|
Article Excerpt
DROPOUT RATES FROM SUBSTANCE-USE disorder (SUD) treatment programs range from 10% to 30% depending on the sample and the setting (De Leon, 1991; Marlatt et al., 1997; Mertens and Weisner, 2000; Rabinowitz and Marjefsky, 1998). Most studies of dropouts from SUD treatment programs focus on predicting dropout status from baseline patient variables, especially demographic and disorder severity characteristics. The ultimate purpose of this research is to identify patients at high risk of dropout, develop interventions to increase the likelihood that they will stay in treatment, and provide clinical managers with an empirical basis for training patient care staff and altering clinical practice.
One factor often overlooked in this literature, however, is the limited nature of information about what happens to patients after they drop out of treatment. Most studies examine only the power of baseline variables to predict dropout and do not obtain follow-up data on patients after they leave treatment. When subsequent data on patients who have dropped out of treatment are collected, it is typically in the context of treatment re-engagement. We do not know of any prior long-term outcome study of dropouts who do not return to treatment. Do such dropouts experience deterioration, improvement, or stability in the course of their SUD? How do the trajectories of change among dropouts compare with those among treatment completers? Further, is it possible to predict differential outcomes of dropouts who do not return to treatment?
Knowing more about which factors predict long-term outcome for dropouts who do not return to treatment could be helpful in guiding SUD interventions before and after dropout. This information could be used to help prioritize clinical activities to ensure that such factors are addressed in the earliest stages of treatment, before dropout occurs. Also, such data could aid in identifying high-risk dropouts for more assertive treatment re-engagement efforts.
Despite the lack of research on the long-term outcomes of patients who drop out and remain out of treatment, two groups of studies are relevant to our project: (1) studies of predictors of dropout (Stark, 1992) and (2) studies of predictors of outcome among untreated individuals. Previous research has identified some of the predictors of treatment dropout, including the following: younger age (Joe et al., 1999; Leigh et al., 1984; Mammo and Weinbaum, 1993; Marlatt et al., 1997; Rabinowitz and Marjefsky, 1998; Siqueland et al., 1998), lower education (Mammo and Weinbaum 1993; Siqueland et al., 1998), lower levels of motivation for treatment (Cahill et al., 2003; Joe et al., 1999; Ryan et al., 1995; Simpson and Joe, 1993), greater severity of substance-use problems and levels of dependence (Joe et al., 1999; Rees et al., 1984; Ryan et al., 1995), and poorer cognitive functioning (Erwin and Hunter, 1984). Although these findings may be helpful in guiding clinical attempts to reduce dropout, it is unknown whether the variables associated with dropout predict how well patients do after they drop out of treatment.
To the extent that patients who drop out of treatment represent a sample of minimally treated individuals, longitudinal studies of untreated individuals with SUDs and general population studies may also be relevant. Predictors of better outcomes for untreated individuals include less severe SUDs (Dawson et al., 2005; Moos and Moos, 2005; Weisner et al., 2003), self-efficacy beliefs (Moos and Moos, 2003), and level of stressors and support from family members and friends (Dawson et al., 2005; Moos and Moos, 2005; Schutte et al., 2003; Weisner et al., 2003). In addition, participation in informal self-help groups, such as Alcoholics Anonymous or Narcotics Anonymous (Humphreys et al., 1995), predict outcomes in untreated individuals and may predict outcome for individuals who drop out of treatment.
The current study examines the long-term outcomes of a group of patients who dropped out of and did not return to SUD treatment. We focused on three questions: (1) Compared with treatment completers, for those patients who drop out of and do not return to treatment, how well do they do at follow-ups during a 5-year period in terms of substance-related problems? (2) Among patients who drop out of treatment, what characteristics predict individual differences in substance-related problems 5 years after leaving treatment? (3) Among patients who drop out of treatment, what characteristics predict the individual variability in change in substance-related problems during the course of 5 years (slope analysis)?
Method
Participants
All male patients at 15 geographically diverse Department of Veterans Affairs (VA) SUD residential treatment programs were medically detoxified and invited to participate in an evaluation of treatment effectiveness. The treatment programs were designed to last 28 days, used individual and group therapy to assist patients in meeting their treatment goals, and were multidisciplinary in nature. In each program, consecutive admissions were approached, unless it was determined that the patient volume would be in excess of data-collection capabilities. If so, a sampling procedure was implemented in which every other admission or every third admission was recruited. A total of 4,192 patients were invited to participate (90% of those eligible); the other 10% left the program before completing detoxification treatment or were not invited to participate because of scheduling problems. A total of 494 patients refused to participate, leaving 3,698 patients enrolled at intake. A subsample of 49 of these patients was transferred to a different treatment setting because of medical instability, leaving 3,649 patients who provided pretreatment data. A total of 445 individuals (12% of the total sample) dropped out of treatment, that is, left the program before completion. Because of the current study's focus on dropouts who left treatment and did not return, the 252 patients who dropped out and returned to treatment were excluded, leaving 3,397 patients who provided pretreatment data.
This study focuses on the 193 individuals who dropped out and never returned to either inpatient or outpatient VA treatment during the 5 years after dropping out of the index treatment episode. Dropouts were defined by program staff as those who left treatment before completion of the prescribed length of treatment (approximately 28 days). In initial analyses, the intake characteristics of these 193 dropouts are compared with those of the 3,204 participants who completed treatment.
Procedure
Patients who completed an intake questionnaire were asked to provide follow-up information 1, 2, and 5 years after leaving treatment. Data-collection procedures included a combination of interview and self-report occurring either...
|
