186) 40.74 (9.07) Homeless 3.6% (81) 19.77 (10.76) Income Source*** Wages/Salary 41.6% (943) 39.05 (10.63) Family/Friends 23.3% (527) 31.80 (12.71) Fixed Assistance 7.5% (170) 31.38 (12.50) None 27.6% (625) 33.18 (13.01) Successful Treatment Completion*** Yes 49.3% (1026) 38.87 (11.05) No 50.7% (1054) 31.56 (12.69) 1 Differences among groups were tested using Kruskal-Wallis one-way ANOVA. Note. ARC = Assessment of Recovery Capital; EAP = employee assistance program; DUI/OWI = driving under the influence/operating while intoxicated; DHS = Department of Human Services. *** p < 0.001

ARC Scores and Successful Treatment Completion

Percent Clients Succesful

80% 70% 60%

68%

50%

60%

40% 30% 20% 10%

40% 30%

0% Q1

Q2 Q3 ARC Score Quartiles

Q4

Figure 1. Quartile groups represent percentage of clients (error bars reflect 95% confidence internals) who successfully completed substance abuse treatment based on their score on the Assessment of Recovery Capital (ARC) scale.

Conflict of Interest: None to declare.

Highlights:



Assessment of Recovery Capital (ARC) predicts substance abuse treatment completion.



Personal and environmental contextual factors predicted ARC scores.



Identifying as Hispanic/Latino was associated with successful treatment completion.



Employer, court, and probation referrals were associated with treatment completion.



Heroin substance use and homelessness were barriers to treatment completion.