Methadone maintenance: Predictors of outcome in a Canadian milieu

Methadone maintenance: Predictors of outcome in a Canadian milieu

Journal of Substance Abuse Treatment, Vol. 10, pp. W-89, 0740-5472/93$6.00 + .W Copyright 0 1993Pergamon Press Ltd. 1993 Printed in the USA. All r...

517KB Sizes 0 Downloads 18 Views

Journal of Substance Abuse Treatment,

Vol. 10, pp. W-89,

0740-5472/93$6.00 + .W Copyright 0 1993Pergamon Press Ltd.

1993

Printed in the USA. All rights reserved.

INTERNATIONAL

PERSPECTIVE

Methadone Maintenance: Predictors of Outcome in a Canadian Milieu FRANCOIS LEHMANN, MD, CCFP ,* PIERRE LAUZON, MD,*? AND RHONDA AMSEL, MS~$ *Department of Family Medicine, University of Montreal; tCentre for Research and Treatment of Narcotic Addiction, Montreal, Canada; $Department of Psychology, McGill University

Abstract - This hypothesis-generating study attempts to identify patient characteristics predicting the successful outcome of methadone maintenance in the treatment of narcotic addiction. The sociodemographic characteristics as well as the general emotional health of 51 addicted individuals in Montreal were studied at entry into the program and are correlated with success one year after entry. The success of the Montreal program, defined as retention in the program and urine samples negative for opiates, is comparable to results reported from many North American clinics. A history offoster care before age 15 is the only characteristicsigntficantly correlated with both retention and negative urines. Emotional health, previous treatment experiences, job status, and level of education do not correlate with success. All addicted individuals applying to a treatment program should therefore be accepted for a trial of treatment while further research on greater numbers of subjects may eventually identify predictors of successful outcome.

Keywords-methadone;

outcome;

addiction.

INTRODUCTION

Schmeiller, 1976) reports that only the number of prior convictions and job status predict retention in the program (his definition of success). Similarly Simpson (Simpson dz Savage, 198l-1982) finds that addicted individuals with greater criminal involvement showed less favorable outcome. On the other hand Kosten et al. (Kosten, Rounsaville, & Kleber, 1987) reports that patients whose income was from criminal activities did better in finding jobs and avoiding legal problems than those whose income derived from welfare. The latter fared better in the areas of medical and psychiatric status. One paper suggests that the outcome of those with severe family or employment problems is improved in inpatient settings and that prior drug abuse severity is not generally important in predicting outcome (McLelIan, 1986). Others have suggested that those with a high degree of psychiatric dysfunction, especially those with a major depression, do not do well on methadone maintenance (Rounsaville, Kosten, Weissman, & Kleber, 1986; Sutker, Cohen, 8z Allain, 1976). The differing conclusions may be explained by the different definition each author gives to success, by the quality of different programs, by the rules in use for administrative

as a treatment modality for opiate addiction has been used since the 196Os, and there have been numerous reports of its effectiveness especially in the areas of further opiate use and criminal activity. As a result, this treatment has received wide acceptance in many countries and has been the subject of much study, especially in the United States. Most of these studies focus on the different treatment modalities used by various methadone treatment programs and on comparisons with other treatment modalities such as therapeutic communities. Some research also addresses outcomes and usually correlates outcomes to the type of treatment given or to the length of time the addict remained in treatment. Few studies look at the addicted individual’s past history before treatment as a predictor of success. Those who have addressed this last question do not report the same findings. Babst (Babst, Ellis, &

METHADONE MAINTENANCE

Requests for reprints should be addressed to Francois Lehmann, Centre Hospitalier de Verdun, 4000 Boul. Lasalle, Verdun (Quebec), H4G 2A3 Canada.

85

86

F. Lehmann et al.

termination, and sometimes by the different populations studied, though most were mainly composed of men, many of them black, from the inner city core, or of veterans. It is the authors’ (EL., P.L.) clinical impression after 15 years of experience in the treatment of addicted individuals, that success, as evidenced by retention in a treatment program and abstinence from opiates, is difficult if not impossible to predict. Since American studies differ in their findings, and since it has not been shown that the results of these studies can be applied to a Canadian population of addicted individuals which is demographically different, exclusively white, 30% female, and residing outside the inner city core, the present study addresses the following question: Can one generate hypotheses about which factors predict outcome for the Montreal addict population enrolled in a methadone maintenance program?

METHODS Fifty one addicted individuals were admitted to the Montreal program between May 1986 and January 1987. All applicants were accepted if they were older than 18 years and had regularly used opiates for 2 years or more. Within a month of admission, a sociodemographic questionnaire was completed. From this entry information, basic data such as age, sex, and education, items that other studies had found important and items that are clearly related to mental health were chosen as independent variables. An item on previous methadone treatment was included to ascertain whether past treatment failure would predict poor outcome. These items are listed in Table 1. There was no test of the reliability of this particular questionnaire but many authors have shown that history taking with addicts is reliable (Mad-

TABLE 1 Variables at Admission Age (mean 3 1 years) 18-24 25-34 35+ Sex Male Female Language English French Other Marital status Legal or common law Education (mean 12 years) 11 years or less College or university Total number of months in prison none 1-2 months 3-5 months 6-l 0 months 1 1 + months Total number of arrests in the 6 months previous to admission no arrests l-2 arrests 3-5 arrests 6- 10 arrests 1 1 + arrests Occupation Work or studies Past treatment in a methadone program Parental abuse before age 16 Placement in an institution or a foster home before age 16 Behavior problems in school before age 16 Learning problems in school before age 16 Loss of a parent through death or abandonment before age 16 Number of days of reported opiate use in past month (range) Score on the Heimler Scale of Social Functioning Below average Above average

10%

58% 32% 68% 32% 34% 46% 20% 18% 72% 28% 39% X = 15.2 overall, n = 51 14% 8% 8% R = 24.2 n = 32 31% SD = 30.22 21% ii = 6.9 overall, 18% 20% x = 8.8, 2l%SD=8.7 20% 60% 76% 40% 18% 50% 24% 40% O-8 55% 45%

n = 51 n = 40

Predictors of Outcome in Canada

dux & Desmond, 1975; Stephens, 1972; Amsel, Mandell, Mattias, Mason, & Hocherman, 1976; Ball, 1967; Cox & Longwell, 1974). The Heimler Scale of Social Functioning was also administered. This scale has been shown to correlate with emotional well-being and social functioning (Burnell & Norfleet, 1982-1983), and its concurrent validity, comparing it with the MPI and a number of other scales, is satisfactory (Rodway, 1977). The items used as dependent or outcome variables were 1-abstinence from heroin (as shown by negative weekly urine samples over the last 6 months of treatment), 2 -retention in the program, 3 -abstinence from cocaine and benzodiazepines. Abstinence from heroin was chosen because as Newman reminds us, methadone maintenance is primarily a treatment of opiate abuse and should be evaluated as such (Newman, 1987). Abstinence from opiates is therefore the gold standard of success. Retention in the program was included because many authors have demonstrated that retention is a condition most often necessary for the achievement of abstinence (Dole & Joseph, 1977). Cocaine and benzodiazepine use were studied because when their use is curtailed in addition to the curtailment of heroin abuse, treatment success is of a much greater quality. For those addicts no longer on the program (all of these were unapproved termination or exclusions because of incarceration, violence, or theft in the clinic), urines were assumed to be positive for all 3 substances (worst possible scenario). The urines were analyzed by the EMIT method. Chi-square or Fisher exact tests compared each independent to each outcome variable. When the independent variable was numeric (e.g., number of arrests) it was grouped into 2 or 3 categories. Since previous studies suggest that combinations of variables might more successfully predict outcome than variables taken alone, 8 combinations were used. These are listed in Table 2. Chi-square tests compared subjects meeting all criteria within each combination to everyone else in the sample with regard to 2 outcome variables-retention in the program and reported abstinence from heroin in the last month of the 1Zmonth study period.

TABLE 2 Combinations of Variables Used to Predict Outcome

Education > 11 years and work or study full or part time Married (or common law) and work or study full or part time Married (or common law) and caring for own child Male and no prior arrests Female and no prior arrests Male and married (or common law) Female and married (or common law) Behavior problems and parental abuse and learning problems

87

RESULTS Thirty seven addicted individuals completed one year in the program, and 14 left or were excluded from the program. Only 5 of the 64 contingency tables on single variables show statistical significance with ap < .05. These are listed in Table 3. Using combinations of variables, there are 16 possible contingency tables. None of the combinations significantly predict abstinence from opiate use or retention in the program. DISCUSSION There is only one standard for measuring the success of narcotic addiction treatment: objective proof that abuse has stopped or has markedly decreased in frequency. In our population of addicts, foster care is the only variable that correlates significantly with the absence of opiates in the addicted individual’s urine. The other accepted measure of success is retention in treatment. Here the x2 with foster care is significant before Yates’ correction, but only marginal when the correction is used. A strong relationship exists between the outcome measures retention in treatment and negative urines because of the supposition that those leaving treatment would test positive. Being of the male sex, working, or being less engaged in criminal activity are each significant predictors of less benzodiazepine use, but are not individually related to the presence or absence of opiates in urine or to retention in the program. This latter outcome is significantly improved if a subject is both of the male sex and has had no prior arrests. This is probably due to the fact that some subjects had to leave the program because they went to prison. Those with no arrests did not have that same risk.

Entry Characterlstlcs

TABLE 3 Slgnlflcantly Related to Outcome

One Year

Urine Negative

Retention

for Opiates

Subjects with foster care 100% (9 of 9) 87% 67% Subjects without foster care 31% (/I; Yjf4) (p = 0.001) (0.004)8 Percentage of urines negative for benzodiazepines Males Females No prior arrest More than 6 arrests Working or studying Not working or studying 8With Yates’ correction.

88% 7% 73% 38% 58% 23%

(p = .004) (O.OOl)a

(p = .Ol) (p = .Ol) (0.02)”

88 It is striking that neither poor emotional health (as reflected by low scores on the Heimler scale), previous treatment experiences, or poor education show significant relation to outcome. Although the risk of Type II error (or false negative) contingency results is around 20% according to tables of statistical power analysis, the results confirm the authors’ clinical experience that one cannot predict which addict will achieve abstinence. One has to consider the very real possibility that, within the context of our present knowledge, there are no reliable predictors of success one year after entry. This is an unpopular opinion, especially for physicians who are trained to offer very specific treatments for each pathological condition - faced with a specific infection, shotgun therapy is frowned upon. But faced by those suffering from heroin addiction, especially by those who wish to enroll in the Montreal methadone program, one should accept the fact that the present state of knowledge does not enable the clinician to make a scientifically based choice. A trial of therapy should be offered to all who apply and meet the legal admission criteria. We hope this will be a short-lived state of affairs, for if predictors of outcome can be found, the physician may be able to target those who might benefit from extra services, such as inpatient treatment, group therapy, or intensive individual therapy. Further studies will also indicate if there are specific predictors for other forms of therapy, particularly therapeutic communities, selfhelp groups, and short-term methadone maintenance or withdrawal. Trying to find variables related to success would be a futile exercise if the Montreal program were a failure. But there was a retention rate of 73%, which is good when compared to the rates reported by American programs. Fifty-eight percent (58%) of all who completed one year produced urines negative for opiates in the last 6 months of treatment, 86% were employed or at school (at least part time) after one year, versus 60% at entry, and 71% had no links with the judicial system compared to 42% at entry. These are criteria utilized by Dole (Dole & Joseph, 1977) and others when looking at success, and the results of the Montreal program demonstrate a level of success at least equal to that reported by the American programs. With a sample size of 50 and 64 contingency tables for single variables as well as 16 tables for combined variables, there is an inflated probability of Type I error, or false significance even when the conservative (Yates’ corrected) test is used. Dunn’s procedure would define significance as a p c .0006. This would effectively disallow the one item that is significant, foster care, a finding that is quite unexpected. On the other hand, the relationship between foster care and success in a methadone program may well be real and could be related to one aspect of the program. As in most American clinics, addicted individuals in the Montreal

E Lehmann et al. clinic are not just handed a prescription or a given quantity of medication. They meet with a counsellor (a nurse or a social worker) initially one hour a week for the first 3 months and then every 2 to 3 weeks if they are doing well. Addicted individuals placed in foster homes may never have had a significant positive relationship with an adult in a care-giving or parental role, and the therapeutic relationship with the counsellor may have had a major impact on them. Of course, the majority of addicted individuals, more than 3,000 in the Montreal area, never apply for methadone treatment. Why they do not apply is not known, but it is possible that methadone treatment appeals only to a particular subset of Montreal’s addicted individuals who already in some way know that their chances of success are fair. Self-selection may in fact have eliminated addicted individuals whose characteristics would have permitted a prediction of outcome. Another explanation could be that women and men have different predictors of success or failure and that analyzing both sexes together renders the correlation nonsignificant. CONCLUSIONS The study reported in this paper has not been able to generate hypotheses for predicting favorable outcomes for addicts on methadone maintenance treatment with the possible exception of placement in foster care. It has suggested that the few predictors listed by some American studies may not be valid for the Montreal population studied. The results do not permit an argument in favor of refusing methadone maintenance to a certain preselected clientele. They do not indicate which addicts should be referred to another treatment modality such as therapeutic community. More importantly the results should remind both government and health care workers that, at least in Montreal, methadone maintenance should be offered to all addicted individuals who desire it and who show a willingness and a capacity to follow the rules of the program. There is no indication for exclusion based on past treatment failure or a history of family or psychiatric problems. Methadone has been clearly shown, both in Canada and elsewhere, to be a successful therapeutic modality for some addicted individuals. As long as no well-founded criteria exist for selecting these addicted individuals, methadone treatment should be offered to all, although about 30% will not be able or willing to follow the rules of the program and will thus exclude themselves. Further research using large samples might succeed in discovering other criteria. REFERENCES Amsel, S., Mandell, W., Mattias, L., Mason, C., & Hocherman, I. (1976). Reliability and validity of self-reported illegal activities

89

Predictors of Outcome in Canada and drug use collected from narcotic addicts. InternationalJournal of the Addictions, 2, 325-336. Babst, D.V., Ellis, R., & Schmeiller, J. (1976). Testing predictive efficiency of patient classifications for methadone maintenance cllents. British Journal of Addiction, 71, 359-367. Ball, J. (1967). The reliability and validity of interview data obtained from 59 narcotic drug addicts. American Journal of Sociology, 12, 650-654.

Burnell, GM., & Norfleet, M.A. (1982-1983). Evaluating psychosocial stress: Preliminary report on a brief and convenient instrument for health professionals. International Journalof Psychiatry in Medicine, 12(2), 141-149. Cox, T., & Longwell, B. (1974). Reliability of interview data concerning current heroin use from heroin addicts on methadone. International Journal of the Addictions,

9, 161-165.

Dole, V., & Joseph, H. (1977). Methadone maintenance: Outcome after termination. New York State Journal of Medicine, 71, 1409-1412.

Kosten, T., Rounsaville, B.J., & Kleber, H.D. (1987). Predictors of 2.5 year outcome in opioid addicts: Pretreatment source of income. American Journal of Drug and Alcohol Abuse, 13 (l&2), 19-32.

Maddux, J., & Desmond, D. (1975). Reliability and validity of information from chronic heroin users. Journal of Psychiatric Research, 12, 87-95.

McLellan, T.A. (1986). “Psychiatric severity” as a predictor of outcome from substance abuse treatments. In R.E. Meyer (Ed.), Psychopathology and addictive disorders (pp. 97-139). New York: Guilford Press. Newman, R.G. (1987). Methadone treatment: Defining and evaluating success. New England Journalof Medicine, 317,447-450. Rodway, M.R. (1977). Validity of the Heimler Scale of Social Functioning In G. de Coq & L. Dick (Eds.), Proceedings of the 1st Conference on Human Social Functioning (pp. 59-66). Calgary, Alberta, Canada: University of Calgary. Rounsaville, B.J., Kosten, T.R., Weissman, M.M., & Kleber, H.D. (1986). Prognostic significance of psychopathology in treated opioid addicts. Archives of General Psychiatry, 43, 739-945. Simpson, D.D., & Savage, J.L. (1981-1982). Client types in different drug abuse treatments: Comparison of follow-up outcomes. American Journal of Drug and Alcohol Abuse, 8, 401-418. Stephens, R. (1972). The truthfulness of addict respondents in research projects. International Journal of the Addictions, I, 549-558.

Sutker, P.B., Cohen, G.H., & Allain, A.N. (1976). Prediction of successful response to multimodality treatment among heroin addicts. International Journal of the Addictions, 11, 861-879.