درمان جامعه درمانی برای افراد سوءمصرف کننده مواد با اختلال شخصیت ضد اجتماعی

Abstract This study compared treatment outcomes of substance abusers with and without antisocial personality disorder (APD) randomly assigned to two therapeutic communities, differing primarily in length of inpatient and outpatient treatment. We hypothesized that APD clients would be less likely to complete treatment, more likely to test positive for drugs and recidivate at follow-up, and that APD clients in the Standard program would have more favorable outcomes than those in the Abbreviated Inpatient program, because of the Standard program's longer inpatient treatment. Self-reports and objective measures of criminal activity and substance abuse were collected at pre- and posttreatment interviews. APD clients were as likely to complete treatment as other clients, and they exhibited the same patterns of reduced drug use and recidivism as did non-APD clients. Treatment program attended was unrelated to outcomes. Substance abusers diagnosed with APD can benefit from treatment in a therapeutic community combined with outpatient care.

Introduction There are several reasons why it is important to learn more about whether a diagnosis of antisocial personality disorder (APD) is related to drug abuse treatment outcomes. First, there is a 40 to 50% prevalence rate of APD in samples of male substance abusers, and approximately 90% of persons diagnosed with APD are substance-abusing criminal offenders Forrest 1992, Gerstley et al. 1990 and Tims, DeLeon, & Jainchill 1994. Second, there is some question regarding the effectiveness of treatment for substance abusers diagnosed with APD. Behavioral characteristics of persons with APD are irresponsibility, reckless/destructive behavior, and criminal activity (Blackburn, 1993). In addition, people with APD are characterized as self-serving, not self-reflective, and may not benefit from the therapeutic community (TC) drug treatment philosophy Abram 1989, Evans & Sullivan 1990, Forrest 1992 and Hare & McPherson 1984. Clients with APD are said to lack the motivation necessary to remain in treatment, and to need at least a year in treatment to reduce drug use and recidivism (Condelli & Hubbard, 1994). Some researchers contend that clients with APD need specialized behavioral approaches in treatment to improve outcomes and have reported, “preliminary results [that] support the viability of reducing drug use in antisocial opioid abusers with a structural behavioral intervention using positive and negative contingencies delivered in a timely manner” (Brooner, Kidorf, King, & Bigelow, 1996, p. 323). Empirical data in this area is greatly lacking, and no data have explored this relationship, in TCs, with random assignment of respondents. Third, the recurring association between APD, substance abuse, and crime is an important social problem. Since 1965, studies have shown that psychiatric patients with a diagnosis of APD were arrested more often than members of the general population, and for more violent crimes Harry & Steadman 1988, Mungas 1983 and Steadman, Cocozza, & Melick 1978. Evidence from samples of incarcerated offenders has also shown that inmates diagnosed with APD have a history of more violent crimes, as compared with other inmates (Hare & McPherson, 1984). Abram (1989) suggests that antisocial substance abusers are committing the most and the worst crimes. While the specific frequency of APD among offending populations is unknown, estimates vary from 20 to 80%, depending on sampling techniques and assessment instruments Abram 1989 and Nigg & Goldsmith 1994. Past and present literature often suggest the relationship between APD and crime seems stronger among men before the age of 40, in the presence of a secondary diagnosis of substance abuse, and for violent rather than nonviolent crime (Hare & McPherson, 1984). Men are also more frequently diagnosed with APD than women (American Psychological Association, 1987). In view of the prevalence of APD among substance-abusing offenders, it is critical that effective treatment strategies are identified for persons with these co-occurring disorders. This article compares treatment outcomes for clients diagnosed with APD with other clients who entered two residential TCs followed by community outpatient care. The data come from the District of Columbia Treatment Initiative (DCI) study, an experiment designed to test the efficacy of providing enhanced inpatient or outpatient drug treatment to clients seeking treatment in Washington, DC. A more detailed description of the DCI appears in Wish, Hoffman, and Nemes (1997), and Hoffman et al. (1995). The current study focuses on the subsample of DCI clients who underwent psychological diagnoses and were randomly assigned to one of two residential TCs. The treatment programs are two-inpatient TC facilities managed by Second Genesis, Inc., a group that has administered TC treatment for the past 25 years. The two treatment facilities were called “Standard” or “Abbreviated Inpatient” programs. The Standard treatment program was designed to reflect TC treatment that was customarily available in the United States, consisting of approximately 10 months of inpatient care followed by 2 months of outpatient care. The Abbreviated Inpatient treatment program provided 6 months of inpatient care, 6 months of outpatient care, and a wide range of extra services. An extensive follow-up study was conducted with the clients assigned to the two TC programs, at an average of 19 months postdischarge. Based on the prior research literature, we hypothesized that clients diagnosed with APD would be less responsive to the TC treatment than the other clients who entered the program. Four hypotheses were tested: Compared with clients without APD, (a) APD clients will be less likely to complete treatment; (b) APD clients will be more likely to test positive for drugs at follow-up; (c) APD clients will be more likely to recidivate postdischarge; and (d) APD clients in the Standard treatment program will have more favorable outcomes than those in the Abbreviated Inpatient treatment program, because of the longer phase of inpatient treatment at the Standard facility.

Results Almost half (49%) of the clients assigned to each program had a SCID diagnosis of APD. Previous analyses, not shown here, revealed that there were no differences between clients with no disorder and clients with disorders other than APD, with regard to the three outcome variables examined in this paper. To simplify our presentation, the analyses in this paper combine clients with no disorders and those with other disorders excluding APD into the no APD group. Fifty-one percent of the clients diagnosed with APD were in the Standard program and 49% were in the Abbreviated Inpatient program. Consistent with previous literature, clients diagnosed with APD were more likely to be male (80% vs. 66%, p < .01), younger at admission (mean age = 31.5 vs. 33.0, p < .05), and more likely to have less education (mean = 10.6 years vs. 11.3 years, p < .01). We found no differences in ethnicity, marital status, employment, or the presence of another psychiatric disorder (41% with no APD had another disorder; 47% with APD had another disorder). About one quarter of the APD and no APD clients had a diagnosis of depression. To assess the construct validity of the SCID diagnosis of APD, we compared APD and no APD clients on our measures of deviant behavior from interviews and official records. Clients with APD had more arrests before treatment (mean = 9.2 arrests vs. 7.1, p < .01) and were first arrested at a younger age (mean = 17.9 years vs. 22.5, p < .01). Seventy-seven percent of the APD group was under some form of criminal justice supervision at admission (primarily probation), compared with 59% of the non-APD group (p < .01). While both groups of clients were likely to be dependent on cocaine with or without heroin at admission, clients with APD were more likely to have multiple drug dependencies (mean = 3.3 vs. 2.5, p < .01) and were more likely to have used needles to inject drugs (38% vs. 27%, p < .05). The APD group was also more likely to have first used alcohol and marijuana at younger ages (mean age first alcohol = 13.6 vs. 15.3, p < .01; mean age first marijuana = 14.5 vs. 15.9, p < .01). These findings show that the clients diagnosed with APD exhibited the expected patterns of greater deviant behavior (see Table 1). Table 1. The Relationship of Antisocial Personality Disorder (APD) Diagnosis to Other Deviant Behaviors (N = 338) a Characteristics No APD (n = 172) APD (n = 166) Mean number of prior arrests (official arrest data) 7.1** 9.2** Mean age at first arrest (self-report) 22.5** 17.9** Under Criminal Justice System supervision at admission b (official arrest data) 59%** 77%** SCID drug diagnosis (hierarchical c) Other d 5% 5% Heroin 3% 3% Cocaine 55% 44% Heroin and cocaine 37% 48% Total 100% 100% Mean number of drugs dependent 2.5** 3.3** Used needles to inject drugs 27%* 38%* Mean age first used alcohol 15.3** 13.6** Mean age first used marijuana 15.9** 14.5** SCID = Structured Clinical Interview for DSM-III-R. a Numbers vary slightly because of missing data. b Includes parole, probation, jail, and bail information from Criminal Justice Records. c Hierarchical drug categories may include one or more of the previous dependencies. For example: Clients with a heroin dependency may also have other drug dependencies (excluding cocaine). Clients with a cocaine dependency may also have other drug dependencies (excluding heroin). Clients with a heroin plus cocaine dependency may also have other drug dependencies. Subjects do not appear more than once in the above table. d Includes alcohol, marijuana, hallucinogens, and other drugs. * p < .05. ** p < .01. Table options Bivariate associations between a diagnosis of APD and the three outcome variables were assessed, by treatment program. There were no statistically significant differences in treatment completion, drug use at follow-up, or postdischarge arrest, between APD and no APD clients in the two treatment programs. Clients in the Standard program had outcomes that were virtually identical to those of clients in the Abbreviated Inpatient program. However, the bivariate relationships do not take into account the effects of other factors predictive of treatment outcomes that might mask the effect of APD. We examined several other variables for an association with the three outcome variables. All variables that were significantly related to an outcome variable were controlled for in subsequent logistic regression equations. The next section tests each of our hypotheses regarding the relationship of APD to treatment outcomes, while controlling for other factors related to these outcomes. Hypothesis 1: APD and Treatment Completion Table 2 presents the analysis of the relationship of APD to treatment completion. Contrary to our hypothesis, after controlling for related variables, a diagnosis of APD was not significantly related to treatment completion. Clients with APD were as likely to complete treatment as those without APD. As expected, older clients and clients who were under some form of criminal justice supervision at admission were more likely to complete treatment. Clients with a higher number of prior arrests were less likely to complete treatment. Clients who were dependent on heroin alone were more likely to complete treatment than those dependent on only alcohol, marijuana, or hallucinogens. The presence of other disorders, with APD or without APD, was not significantly related to treatment completion. Program attended was also not related to treatment completion. Table 2. Logistic Regression Assessing Treatment Completion (N = 330) Variable B SE df Sig. Exp(B) SCID diagnosis [No APD] APD .4186 .2579 1 .1045 1.5199 Other disorders [No other disorders] Other disorders occur −.0545 .2586 1 .8331 .9470 Gender [Female] Male −.3158 .2923 1 .2799 .7292 Age at admission .0482 .0208 1 .0205* 1.0494 Years of education −.0700 .0632 1 .2679 .9324 Number prior arrests −.0391 .0203 1 .0541* .9617 Criminal status at admission [No] Yes .6338 .2894 1 .0285* 1.8848 SCID drug diagnosis at admission 3 .0477* [Alcohol/marijuana/PCP] Heroin 2.8345 1.1974 1 .0179* 17.0227 Cocaine/crack −.2149 .5618 1 .7020 .8066 Heroin and cocaine −.0637 .6059 1 .9163 .9383 Number drugs dependent .1375 .0967 1 .1551 1.1474 Treatment sites [Standard] Abbreviated inpatient .1079 .2384 1 .6508 1.1140 Constant −1.7220 1.0956 1 .1160 Note. [brackets] indicate reference category. SCID = Structured Clinical Interview for DSM-III; APD = Antisocial personality disorder. * p ≤ .05. Table options Criminal status at admission, total number of prior arrests, and number of drug dependencies are, by definition, related to a diagnosis of APD. To determine if their correlation with APD was diminishing the effect of APD on treatment completion; we repeated each model excluding these three variables. Removing these variables from the regression models made no difference in outcomes, with the exception of the treatment completion model. Contrary to our hypothesis, clients with APD were more likely to complete treatment when these variables were removed from the model. Hypothesis 2: APD and Positive Urinalysis at Follow-Up Time in treatment has been shown to be a strong predictor of treatment outcomes. We, therefore, first looked at the relationship of APD to a positive (for any drug) urinalysis and postdischarge arrest, while controlling for treatment completion. We found no differences between APD and non-APD clients with regard to drug use and recidivism. However, as expected, persons who completed treatment were less likely to test positive for drugs (28% vs. 55%, p < .01) and to recidivate at follow-up (29% vs. 62%, p < .01). The effect of treatment completion was so great that we tested our remaining hypotheses by including in our models a variable that coded the four possible combinations of treatment completion and APD diagnosis (see Table 3). This variable allowed us to test, for example, whether APD clients who completed treatment had similar outcomes as no APD clients who completed treatment. Table 3. Treatment Outcomes by Treatment Completion and Antisocial Personality Disorder (APD) Diagnosis Not Completed Completed Characteristics No APD APD Total No APD APD Total Positive urine test (74) 53% (58) 59% (132) 55%* (53) 32% (69) 25% (122) 28%* Postdischarge arrest (113) 63% (90) 61% (203) 62%* (59) 25% (76) 32% (135) 29%* * p < .01. Table options The only significant predictor of a positive (primarily for cocaine) urinalysis at follow-up was treatment completion. Completing treatment appeared to be associated with reductions in drug use for all clients, regardless of a diagnosis of APD. Moreover, the presence of a psychological disorder other than APD was not related to a positive urinalysis. Hypothesis 3: APD and Postdischarge Arrest Because the time between treatment discharge and the cut-off date for obtaining criminal justice record information varied considerably among our sample members, we controlled for time at risk by adding a variable to our model that reflected the number of months between client discharge and the last date for which arrest information was obtained from criminal records. Table 4 shows that clients both with or without APD who completed treatment had marked reductions in the likelihood of a postdischarge arrest. Clients dependent on multiple drugs at admission were also less likely to recidivate. Men and younger clients were more likely to recidivate, as were clients with many arrests before treatment and clients under some form of criminal justice supervision at admission. Clients who attended the Abbreviated Inpatient treatment facility were also more likely to recidivate. The occurrence of psychological disorders other than APD was not significantly related to postdischarge arrest. Table 4. Logistic Regression Assessing Recidivism (N = 330) Variable B SE df Sig. Exp(B) APD × treatment status 3 .0001** [No APD/dropped out] No APD/completed −1.4419 .4444 1 .0012** .2365 APD/dropped out −.5206 .3604 1 .1487 .5942 APD/completed −1.7824 .4130 1 .0001** .1682 Other disorders [No other disorders] Other disorders occur .3552 .3006 1 .2373 1.4265 Gender [Female] Male .7481 .3363 1 .0261* 2.1131 Age at admission −.1173 .0270 1 .0001** .8894 Years of education −.0186 .0722 1 .7971 .9816 Criminal status at admission [No] Yes .6981 .3174 1 .0278* .4975 Number prior arrests .0606 .0228 1 .0078** 1.0625 SCID drug diagnosis at admission 3 .6247 [Alcohol/marijuana/PCP] Heroin .3703 1.0263 1 .7183 1.4481 Cocaine .0105 .7147 1 .9883 1.0105 Heroin and cocaine .4329 .7639 1 .5709 1.5417 Number drugs dependent −.2333 .1125 1 .0382* .7919 Marital status 2 .4714 [Never married] Married/living together −.1804 .3852 1 .6392 .8349 Divorced/separated −.5221 .4290 1 .2236 .5932 Treatment sites [Standard] Abbreviated inpatient .6415 .2696 1 .0173** 1.8993 Number of months to criminal justice coding .0623 .0201 1 .0020** 1.0643 Constant 2.7936 1.3632 1 .0404* Note. [brackets] indicate reference category. * p < .05. ** p < .01. Table options Hypothesis 4: Treatment Program Effects for APD Clients Our fourth hypothesis stated that APD clients assigned to the Standard program would have better outcomes than APD clients assigned to the Abbreviated Inpatient program. The bivariate results had shown no differences in treatment outcomes for APD clients in the two treatment programs. We performed logistic regressions with the sample of only APD clients to compare outcomes of Standard and Abbreviated Inpatient program clients, while controlling for other variables related to these outcomes. The separate regression analyses, not presented, revealed that APD clients randomly assigned to the Standard program were as likely to complete treatment as APD clients in the Abbreviated Inpatient program. Moreover, APD clients who completed treatment in the Standard and Abbreviated Inpatient programs had no differences in their likelihood of testing positive for drugs at follow-up or having a postdischarge arrest.