Ethics of Mandated Treatment

Substance Use & Misuse, 43:1785–1801 Copyright © 2008 Informa Healthcare USA, Inc. ISSN: 1082-6084 (print); 1532-2491 (online) DOI: 10.1080/10826080802297484

Identifying Client-Level Indicators of Recovery Among DUI, Criminal Justice, and Non–Criminal

Justice Treatment Referrals

ROBERT WALKER, JENNIFER COLE, AND T. K. LOGAN

Center on Drug and Alcohol Research, University of Kentucky, Lexington, Kentucky, USA

This study is part of a mandated treatment outcome study on all government-funded programs in a rural state. This naturalistic study included a sample of 888 clients who served between July 2003 and June 2004 in a state-funded treatment for substance misuse and were included in a follow-up interview 12 months after treatment. To examine differences in treatment outcome, clients were examined in three referral conditions: (1) driving under the influence (DUI) referral; (2) criminal justice referral; and (3) non–criminal justice referral. While more DUI referrals reported alcohol use at 12- month follow-up, there were no other differences between referral conditions. Instead, controlling for factors like age, gender, and race, recovery intent at intake, and 12- step program participation at follow-up predicted positive treatment outcomes, while persistent depression predicted negative outcomes. This study of clients in state-funded treatment for substance misuse provides additional evidence that referral condition does not predispose clients toward positive or negative outcomes. Secondly, client-level factors related to recovery practices and intent to reduce or stop using substances may need closer attention in the clinical process. Study limitations included data being collected by clinicians during intake, which may have resulted in reliability questions about how data are entered.

Keywords recovery indicators; recovery intent; outcome indicators; treatment out- comes; naturalistic environment

Introduction

There is increasing interest in the outcomes associated with treatment for substance use– related disorders, along with an emphasis on the use of evidence-based practices with substance use–related disorders. In 2007, the Substance Abuse and Mental Health Ser- vices Administration (SAMHSA) issued a requirement for states to collect the National Outcomes Measures, which SAMHSA describes as “the lifeblood of quality assurance at each level of administration—Federal, State, and local” (SAMHSA, 2007). For substance user treatment,1 the most critical outcome objective is to attain and sustain “abstinence

This study was funded by the Kentucky Division of Mental Health and Substance Abuse under a contract with the University of Kentucky Center on Drug and Alcohol Research.

Address correspondence to Robert Walker, Center on Drug and Alcohol Research, University of Kentucky, 915B South Limestone Street, Lexington, KY 40536. E-mail: [email protected]

1Treatment can be briefly and usefully defined as a planned, goal-directed change process, of necessary quality, appropriateness, and conditions (endogenous and exogenous), which is bound (by culture, place, time, etc.) and can be categorized into professional-based, tradition-based, mutual

1785

1786 Walker et al.

from drug use and alcohol abuse,” along with improved functioning (SAMHSA, 2007). In response to these policies, providers have an increasing need to identify what works, for whom, and under what conditions. The focus on attaining positive treatment out- comes is intensified by the fact that only a small percent of persons needing treatment ever receive it (SAMHSA, 2006). For example, in 2005 there were an estimated 22.2 mil- lion people over the age of 12 in the United States with substance abuse or dependence problem, but only 3.9 million had received any substance abuse services in the past 12 months, and 2.2 million had received services from a self-help group, and 1 million (4.5%) had received services at a mental health center in the past 12 months (SAMHSA, 2006).

For several decades there has been interest in the outcomes of treatment for substance use–related disorders, with a preponderance of evidence suggesting that positive outcomes result from a variety of different clinical approaches and modalities, including residential and outpatient counseling (Floyd, Monahan, Finney, and Morley, 1996; Morley, Finney, Monahan, and Floyd, 1996; Moyer, Finney, and Swearingen, 2002; Swearingen, Moyer, and Finney, 2003). Further, length of treatment has been demonstrated to be associated with better treatment outcomes in several studies (Hser, Evans, Huang, and Anglin, 2004; Moos and Moos, 2003; Moos, Moos, and Andrassy, 1999). Studies have also demonstrated that client characteristics as well as motivation and creation and maintenance of therapeutic alliance contribute to outcomes (Cacciola, Dugosh, Foltz, Leahy, and Stevens, 2005; Ilgen, McKellar, Moos, and Finney, 2006; Joe, Simpson, Dansereau, and Rowan-Szal, 2001). Thus, providers who have an increased investment in achieving positive outcomes may need to not only use evidence-based practices and skilled clinicians but also pay close attention to client-level variables that may foster posttreatment recovery. Clinicians may benefit from being able to screen for indicators of client recovery intent as a way of identifying clients most likely to benefit from treatment services. Clients who are identified as having lower potential for positive outcomes may require additional motivational approaches or pretreatment services.

Community treatment for substance misuse receives many, if not most, of its clients from the criminal justice system (Farabee and Leukefeld, 2001). Criminal justice–referred clients may underreport substance use and related problems and may lack internal motiva- tion to engage in treatment processes (Farabee and Leukefeld, 2001). However, research has largely dispelled mistaken beliefs about criminal justice system– and DUI-referred clients not benefiting from treatment or recovery (Cavaiola, Strohmetz, and Abreo, 2007; DeYoung, 1997; Gregoire and Burke, 2004; Hiller, Knight, Rao, and Simpson, 2002; Lo- gan, Hoyt, McCollister, French, Leukefeld, and Minton 2004; Kelly, Finney, and Moos, 2005; Miller and Flaherty, 1999; Ninonuevo and Hoffmann, 1993). However, how clients are referred to treatment (due to DUI charges, criminal justice, or other, non–criminal jus- tice referral) may still interact with other important factors that affect outcomes or may have an independent impact on treatment outcomes in a naturalistic treatment setting. The literature has identified other client characteristics that have influenced negative treatment outcomes such as a history of unemployment, depression, and other mental health problems

help–based (AA, NA, and the like), and self-help (“natural recovery”) models. There are no unique models or techniques used with substance users—of whatever type—which aren’t also used with nonsubstance users. In the West, with the relatively new ideology of “harm reduction” and the even newer quality of life (QOL) treatment–driven model there are now a new set of goals in addi- tion to those derived from/associated with the older tradition of abstinence-driven models. Editor’s note.

Indicators of Recovery at Intake 1787

(Rounsaville, Dolinsky, Babor, and Meyer, 1987; Sinha and Schottenfeld, 2001). For ex- ample, pretreatment employment has been identified as an important indicator of positive substance user treatment outcome (Cebulla, Smith and Sutton, 2004; Galaif, Newcomb, and Carmona, 2001; McCaul, Svikis, and Moore, 2001; McLellan, 1983; Slaymaker and Owen, 2006; Sterling, Gottheil, Glassman, Weinstein, Serota, and Lundy, 2001; Vaillant, 1988). Also, low social functioning and overall severity of mental health symptoms have been demonstrated to predict negative treatment outcomes (McLellan, Alterman, et al., 1994), and depression, in particular, may predict decreased likelihood of abstinence following treatment (Dodge, Sindelar, and Sinha, 2005). However, it is unclear whether clinicians in publicly funded treatment programs, who may have biases about “unmotivated” court- referred clients, trust the findings from controlled research studies of criminal justice– and DUI-referred clients and their treatment outcomes. In addition, it is unclear whether the findings from controlled studies about the treatment outcomes of criminal justice and DUI clients are actually replicated in naturalistic studies of publicly funded treatment. While the drug abuse treatment outcome study (DATOS) suggested positive treatment outcomes across a wide range of treatment sites, the study was carried out among carefully recruited treatment sites and for a specified 2-year period (Fletcher, Tims, and Brown, 1997; Flynn, Craddock, Hubbard, Anderson, and Etheridge, 1997). There is a need for ongoing studies that are embedded in everyday practice settings on a routine basis to help identify predictors of better outcomes in terms of abstinence. Naturalistic studies of substance user treatment outcomes possess realism and external validity because they examine outcomes in real-world situations (Timko, Moos, Finney, and Connell, 2002). One other important component of naturalistic research on substance user treatment outcomes is that self-reported client re- covery activities and intentions can be examined along with their clinical characteristics and referral conditions and in the wide mix of treatment types and approaches that occur under the “treatment-as-usual” condition.

Recovery activity or intent toward recovery, while subject to influence through motiva- tional approaches (Miller and Rollnick, 2002), is distinct from actions taken by treatment providers, since these two factors are within clients’ sphere of experience and control. The two are independent of treatment per se (McLellan, Chalk, and Bartlett, 2007). Recovery is a term with many different denotations that has overlap with treatment outcomes but has clear connotations associated with using mutual help (McLellan, et al., 2007; Tims, Leukefeld, and Platt, 2001). Recovery activity is also a client-level factor rather than a treatment ac- tivity or program-related factor. For the purposes of this study recovery is understood as abstinence from alcohol or drugs.

Study Objectives

To better understand the relative role of client-level recovery activity and intent to end or reduce substance use under different referral conditions, we examined outcomes in the nat- uralistic environment of publicly funded treatment in one state, by focusing on follow-up data to identify intake client characteristics that predict factors related to positive outcomes 12 months after treatment. The study examined client-level clinical characteristics associ- ated with substance use, referral condition, and clients’ self-reported intent to be substance free and their participation in mutual help 1 year after intake. The study hypothesis was that neither clinical characteristics nor referral condition would predict treatment outcomes but that clients’ report of positive recovery intent and use of mutual help at intake and/or follow-up would predict positive outcomes.

1788 Walker et al.

Method

Procedures

In Kentucky, all state-funded programs treating substance misuse participate in a statuto- rily mandated treatment outcome study. After informing clients about the purpose of the follow-up study and the study’s confidentiality protections, clinicians in outpatient, intensive outpatient, and residential settings collect data on clients during the intake and assessment phase of services. The Kentucky Substance Abuse Treatment Outcome Study (KTOS) is conducted annually, using intake data collected by clinicians in the course of substance use assessment. The data were collected using a personal digital assistant (PDA)–based instru- ment that is administered by the clinician. The intake data were synchronized via modem on a regular basis to the University of Kentucky Center on Drug and Alcohol Research (CDAR) for analysis.

Clients who voluntarily agreed to participate in the follow-up study gave informed consent to participate before giving personal locator information that was used to locate them for follow-up telephone interviews 12 months after treatment. Research staff from CDAR then sampled clients for follow-up interviews. In state fiscal year 2004, there were 9,876 intake records, and 3,136 clients consented to follow-up interviews and had face valid contact information. The initial sample was 50% of these (1,568) with 249 being ineligible (in controlled living conditions or deceased), and 431 could not be located, with a final follow-up sample of 888 clients. The follow-up rate was 67.3%. All data are client self-reports. No incentive was given for participation in the study at intake. Participants received $20 for completing the follow-up interview. All study procedures were approved by the University of Kentucky institutional review board.

Participants

Overall there were 9,876 intake records of client entering state-funded treatment in the Commonwealth of Kentucky during a 12-month period (from July 1, 2003 to June 30, 2004). The sample for this analysis was 888 adults who participated in a follow-up inter- view approximately 12 months later. The treatment programs included outpatient, intensive outpatient, case management, and short-term (30-day) residential settings statewide, rang- ing from urban to very rural sites. Clients providing intake information included even those who came only for assessment visits. Just over one fifth of clients (20.4%) received 4 or fewer services, 28.3% received 5–15 services, 21.5% received 16–30 services, 17.2% re- ceived 31–50 services, and only 12.6% received 51 or more services. Clients often received a combination of residential and outpatient services.

Measures

Substance Use. Substance use measures were taken from the SAMHSA Center on Sub- stance Abuse Treatment (CSAT) Government Performance and Results Act (GPRA) data collection tool, which has been used to examine treatment outcomes in treatment capacity expansion and other CSAT-funded programs (Mulvey, Atkinson, Avula, and Luckey, 2005). The CSAT GPRA is based on the Addiction Severity Index (ASI) (Kosten, Rounsaville, and Kleber, 1983; McLellan, Kushner, et al., 1992), and it measures substance use, criminal activity, employment, and other related behaviors during the past 30 days. For the this study, the GPRA was modified to include past 12-month and lifetime use as well as past 30-day

Indicators of Recovery at Intake 1789

use of all substances. Clients were asked if they had ever used each class of substance (e.g., alcohol, and illicit drugs like marijuana, opiates, tranquilizers, cocaine, stimulants, nonprescription methadone, inhalants, and hallucinogens), and if so, how many months out of the past 12 months they had used each class of substance. A composite measure of any illicit drug use was computed from clients’ reports of individual classes of illicit drugs, by computing the maximum number of months that illicit drugs were reported.

Recovery Intent. Questions were added to the core instrument to examine self-reported 12-step program participation at intake and follow-up as well as clients’ own rating of the odds of being able to get off and stay off drugs or alcohol. These questions do not examine motivation, but were developed to characterize recovery intent and use of recovery activities independent of treatment. First, attendance at Alcoholics Anonymous (AA) or Narcotics Anonymous (NA) meetings during the 30 days prior to intake and follow-up were included in the analyses as two separate variables. Second, clients were asked at intake, “Based on what you know about yourself and your situation, how good are the chances that you can get off and stay off of drugs/alcohol?” The values ranged from 1 (very good) to 5 (very poor). They were also asked at intake and follow-up, “How many days in the past 30 days have you attended AA, NA or other mutual-help group meetings?”

Mental Health Problems. The mental health measures were taken from the ASI and included self-reported depression, anxiety, trouble in concentration, difficulty in controlling violence, hallucinations, as well as suicidal thoughts and attempts in the past 12 months (McLellan et al., 1992). Since depression at intake can be directly a function of substance use (due either to intoxication or withdrawal effects), clients who reported experiencing serious depression at both intake and follow-up were categorized as experiencing persistent depression to exclude substance-affected depressed mood of a more transient nature.

Criminal Justice System Involvement. Criminal justice referral conditions were derived from ASI-adapted measures of referral source. Clients were asked if the admission was prompted or suggested by the criminal justice system and whether the admission resulted from a DUI charge. Questions about number of arrests in the past 12 months and the past 30 days were modified from the ASI.

Data Analysis

Two logistic regression models were run to examine the relationship between clients’ in- volvement with the criminal justice system, indicators of intent to achieve and maintain abstinence, mental health problems, and recovery from alcohol use and illicit drug use ap- proximately 12 months after intake into substance abuse treatment. In one model alcohol use in the 12-month follow-up period was the outcome variable, and in the second model any illicit drug use in the 12-month follow-up period was the outcome variable. Involvement with the criminal justice system was operationalized as the three groups: (1) clients who were referred to treatment by the criminal justice system for any charge other than a DUI were categorized into the CJ group (n = 296); (2) clients who were referred to treatment by the criminal justice system based on a DUI charge were categorized into the DUI group (n = 273); and (3) clients who had a referral condition not related to the criminal justice system were categorized as belonging to the Non CJ group (n = 317). Two clients were dropped from the group analysis because no data were available on their referral condition at intake. In the logistic regression models the Non CJ group was used as the reference

1790 Walker et al.

group. Number of arrests in the 12 months before intake was also included as predictor variable in the logistic regression models. Indicators of intent to reduce or end substance use were taken from two items. The first one was the clients’ rating at intake of their chances of staying off alcohol/drugs. Second, attendance at AA/NA meetings in the 30 days before intake and the 30 days before follow-up were included in the analyses as two separate vari- ables. Attendance at AA/NA meetings in the 30 days before follow-up was used to indicate recovery activity independent of treatment. The correlation between attendance in mutual help groups in the 30 days before intake and the 30 days before follow-up was small (Pear- son r = 0.294). Also, clients who reported experiencing serious depression at both intake and follow-up were categorized as experiencing persistent depression. Control variables included gender, race, age, employment status at treatment intake, and the highest level of education attained. Control variables were selected because each has been associated with independent contributions to outcomes, and there were significant differences in these variables across the three referral conditions.

In order to assess recovery from use of alcohol and illicit drugs, only individuals who reported use of each class of substance in the 12 months before intake were included in each of the logistic regression models. Because interpretation of adjusted odds ratios is difficult to interpret, relative risk was used when possible (Holcomb, Chaiworapongsa, Luke, and Burgdorf 2001; Osborne, 2006; Zhang, 1998).

Results

Sample Descriptives by Referral Condition

Table 1 presents the results of bivariate analyses of the criminal justice groups on demo- graphic variables. The vast majority of individuals in the DUI group were male, and a significantly greater proportion of the DUI group was male compared to individuals in the CJ group and the Non CJ group. Individuals in the CJ group were significantly younger than individuals in the Non CJ group and individuals in the DUI group. The DUI group was composed of a larger proportion of White individuals compared to the other two groups. The greatest proportion of individuals reported that they had either never been married (38.6% of the sample) or recently been divorced (27.8% of the sample). The only difference in marital status was that significantly fewer individuals in the DUI group were separated at the intake interview compared to the individuals in the Non CJ group. The average highest level of education attained by the sample was a little less than 12 years of education. Individuals in the CJ group reported significantly more years of education compared to individuals in the other two groups. More individuals in the DUI group were employed full-time at the time of the intake interview compared to individuals in the Non CJ group and the CJ group, and significantly fewer individuals in the DUI group were unemployed compared to the individuals in the other two groups. About 16% of the sample reported disability at intake.

Table 2 presents the results of bivariate analyses of the criminal justice groups on mental health, treatment, perceptions of treatment success, mutual help group participation, and arrests. Significantly more clients in the Non CJ group reported depression at intake, follow- up, and both time periods compared to clients in the CJ group and DUI group. Compared to clients in the DUI group, significantly more clients in the Non CJ group reported that they had ever been in substance abuse treatment before the current treatment; however, there was no difference by CJ group in the number of times individuals had been in treatment among those who had had past treatment. The majority of clients in all the groups rated

Indicators of Recovery at Intake 1791

Table 1 Demographic characteristics of follow-up sample at intake by criminal justice

referral group

No CJ DUI CJ, referral charge non-DUI Statistical

Demographics Response (n = 317) (n = 273) (n = 296) test Gender Masculine 49.8%a 85.7%a,b 54.4%b χ 2(2) = 92.658∗∗ Mean age 34.4a 35.2b 31.7a,b F (2, 883) = 9.724∗∗ Race White 82.6%a 95.2%a,b 83.4%b χ 2(4) = 33.488∗∗

Black 13.6%a 4.0%a,b 15.9%b

Other 3.8% 0.7% 0.7% Marital status Never married 38.9% 33.5% 42.9% χ 2(8) = 22.007∗

Married 20.3% 23.9% 18.9% Divorced 25.3% 33.8% 25.0% Separated 14.6%a 6.3%a 10.8% Widowed 0.9% 2.6% 2.4%

Education Mean education 11.8a,b 11.2a 11.2b F (2, 878) = 10.140∗∗

(years) Employment Current Full time 22.5%a 41.0%a,b 26.1%b χ 2(8) = 35.604∗∗

employment Part time 12.1% 10.3% 10.2% status Unemployed 45.1%a 28.6%a,b 47.5%b

Disabled 17.1% 17.2% 13.9% Other 3.2% 2.9% 2.4%

a,b,c: groups differ significantly at p < .01; ∗ p < .01; ∗∗ p < .001.

their chances of success in treatment as being moderately to very good, with significant differences between the Non CJ group and the DUI group. Clients in the DUI group had the lowest rates of mutual help group participation at both intake and follow-up when compared to the other two groups. Finally, there was no significant difference in number of arrests between the three groups.

A total of 513 clients (58%) reported using alcohol in the 12 months before follow- up, and 275 clients (31.1%) reported using illicit drugs in the 12 months before follow-up. Further, among the clients who reported using alcohol in the 12 months before intake, 69.6% reported using alcohol at follow-up. Among the clients who reported using illicit drugs in the 12 months before intake, 40.7% reported using illicit drugs at follow-up. In addition, at follow-up, 196 clients (38.4%) reported using both alcohol and illicit drugs in the past 12 months.

Multivariate Analysis

Among clients who reported that they had used alcohol in the 12 months preceding the intake interview (n = 634), several predictors were significantly associated with alcohol use at follow-up. First, clients who were referred to treatment based on a DUI charge were 1.28 times more likely to report using alcohol in the 12 months after intake compared to

T ab

le 2

D ep

re ss

io n,

in di

ca to

rs of

m ot

iv at

io n

fo r

tr ea

tm en

t, an

d ar

re st

s fo

r fo

ll ow

-u p

sa m

pl e

by cr

im in

al ju

st ic

e gr

ou p

N o

C J

D U

I C

J, re

fe rr

al ch

ar ge

no n-

D U

I S

ta ti

st ic

al R

es po

ns e

(n =

31 7)

(n =

27 3)

(n =

29 6)

te st

M en

ta l

h ea

lt h

D ep

re ss

io n

at in

ta ke

Y es

62 .7

% a,

b 26

.4 %

a 34

.2 %

b χ

2 (2

) =

90 .1

31 ∗∗

D ep

re ss

io n

at fo

ll ow

-u p

Y es

46 .4

% a,

b 27

.1 %

a 30

.7 %

b χ

2 (2

) =

27 .7

60 ∗∗

P er

si st

en t

de pr

es si

on Y

es 36

.6 %

a, b

13 .6

% a

18 .0

% b

χ 2 (2

) =

50 .4

87 ∗∗

P a

st tr

ea tm

en t

E ve

r be

en in

su bs

ta nc

e ab

us e

tr ea

tm en

t be

fo re

Y es

68 .3

% a

64 .3

% 55

.5 %

b χ

2 (2

) =

10 .9

84 ∗

M ea

n no

.o f

ti m

es tr

ea tm

en t

ha s

be en

ta ke

n (f

or th

os e

w ho

ha d

ev er

be en

in tr

ea tm

en t)

2. 7

2. 7

2. 4

F (2

,5 49

) =

0. 71

4

P er

ce p

ti o

n o

f ch

a n

ce s

o f

su cc

es s

S el

f- re

po rt

ed ch

an ce

s of

su cc

es s

in tr

ea tm

en t

M od

er at

el y

to ve

ry go

od 77

.8 %

a 83

.1 %

89 .9

% a

χ 2 (4

) =

17 .0

53 ∗

U nc

er ta

in 18

.4 %

a 15

.1 %

8. 4%

a

V er

y to

m od

er at

el y

po or

3. 8%

1. 8%

1. 7%

S el

f- h

el p

g ro

u p

p a

rt ic

ip a

ti o

n A

tt en

de d

A A

/N A

m ee

ti ng

s in

th e

30 da

ys be

fo re

in ta

ke Y

es 39

.2 %

a 12

.1 %

a, b

33 .4

% b

χ 2 (2

) =

56 .9

42 ∗∗

A tt

en de

d A

A /N

A m

ee ti

ng s

in th

e 30

da ys

be fo

re fo

ll ow

up Y

es 35

.5 %

a 14

.4 %

a, b

43 .7

% b

χ 2 (2

) =

58 .8

15 ∗∗

A rr

es ts

M ea

n no

.o f

ar re

st s

in th

e 12

m on

th s

be fo

re in

ta ke

0. 70

1. 3

1. 5

F (2

,8 81

) =

3. 84

6

a, b :

gr ou

ps di

ff er

si gn

ifi ca

nt ly

at p

< .0

1; ∗ p

< .0

1; ∗∗

p <

.0 01

.

1792

Indicators of Recovery at Intake 1793

Table 3 Logistic regression predicting alcohol use at follow-up

β Wald Odds ratio (C.I.)

Gender (0 = Masculine) −0.432 4.516 0.649 (0.384, 1.096) Age −0.021 4.844 0.979 (0.955, 1.004) Race (0 = White) −0.130 0.220 0.878 (0.431, 1.789) Employed (0 = Employed) 0.126 0.423 1.135 (0.688, 1.871) Highest level of education completed 0.088 3.600 1.092 (0.969, 1.231) Persistent depression (0 = No) 0.869 13.215∗∗ 2.385 (1.288, 4.414) No. of arrests in the past 12 months −0.057 0.758 0.944 (0.797, 1.119) No. of times in substance abuse treatment

in lifetime 0.011 0.093 1.011 (0.924, 1.106])

Chances of staying off alcohol or drugs 0.144 1.796 1.155 (0.876, 1.522) Self-help in the 30 days before intake −0.107 0.241 0.899 (0.513, 1.574) Self-help in the 30 days before follow-up −0.587 7.963∗ 0.556 (0.326, 0.950) DUI-referred 0.907 12.753∗∗ 2.476 (1.288, 4.763) CJ-referred 0.076 0.109 1.079 (0.596, 1.952)

∗ p < .01; ∗∗ p < .001.

individuals who were not referred to treatment by the criminal justice system (RR = 1.28). Second, clients who reported persistent depression were 1.23 times more likely to report using alcohol in the 12 months after intake compared to clients who did not report persistent depression (BRR = 1.23). Third, clients who reported attending AA/NA meetings in the 30 days before follow-up were less likely to report alcohol use in the 12 months before follow-up (RR = 0.834). None of the other variables were significantly associated with alcohol use during the follow up period (Nagelkerke R2 = 0.130).

Table 4 Logistic regression predicting illicit drug use at follow-up

β Wald Odds ratio (C.I.)

Gender (0 = Male) −0.341 2.899 0.711 (0.424, 1.191) Age −0.026 5.918 0.975 (0.949, 1.002) Race (0 = White) −0.346 1.518 0.707 (0.343, 1.459) Employed (0 = Employed) .263 1.762 1.300 (.781, 2.164) Highest level of education completed 0.127 6.447 1.136 (0.998, 1.292) Persistent depression (0 = No) 1.086 24.644∗∗ 2.963 (1.686, 5.206) No. of arrests in the past 12 months −.040 .501 .961 (.830, 1.112) No. of times in substance abuse treatment

in lifetime 0.092 5.683 1.096 (0.993, 1.211)

Chances of staying off alcohol or drugs 0.423 15.983∗∗ 1.526 (1.162, 2.003) Self-help in the 30 days before intake −0.324 2.398 0.723 (0.421, 1.240) Self-help in the 30 days before follow-up −0.042 0.041 0.959 (0.565, 1.628) DUI-referred −0.433 2.389 0.649 (0.315, 1.334) CJ-referred −0.116 0.284 0.891 (0.509, 1.559)

∗ p < .01; ∗∗ p < .001.

1794 Walker et al.

Among clients who reported that they had used illicit drugs in the 12 months preceding the intake interview (n = 568), several predictors were significantly associated with illicit drug use at follow-up. First, persistent depression (RR = 1.78) was positively associated with illicit drug use at follow-up. In other words, clients who reported experiencing persistent depression were 1.78 as likely to report using illicit drugs during the follow-up period compared to clients who did not experience persistent depression. Second, clients’ ratings at intake of their chances of staying off alcohol/drugs were significantly associated with using illicit drugs during the follow-up period. Clients who rated their chances of staying off alcohol/drugs as better, thus expressing intent toward recovery, were less likely to report using illicit drugs at follow-up. Level of involvement with the criminal justice system was not associated with the likelihood of using illicit drugs during the follow up-period. No other variables were significantly associated with reporting illicit drug use during the follow-up period (Nagelkerke R2 = 0.166).

Discussion

We found that the hypothesis was in part substantiated. One referral condition (DUI) did predict negative treatment outcomes with DUI offenders more likely reporting alcohol use at follow-up. However for all other referral conditions, there was no alcohol or drug use outcome effect. We also found that self-reported intent toward recovery and use of mutual help predicted substance use outcomes with those who reported using mutual help at follow- up being less likely to report alcohol use at follow-up. Also, clients who reported good or very good chances of stopping illicit drug use at intake were less likely to report any illicit drug use at follow-up. Persistent depression was not included in the hypothesis, but it too predicted more likely negative outcomes.

This naturalistic study examined 886 substance-using clients who entered community- based treatment with one of three referral conditions—criminal justice, DUI, or non– criminal justice. The clients received substance abuse treatment from a variety of community-based, publicly funded programs and in varying intensity of services.

To better identify indicators of positive treatment and recovery outcomes in state- funded community treatment this study considered referral conditions, client-level clinical characteristics, as well as self-reported recovery intent and use of mutual help. Given the many factors that can contribute to outcomes we used a multivariate analysis to control for alternative explanations for recovery-related outcomes. By examining the outcomes of clients in three referral conditions it was clear that there were very few differences between them after controlling for other variables. There were four key findings that have importance to clinical providers in publicly funded treatment: (1) referral conditions such as court or probation referrals or DUI-initiated treatment did not predict treatment outcomes positively or negatively with the exception of DUI-referred clients being more likely to report alcohol use at follow-up; (2) persistent depression (that is, depression that was present at intake and still at follow-up) predicted a greater likelihood of alcohol and illicit drug use at follow- up; (3) client reports at intent to achieve abstinence of intake were significantly associated with lower likelihood of reporting illicit drug use 12 months after treatment; and (4) while reporting use of mutual help at intake was not associated with outcomes, clients who reported use of mutual help in the period before follow-up were significantly less likely to report alcohol use at follow-up. Findings on the role of depression on substance use outcomes are consistent with research studies that have demonstrated that clients with co-occurring mental health problems have poorer substance use outcomes than those without mental

Indicators of Recovery at Intake 1795

health problems (Dodge, Sindelar, and Sinha, 2005; Ritscher, Moos, and Finney, 2002). In fact, depression is being increasingly identified as a risk factor for overall mortality among all disease-related causes of death (Mykletun et al., 2007). Its prominence as a contributor to mortality as well as disease expression suggests that clinical attention to it in substance misuse treatment and recovery support should be of paramount importance. In addition, depression may interfere with help-seeking and recovery behaviors (Mykletun et al., 2007). Thus, even providers of recovery supports, including members of the recovery community, might be alerted to the importance of facilitating treatment for depression to aid recovery from substance use. This education of the mutual-help community might include clarification of the actions of antidepressant medication compared to other psychoactive substances to dispel concern about the use of antidepressants being simply another form of drug dependence.

In addition, this study adds two important findings for clinical practice in regard to recovery in relation to treatment outcomes. Both of the key findings suggest the importance of focus by clinicians on clients’ own contributions to recovery instead of merely adding more treatment. First, clients’ own rating of their chances of getting off and staying off drugs or alcohol at intake and assessment was significantly associated with lower rates of reported illicit drug (but not alcohol) use at follow-up. Whether this measure was related to treatment motivation was not examined in this study. However, it is a simple measure to use in clinical practice, and client responses may be important cues to treatment and recovery intent. Second, client reports of taking mutual help at intake did not significantly predict abstinence outcomes, but use of mutual help after intake did predict greater likelihood of reporting abstinence from alcohol. While clinicians may not be able to directly monitor clients’ use of mutual help, these findings suggest that encouraging and guiding clients to use mutual help may be a significant contribution to treatment outcomes. This study’s findings on mutual help may have differentiated between clients who report mutual help at intake as a way to manage an impression of seriousness and clients who stick with mutual help 12 months later. The latter group clearly reports recovery activity, whereas the intake reports may be associated with managing how probation officers and treatment providers view the client. Other research has suggested that mutual help may be an important determinant of sustained abstinence (Moos and Moos, 2006; Moos, Schaefer, Andrassy, and Moos, 2001). A long-term follow-up study of alcohol dependent persons who were initially untreated reported that 12-step program participation in the first year of the study predicted better outcomes 16 years later (Moos and Moos, 2006). Furthermore, encouraging the use of mutual help means promoting recovery activity, which places more emphasis on what clients can do above mere participation in treatment. These simple ways to ask the client about intent and use of recovery activities may in part address complex questions about internal versus external motivations for treatment that arise with criminal justice and other court-related referrals (Leukefeld, Tims, and Platt, 2001).

Study’s Limitations

There were limitations to this study. First, the follow-up sample was taken from clients who consented at intake to participate in the follow-up study; therefore, it is possible that the clients in the follow-up sample do not represent all clients who enter treatment in state-funded substance user treatment. There were only three significant differences ( p < .01) between the follow-up and non–follow-up samples: the follow-up sample contained more females than the non–follow-up sample (37.6% vs. 32.4%); the follow-up sample had completed more education (11.4 years vs. 11.2 years); and the follow-up sample reported a

1796 Walker et al.

lower average number of arrests in the 30 days before intake compared to those who were not followed up (0.1 vs. 0.2).

All the data are client self-reports, and there were over 150 clinicians collecting data in a wide variety of clinical settings. While the intake data were collected by clinicians who may be under obligations to report to the court or probation/parole, the follow-up interviews were conducted by research staff under the direction of a study coinvestigator. Furthermore, participants were informed that interview data were covered by a federal Certificate of Confidentiality. The validity and reliability of self-reports of substance use has been supported by a number of studies (Del Boca and Noll, 2000; Rutherford, et al., 2000). Earlier studies have found that the context of the interview influences reliability (Babor, Stephens, and Marlatt, 1987), and generally self-reports at the beginning of treatment as well as during treatment have been demonstrated to be reliable (Rutherford et al., 2000). In addition, it is important to understand the reliance on self-reports in health research as well as in substance use and misuse studies. For example, research on other chronic health problems that have behavioral and recovery components such as diabetes, chronic headache, obesity, hypertension, and heart disease often depend on self-reported diet, exercise, medication compliance, and weight reduction efforts (Holroyd et al., 2001; Mokdad et al., 2001; Pereira et al., 2002). In addition, the depression measure did not include specific depression-related symptoms or criteria that are included in the DSM-IV-TR diagnosis.

This naturalistic study of treatment outcomes among clients in state-funded treatment for substance misuse has several implications for the practice community. First, findings provide additional evidence that referral condition does not predispose clients toward posi- tive or negative outcomes, with the exception of DUI referral being associated with alcohol use at follow-up. Second, client-level factors related to recovery practices and intent to reduce or stop using substances may need closer attention in the clinical process. There are two uses of this information: (1) client self-reports of intent to end or reduce substance use may provide important indicators of level of intensity of services that should be used; and (2) clinicians may need to more intently encourage engagement with self-help activities such as AA and NA.

The recognition of the importance of client-level factors in the recovery and treatment outcome process suggests that an exclusive focus on evidence-based or best practices may miss important factors related to recovery. This study suggests that clinicians may take into greater consideration clients’ intent level to end or reduce substance use and client recovery. The identification of clients who report little recovery intent may need either increased motivational approaches or pretreatment services. Alternatively, with low levels of funding for an ever-increasing demand in treatment services, clinicians may need to focus treatment efforts on those who convey the greatest intent toward recovery. These findings also suggest the possibilities for empowering clients to take more charge of their own recovery processes as a way to better treatment outcomes.

RÉSUMÉ

Identification d’indicateurs de guérison au niveau du client parmi les personnes en traitement de toxicomanie pour conduite sous influence, pour des raisons de justice

criminelle, et des raisons non-criminelles

La présente étude fait partie d’un projet d’évaluation des résultats du traitement concer- nant tous les programmes subventionnés par le gouvernement dans un état rural aux Etats

Indicators of Recovery at Intake 1797

Unis. Un des buts de l’étude est de générer des connaissances concernant les caractéristiques et les résultats des clients qui peuvent être utilisés pour améliorer les services. Cette étude utilise un échantillon de 888 clients recrutes entre juillet 2003 et juin 2004 en traitement d’abus de substances subventionné par l’État et qui ont participe a un entretien de suivi 12 mois après le traitement. Trois catégories de clients étaient examinées selon la raison pour leur entrée dans le programme pour examiner les différences de résultats: (1) Conduite sous influence (DUI – ‘driving under the influence’); (2) justice criminelle ; et (3) autres raisons. Tandis qu’un plus grand nombre de personnes dans le groupe DUI affirmaient consommer de l’alcool lors du suivi 12 mois plus tard, aucune autre différence n’a été constatée entre les groupes. Apres avoir contrôler pour l’âge, le sexe, la race et d’autres facteurs, l’intention de guérison à l’entrée et la participation aux 12 étapes au suivi prédisaient un résultat positif du traitement, tandis que la dépression persistante prédisait des résultats négatifs. Cette étude de clients dans le traitement pour abus de substances subventionné par l’État fournit des preuves supplémentaires que la raison pour entrer en traitement ne prédispose pas le client pour un résultat positif ou négatif. Deuxièmement, les facteurs au niveau du client qui sont liés aux pratiques de récupération et l’intention de réduire ou d’arrêter l’utilisation de substances pourraient nécessiter une plus grande attention dans le processus clinique.

RESUMEN

Definición de indicadores de curación en el cliente entre las personas en tratamiento de toxicomanı́a para conducta bajo influencia, por razones de justicia criminal,

y razones no criminales

El presente estudio forma parte de un proyecto de evaluación de los resultados del tratamiento relativo todos los programas subvencionados por el Gobierno en un estado rural en los Estados Unidos. Uno de los objetivos del estudio es generar conocimientos relativas a las caracterı́sticas y los resultados de los clientes que pueden utilizarse para mejorar los servicios. Este estudio utiliza una muestra de 888 clientes reclutados entre julio de 2003 y junio de 2004 en tratamiento de abusos de sustancias subvencionado por el Estado y que tienen participa tiene un mantenimiento de seguimiento 12 meses después del tratamiento. Se examinaban tres categorı́as de clientes según la razón para su entrada en el programa para examinar las diferencias de resultados: (1) Conducta bajo influencia (DUI—‘driving under the influence’); (2) justicia criminal; y (3) otras razones. Mientras que un mayor número de personas en el grupo DUI afirmaban consumir alcohol en el seguimiento 12 meses más tarde, ninguna otra diferencia se constató entre los grupos. Después de controlar para la edad, el sexo, la raza y de otros factores, la intención de curación a la entrada y la participación en las 12 etapas al seguimiento predecı́an un resultado positivo del tratamiento, mientras que la depresión persistente predecı́a resultados negativos. Este estudio de clientes en el tratamiento para abuso de sustancias subvencionado por el Estado proporciona pruebas suplementarias que la razón para entrar en tratamiento no predispone al cliente para un resultado positivo o negativo. En segundo lugar, los factores en el cliente que están vinculados a las prácticas de recuperación y la intención de reducir o decidir la utilización de sustancias podrı́an requerir una mayor atención en el proceso clı́nico.

1798 Walker et al.

THE AUTHORS

Robert Walker, MSW, LCSW, is an assistant professor of psychiatry at the University of Kentucky Center on Drug and Alcohol Research with conjoint appointments in behavioral science and social work. His over fifty publi- cations span a wide range of health and behavioral health topics including substance abuse, professional ethics in clinical practice, partner violence perpetration and vic- timization, and traumatic brain injury. He is the principal investigator for a state-mandated substance abuse treat- ment outcome study, a statewide outcome study of case management services for special education courses (SED) children and youth, and he is the evaluator for two feder- ally funded (CSAT) and four other state-funded projects.

Before coming to the university, he had over 25 years’ experience in the community mental health system as a clinician and Community Mental Health Center (CMHC) director, and he maintains close relationships with the mental and other health providers throughout the state. He has taught psychopathology as well as research in the master’s program in the College of Social Work. He has been a coinvestigator on partner violence studies in rural and urban areas and has been an evaluator of substance abuse treatment programs in rural and inner-city programs.

Jennifer Cole, MSW, is a PhD candidate in the College of Social Work at the University of Kentucky. She cur- rently works on the Kentucky Treatment Outcome Study Follow-Up as a research coordinator. She has worked as a project coordinator for a National Institute on Alcohol Abuse and Alcoholism (NIAAA) study, which examined alcohol, violence, mental health, health status, and service utilization among rural and urban women with protective orders against male partners, and a project coordinator on a National Institute on Drug Abuse (NIDA) study, which examined the nature, extent, and co-occurrence of HIV- risk behavior, violence, and crack use. Her primary inter- ests are in the areas of HIV sexual risk, intimate partner

violence, sexual violence, revictimization, and mental health issues of women.

Logan, PhD, is currently a professor in the department of behavioral science at the University of Kentucky and the Center on Drug and Alcohol Research, with joint ap- pointments in psychiatry, psychology, and social work. Dr. Logan has been funded by the NIDA, the NIAAA, and the National Institute of Justice (NIJ) to examine victimiza- tion, mental health, and substance use among women. She has a particular interest in understanding the intersection of intimate partner and sexual assault victimization, the health and mental health manifestations of victimization, help-seeking, and the justice system response to intimate partner and sexual assault victimization. She also has a

Indicators of Recovery at Intake 1799

particular interest in intimate partner stalking. Dr. Logan has coauthored several books including Women and Victimization: Contributing Factors, Interventions, and Implications and Partner Stalking: How Women Respond, Cope, and Survive.

Glossary

Recovery: Recovery as used in this study refers to abstinence. It is in contrast to another group of clients in this study who are defined as being in harm reduction with reduced substance use at follow-up.

Recovery intent: This is a new concept that is not synonymous with motivation, which is a more complex construct. Recovery intent, as used in this study, refers to clients’ vision of intended outcome as expressed as chances of becoming and remaining substance free.

References

Babor, T. F., Stephens, R. S., Marlatt, A. (1987). Verbal report methods in clinical research on alcoholism: Response bias and its minimization. Journal of Studies on Alcoholism 48:410–424.

Cacciola, J. S., Dugosh, K., Foltz, C., Leahy, P., Stevens, R. (2005). Treatment outcomes: First time versus treatment-experienced clients. Journal of Substance Abuse Treatment 28(Suppl.):S13– S22.

Cavaiola, A. A., Strohmetz, D. B., Abreo, S. D. (2007). Characteristics of DUI recidivists: A 12-year follow-up study of first time DUI offenders. Addictive Behaviors 32:855–861.

Dennison, S. J. (2005). Substance use disorders in individuals with co-occurring psychiatric dis- orders. In J. H. Lowinson, P. Ruiz, R. B. Millman, & J. G. Langrod (Eds), Substance abuse: A comprehensive textbook (4th ed.) (pp. 904–913). Philadelphia, PA: Lippincott Williams & Wilkins.

DeYoung, D. J. (1997). An evaluation of the effectiveness of alcohol treatment, driver license actions and jail terms in reducing drunk driving recidivism in California. Addiction 92:989–997.

Dodge, R., Sindelar, J. Sinha, R. (2005). The role of depression symptoms in predicting drug abstinence in outpatient substance abuse treatment. Journal of Substance Abuse Treatment 28:189–196.

Farabee, D., Leukefeld, C. G. (2001). Recovery and the criminal justice system. In, F. M. Tims, C. G. Leukefeld, & J. J. Platt, (Eds), Relapse and recovery in addictions (pp. 40–59). New Haven, CN: Yale University Press.

Fletcher, B. W., Tims, F. M., Brown, B. S. (1997). Drug Abuse Treatment Outcome Study (DATOS): Treatment evaluation research in the United States. Psychology of Addictive Behaviors 11:216– 229.

Floyd, A. S., Monahan, S. C., Finney, J. W., Morley, J. A. (1996). Alcoholism treatment outcome studies, 1980–1992: The nature of the research. Addictive Behaviors 21:413–428.

Flynn, P. M., Craddock, S. G., Hubbard, R. L., Anderson, J., Etheridge, R. M. (1997). Methodolog- ical overview and research design for the Drug Abuse Treatment Outcome Study (DATOS). Psychology of Addictive Behaviors 11:230–243.

Gregoire, T. K., Burke, A. C. (2004). The relationship of legal coercion to readiness to change among adults with alcohol and other drug problems. Journal of Substance Abuse Treatment 26:337–343.

Hiller, M. L., Knight, K., Rao, S. R., Simpson, D. D. (2002). Assessing and evaluating mandated cor- rectional substance-abuse treatment. In C. G. Leukefeld, F. Tims, & D. Farabee (Eds.), Treatment of drug offenders (pp. 41–56). New YorkSpringer.

Holcomb, W. L., Chaiworapongsa, T., Luke, D. A., Burgdorf, K. D. (2001). An odd measure of risk: Use and misuse of the odds ratio. Obstetrics and Gynecology 98:685–688.

1800 Walker et al.

Holroyd, K. A., O’Donnell, F. J., Stensland, M., Lipchik, G. L., Cordingley G. E., Carlson, B. W. (2001). Management of chronic tension-type headache with tricyclic antidepressant medication, stress management therapy, and their random combination. Journal of the American Medical Association 285:2208–2214.

Hser, Y. I., Evans, E., Huang, D., Anglin, D. M. (2004). Relationship between drug treatment services, retention, and outcomes. Psychiatric Services 55:767–774.

Ilgen, M. A., McKellar, J., Moos, R., Finney, J. W. (2006). Therapeutic alliance and the relationship between motivation and treatment outcomes in patients with alcohol use disorder. Journal of Substance Abuse Treatment 31:157–162.

Joe, G. W., Simpson, D. D., Dansereau, D. F., Rowan-Szal, G. A. (2001). Relationships between counseling rapport and drug abuse treatment outcomes. Psychiatric Services 52:1223–1229.

Kelly J. F., Finney J. W., Moos, R. (2005). Substance use disorder patients who are mandated to treatment: Characteristics, treatment process, and 1- and 5-year outcomes. Journal of Substance Abuse Treatment 28:213–223.

Kosten, T. R., Rounsaville B. J., Kleber, H. D. (1983). Concurrent validity of the Addiction Severity Index. Journal of Nervous and Mental Disorders 171:97606–97610.

Leukefeld, C. G., Tims, F. M., Platt, J. J. (2001). Future directions in substance abuse relapse and recovery. In F. M. Tims, C. G. Leukefeld, & J. J. Platt (Eds.), Relapse and recovery in addictions (pp. 401–410). New Haven, CN: Yale University Press.

Logan, T., Hoyt, W., McCollister, K., French, M., Leukefeld, C., Minton, L. (2004). Economic eval- uation of drug court: Methodology, results, and policy implications. Evaluation and Program Planning 27(4):381–396.

McCaul, M. E., Svikis, D. S., Moore, R. D. (2001). Predictors of outpatient treatment retention: Patient versus substance use characteristics. Drug and Alcohol Dependence 62:9–17.

McLellan, A. T. (1983). Patient characteristics associated with outcomes. In J. R. Cooper, F. Altman, B. S. Brown, & D. Czechowicz (Eds.), Research on the treatment of narcotic addiction: State of the art (pp. 500–529). Rockville, MD: U.S. Department of Health and Human Services.

McLellan, A. T., Alterman, A. I., Metzger, D. S., Grisson, G. R., Woody, G. E., Luborsky, L., et al. (1994). Similarity of outcome predictors across opiate, cocaine, and alcohol treatments: Role of treatment services. Journal of Consulting and Clinical Psychology 62:1141–1158.

McLellan, A. T., Chalk, M., Bartlett, J. (2007). Outcomes, performance, and quality—What’s the difference? Journal of Substance Abuse Treatment 32:331–340.

McLellan, A. T., Kushner, H., Metzger, D., Peters, R., Smith, I., Grissom, G., et al. (1992). The fifth edition of the Addiction Severity Index. Journal of Substance Abuse Treatment 9:199–213.

Miller, N. S., Flaherty, J. A. (2000). Effectiveness of coerced addiction treatment (alternative conse- quences): A review of the clinical research. Journal of Substance Abuse Treatment 18:9–16.

Miller, W. R., Rollnick, S. (2002). What is motivational interviewing? In W. R. Miller & S. Rollnock (Eds.), Motivational interviewing: Preparing people for change (2nd ed.) (pp. 33–42).New York: The Guilford Press.

Mokdad, A. H., Bowman, B. A., Ford, E. S., Vinicor, F., Marks, J. S., Koplan, J. P. (2001). The continuing epidemics of obesity and diabetes in the United States. Journal of the American Medical Association 286:1195–1200.

Moos, R. H., Moos, B. S. (2003). Long-term influence of duration and intensity of treatment on previously untreated individuals with alcohol use disorders. Addiction 98:325–337.

Moos, R. H., Moos, B. S. (2006). Participation in treatment and Alcoholics Anonymous: A 16-year follow-up of initially untreated individuals. Journal of Clinical Psychology 62:735–750.

Moos, R., Schaefer, J., Andrassy, J., Moos, B. (2001). Outpatient mental health care, self-help groups, and patients’ 1-year treatment outcomes. Journal of Clinical Psychology 57:273–287.

Morley, J. A., Finney, J. W., Monahan, S. C., Floyd, A. S. (1996). Alcoholism treatment outcome studies, 1980—992: Methodological characteristics and quality. Addictive Behaviors 21:429– 443.

Moyer, A., Finney, J. W., Swearingen, C. E. (2002). Methodological characteristics and quality of alcohol treatment outcome studies, 1970–98: An expanded evaluation. Addiction 97:253–263.

Indicators of Recovery at Intake 1801

Mykletun, A., Bjerkeset, O., Dewey, M., Prince, M., Overland, S., Stewart, R. (2007). Anxiety, depression, and cause-specific mortality: The HUNT Study. Psychosomatic Medicine 69:323– 331.

Ninonuevo, F., Hoffman, N. G. (1993). DUI arrestees versus other outpatients in chemical dependency treatment: Initial status and 1-year outcome. International Journal of Offender Therapy and Comparative Criminology 37:177–186.

Osborne, J. W. (2006). Bringing balance and technical accuracy to reporting odds ratios and the results of logistic regression analyses. Practical Assessment, Research, & Evaluation, 11(7). Retrieved April 25, 2007, from http://pareonline.net/getvn.asp?v=11&n=7.

Pereira, M. A., Jacobs, D. R., Van Horn, L., Slattery, M. L., Kartashov, A. I., Ludwig, D. S. (2002). Dairy consumption, obesity, and the insulin resistance syndrome in young adults. Journal of the American Medical Association 287:2081–2089.

Ritsher, J. B., Moos, R. H., Finney, J. W. (2002). Relationship of treatment orientation and continuing care to remission among substance abuse patients. Psychiatric Services 53:595–601.

Rounsaville, B. J., Dolinsky, Z. S., Babor, T. F., Meyer, R. (1987). Psychopathology as a predictor of treatment outcome in alcoholics. Archives of General Psychiatry 44:505–513.

Rutherford, M. J., Cacciola, J. S., Alterman, A. I., McKay, J. R., Cook, T. G. (2000). Contrasts between admitters and deniers of drug use. Journal of Substance Abuse Treatment 18:343–348.

Substance Abuse and Mental Health Services Administration. (2007). SAMHS stat data. Retrieved on June 19, 2007, from http://www.nationaloutcomemeasures.samhsa.gov/.

Substance Abuse and Mental Health Services Administration. (2006). Results from the 2005 national survey on drug use and health: National findings. Rockville, MD: Department of Health and Human Services.

Sinha, R., Schottenfeld, R. (2001). The role of comorbidity in relapse and recovery. In F. M. Tims, C. G. Leukfeld, & J. J. Platt (Eds.), Relapse and recovery in addictions (pp. 172–207). New Haven, CT: Yale University Press.

Slaymaker, V. J., Owen, P. L. (2006). Employed men and women substance abusers: Job troubles and treatment outcomes. Journal of Substance Abuse Treatment 31:347–354.

Swearingen, C. E., Moyer, A., Finney, J. W. (2003). Alcoholism treatment outcome studies, 1970– 1998: An expanded look at the nature of the research. Addictive Behaviors 28:415–436.

Timko, C., Moos, R. H., Finney, J. W., Connell, E. G. (2002). Gender differences in help-utilization and the 8-year course of alcohol abuse Addiction 97:877–889.

Tims, F. M., Leukefeld, C. G., Platt, J. J. (2001). Relapse and recovery. In F. M. Tims, C. G. Leukefeld, & J. J. Platt (Eds.), Relapse and recovery in addictions (pp. 3–17). New Haven, CN: Yale University Press.

Vaillant, G. E. (1988). What can long-term follow-up teach us about relapse and prevention of relapse in addiction? British Journal of Addiction 83:1147–1157.

Zhang, J., Yu, K. F. (1998). What’s the relative risk? A method of correcting the odds ratio in cohort studies of common outcomes. Journal of the American Medical Association 280:1690–1691.