Ethchlorvynol

INTRODUCTION

Numerous studies have examined the relationship between pretreatment characteristics and retention or performance during the course of methadone maintenance [1]. Variables such as age [2, 3], raceor ethnicity [4], marital status [3, 5], employment history [6, 7], criminal activity [3, 7], psychological status [8, 9], and drug use severity or type [3, 10] have all ben examined. As previously summarized by McLellan [1], "the available data suggest that criminal background, poor work history, use of non-opiate drugs, and possibly greater severity of psychological problems are the major background variables which are predictive of poor retention in and performance during methadone" (p. 510).

Methadone detoxification has been less well examined. A previous small sample (N = 20) study from this laboratory found that methadone detoxification patients exhibited a wide range of involvement with illicit opiates during initial treatment weeks, as indicated by urinalysis test results. A group of high frequency opiate users submitted 92.5% opiate positive tests during two weeks of dose stabilization while a group of low frequency users submitted only 7.5% opiate positive urines. However, this study failed to identify any demographic variables that differentiated these patient groups with different treatment response. The research described in this paper sought to expand the examination of outcome predictors during methadone detoxification by studying a larger group of patients and by including measures of environmental exposure to stimuli and interpersonal situations normally associated with opiate use along with more commonly investigated demographic and pretreatment characteristic variables.

In this study, behavioral, environmental, and demographic variables collected prior to subject entry into a 90-day outpatient methadone detoxification were examined in order to derive predictors of treatment efficacy. By systematically determining the relationship between pretreatment variables and treatment performance and retention, it was hoped that we mightL: 1) determine the relative appropriateness of this treatment protocol for different subgroups of patients, 2) identify characteristics that place patients at risk for treatment failure, and 3) better understand the complex of individual and environmental influences on the treatment process.

METHODS

Subjects

Participants included consecutive new admissions to a 90-day methadone detoxification treatment program at the Behavioral Pharmacology Research Unit's methadone treatment clinic. Subjects were informed by the intake staff that this was a treatment research clinic and that they would be expected to participate in a research protocol. Individuals were excluded from the study if they: 1) were older than 55 or younger than 18 years of age, 2) tested positive for pregnancy, or 3) were taking medication by prescription for a physical or psychological disorder. In all, 71 subjects (55 men and 16 women) participated in the research protocol. Table 1 summarizes the standard demographic and background characteristics of the study participants.

All subjects tested positive for opiates at the time of admission. Eighteen (26%) of the subjects tested positive only by opiates. Fifty-three (75%) of the subjects submitted admission urines which tested positive for at least one other drug. Forty-seven (66%) of the admission urines tested positive for cocaine, while 14 (21%) of the urines tested positive for benzodiazepines. Three subjects (21%) tested positive for PCP. In all, 9 (13%) of the subjects tested positive for more than one substance.

General Procedures

The study was approved by the Francis Scott Key Institutional Review Board for human research and was initiated during April 1987. At the intake appointment, which occurred prior to treatment admission, demographic data were collected, study details were read and explained to each subject, and informed written consent was obtained. One or two weeks generally elapsed between intake and treatment admission, depending on receipt of records from other programs or institutions and counselor availability. During this time, subjects reported to the clinic once weekly to participate in data collection procedures. After treatment admission, subjects were required to report on a daily basis to the clinic

[TABULAR DATA OMITTED]

where they ingested their methadone dose in a cherry syrup vehicle under nursing supervision. All subjects received 40 mg methadone during an initial 3-week stabilization period, and their dose was then gradually decreased by 4mg/week over a 10-week period. Medication was delivered in single-blind fashion with patients unaware of their current dose while medication nurses had knowledge of the dosing regimen. Monitored urines were collected from all subjects on Monday, Wednesday, and Friday of each week and temperature tested to ensure veracity. Each urine was screened on-site by EMIT for evidence of opiate, cocaine, or benzodiazepine use. The urine samples were then sent to an independent laboratory for full screen analysis by thin layer chromatography (TLC) which detects a wide variety of opiate and nonopiate drugs including morphine, codeine, hydromorphone, propoxyphene, diazepam, oxazepam, barbiturates, phenothiazines, hydroxyzine pamoate (Vistaril), ethchlorvynol (Placidyl), and amitriptyline (Elavil).

1. The Interview Process. All interviews were administered by the same research assistant. Subjects were interviewed on a weekly basis from the day of intake until treatment termination, using a semistructured internview. With the exception of the first interview, the interview had complete access to the EMIT urinalysis results from the previous week. Inconsistencies between verbal reports and urinalysis results were noted and clarifying probes were asked, particularly when the subject reported no drug use and had submitted a drug-positive sample. Subjects were compensated for their time in the amount of $5 per interview.

The first section of the interview asked a series of identical questions about each and every episode of opiate use within the previous week (excluding the interview day). Subjects were asked to carefully review the day and time of each opiate use occasion in the previous week, starting with the previous day and working backwards. If a subject appeared to have difficulty remembering the events of the previous week, he was prompted with multiple probes such as, "Did you use any opiates after you got off work yesterday?", or "There was a thunderstorm that day...." On occasion, the interviewer would help the subject to recreate entire days.

2. Opiate Purchases: Source; Drug; When; Where; and Amount Obtained. Subjects were asked to recall drug purchases for the previous week, to name the specific opiate drug obtained and the amount, to recall dollar cost of the purchase, when purchases were made, and from whom using a code name to identify their source.

3. Opiate Use: Amount, Route; Location; and Social Context. For each drug use occasion, subjects were asked to identify the drug used, the amount, route of administration, the location of use, the coded identity and social relationship of any individuals using the drug with them, and a question with probes regarding needle sharing. The measure derived was number of episodes of opiate use per week.

4. Opiate use: adjunct substance use; mood state; ratings of withdrawal and high. Questions regarding illicit drug and alcohol use were asked. A profile of mood state (POMS) adjective checklist and an opiate withdrawal symptom questionnaire were also included in the interview in order to get a general picture of mood state and withdrawal discomfort during the week covered by the interview. Specific questions regarding mood state immediately before and after each drug use episode were asked, as was a question regarding the "high" associated with each period of opiate use. Additional questions were asked about the number of interpersonal conflicts in the previous week and general mood state as well as a rating of urge or drug craving. The data from this section were not included in this analysis.

5. Nonopiate use: frequency of opiate-related interpersonal and situational exposures not resulting in opiate use. The second section of the interview was called the "Risk factor exposure report," and focused on nondrug use situations in which the individual had the opportunity to use opiate drug in the previous week but did not. Subjects were asked to quantify the number of times they received handled opiate drugs without using them. They were also asked about the number of times they did not use opiate drugs when they: 1) were in the same room/location as someone else who was using opiate drugs or drugs/alcohol of any kind, 2) were offered a sample of a drug, and 3) used nonopiate drugs or alcohol. The number of opiate-related stimulus exposures was calculated by adding all occasions when: 1) subjects purchased, received, or handled opiate drugs without using them; 2) subjects interacted with a regular opiate drug source without using; 3) subjects were in the same room/location as someone else who was using opiate drugs but did not use themselves; and 4) subjects were offered a sample of opiate drugs but refused.

Data Analysis

Drug use data were collected by the interviewer with the EMIT urinalysis data on hand so that discrepancies could be immediately noted and probes asked regarding the accuracy of the subject's answers. Self-reported quantity (estimated dollar value and episodes of use per week) of opiate use during pretreatment baseline (weeks -2 and -1), dosage stabilization (treatment weeks 2 and 3), and treatment endpoint (2 weeks prior to study termination for each subject) were compared using an ANOVA with repeated measures and post-hoc analyses so as to determine the impact of treatment on the frequency of opiate use.

A backward stepwise multiple regression and a forward stepwise multiple regression were used to evaluate the relationship between pretreatment demographic and drug use characteristics with respect to the number of opiate-free urines submitted by each individual during the course of treatment. The total number of opiate-free urines submitted by each subject was choosen as the dependent measure of treatment outcome because it uniquely combined time in treatment with an objective measure of opiate-abstinence. Two other outcome measures, 1) days in treatment and 2) percent of opiate-free urines (with missing urines counted as opiate positive), were also used as outcome measures in regression analysis to clarify their contribution to the composite outcome measure. Finally, percent of urines free of all illicit drugs (including opiates, cocaine, benzodiazepines, and other sedatives) was examined as a dependent measure to identify

[TABULAR DATA OMITTED]

predictors of total abstinence during methadone detoxification. Data entered into the regression as independent variables included all of the demographic and pretreatment characteristics listed in Table 1, including gender, race, marital status, legal status, age, years of education, income level, months of incarceration, number of previous drug treatments, lifetime duration of opiate use, and current duration of regular opiate use. Also included in the regression were three measures derived from interviews conducted during the 2-week baseline period as presented in Table 2: 1) the average number of opiate use episodes/week, 2) the average number of opiate acquisition sources/week, and 3) the average number of opiate-related interpersonal and situational exposures per week. Finally, nonopiate drug use at treatment entry (coded as present or absent in the intake urine) was also included in the regression analysis.

RESULTS

Impact of Treatment on Opiate Use

Figure 1 depicts the self-reported number of opiate use episode at three different study time points for the subset of patients (n = 51) who were in treatment a minimum of 5 weeks. The 2-week time periods included: 1) baseline--the 2-week period just prior to treatment entry; 2) stabilization--Weeks 2 and 3 of treatment, during which the methadone dose was stabilized at 40 mg; and 3) end point--the 2-week period just prior to treatment termination. The average number of opiate use episodes determined at baseline was 13.3/week, with a dramatic and sustained drop in episodes of opiate use/week to 2.22/week during the stabilization period and 3.37/week just prior to treatment termination. A one-factor ANOVA with repeated measures was used to examine this data, with a highly significant treatment effect (p < .0001; F = 106.344; df = 50,102), significant differences indicated between baseline and stabilization (p < .01; F = 82.911; df = 1,50), and significant differences indicated between baseline and endpoint (p < .01; F = 76.476; df = 1,50). Correspondingly, the percent of opiate-free urines collected during treatment weeks 2-12 ranged from 37 to 60% and was generally stable over time.

Predictors of Opiate Use

As noted, both a backward stepwise regression and a forward stepwise regression were used to derive predictors of treatment outcome with respect to the number of opiate-free urines submitted by each individual. Regardless of statistical approach, the same variables were identified. One subject's data were omitted from the analysis and substituted with the averaged data of four individuals with the most similar demographic profile after it was determined that the individual was an outlier. The subject was an opiate pill user (i.e., Tylox, Dilaudid) who took as many as 70 pills in individual doses during a single week, with each pill counted by the individual as a separate episode of use.

The overall regression was found to be highly significant (p < .001; F = 9.176; df = 4,66), with race (p < .008; t = -3.522; beta = -0.366), gender (p < .0245; t = 2.305; beta = 0.222), and the number of opiate use episodes/week at baseline (p < .0013, t = -3.364; beta = -0.338), identified as independently significant. Current duration of regular opiate use was also marginally significant (p < .1203; t = 1.574; beta = 0.1593).

The upper panel in Fig. 2 depicts the relationship between gender and opiate-free urines, while the lower panel illustrates the relationship between race and opiate-free urines. On average, males submitted approximately seven fewer opiate-free urines than did females, and non-White subjects submitted almost nine fewer opiate-free urines than did White subjects.

In Fig. 3 the predicted number of opiate-free urines were derived from the weighted beta values of each of the items cited above and plotted as a function of the number of actual opiate-free urines submitted by each individual. The x-axis depicts the actual number of opiate-free urines. As shown in Fig. 3, the dependent variable was well distributed along the abscissa, and approximately one-third of the variance is accounted for ([R.sup.2] = 0.357; adjusted [R.sup.2] = 0.318).

A second regression analysis with race, age, and gender excluded as independent variables was conducted in order to derive indices which were related to behavioral and environmental characteristics versus global physical/cultural identification. With race and gender removed, the overall regression was still highly significant (p < .007; F = 8.146; df = 2,68) although less than a fifth of the variance was accounted for ([R.sup.2] = 0.1933; adjusted [R.sup.2] = 1696). The number of opiate use episodes/week at baseline (p < .006; t = -2.836; beta = -0.3175) and the total number of drug-related stimulus cue exposure at baseline (p < .0362; t = -2.137; beta = -0.2393) were found to be independently significant variables in the analysis. As described above, the greater the amount of daily opiate use prior to treatment entry, the smaller the likelihood of submitting opiate-free urines during treatment. Further, the greater the number of daily contacts, with drug-related individuals, stimuli, and drug use situations prior to treatment entry, the smaller the likelihood of submitting opiate-free urines during treatment-regardless of race or gender.

Alternative Outcome Measures

As noted above, number of days in treatment and percent opiate-free urines were also examined separately as dependent measures to clarify their relative contribution to the composite outcome, number of drug-free urines. Only race/ethnicity was independently and significantly related to the number of days in treatment (p < .0457; t = 2.036; beta = 0.244). The overall regression was not significant (p < .0694; F = 2.775; df = 2,68) and accounted for less than 5% of the variance. Two variables were significant predictors of percent opiate-free urines: 1) race (p < .006, t = 3.595, beta = 0.377) and and 2) pretreatment episodes of opiate use per week (p < .0027, t = 3.118, beta = 0.324). Two other variables, the number of previous drug treatments and the duration of regular and uninterrupted drug use prior to treatment, were also marginally significant. The overall regression was highly significant (p < .001; F = 7.506, df = 4,66) and accounted for almost 30% of the variance ([R.sup.2] = 0.313; adjusted [R.sup.2] = 0.271). Thus, it appears that percent opiate-free urines rather than time in treatment was the major determinant of the regression effects with the composite measure.

When the number of urines testing negative for all nonprescribed substances (i.e., opiates, benzodiazepines, or cocaine) was used as the outcome measure, two variables emerged as significant predictors: 1) average number of different opiate acquisition sources (p < .0078; t - 2.742; beta = 0.305) and 2) years of education (p < .029; t = 2.231; beta = -0.249). The overall regression was significant (p < .0052; F = 4.646; df = 3,67), but little of the variance is accounted for ([R.sup.2] = 0.172; adjusted [R.sup.2] = 0.135). When percent of totally drug-free urines was used as outcome, the presence or absence of another drug in intake urine screening emerged as the strongest predictor (p < .0003; t = -3.820, beta = -0.381), with the duration of continuous and regular opiate use prior to treatment entry (p < .0119; t = 2.587; beta = 0.263) and race (p < .0495; t = -0.917 beta = .381) identified as significant and independent predictors. Gender and the weekly frequency of opiate consumption at the pretreatment baseline were also marginally predictive. The overall regression was significant (p < .001; F = 7.790; df = 5,65), and nearly a third of the variance is accounted for ([R.sup.2] = 0.375; adjusted ([R.sup.2] = 0.327).

DISCUSSION

In this study we focused on the relationship between pretreatment demographic characteristics and drug use behaviors with respect to treatment retention, self-reports of opiate use, and the submission of opiate-free urines during the course of a 90-day outpatient methadone detoxification. Our data indicated clear and identifiable population differences with respect to the frequency of opiate use during treatment. Overall, males submitted fewer opiate-free urines than did females, non-White subjects submitted fewer opiate-free urines than did White subjects, and subjects demonstrating a greater frequency of opiate use episodes/week at baseline also submitted fewer opiate-free urines. One other variable, the duration of continuous and regular opiate use prior to treatment entry, was also identified as marginally significant. Altogether, the four variables accounted for almost one-third of the variance in the data set. These results are generally consistent with previous work [1].

Factors contributing to racial and gender differences might include such things as general socioeconomic level, relative ease of drug availability, neighborhood prevalence of individuals associated with drugs (peer influences), availability of behavioral alternatives to drug use, quality of interactions with family members and significant others, and quality of interaction with clinical staff (e.g., related to racial homogeneity of patients and staff). While we were not able to evaluate every possibility, we were able to explore the issue of neighborhood social pathology. This factor was quantified by using Nurco's ratings of social pathology in Baltimore's various census tracts [11]. The index of social factors included in the ratings were narcotic arrests, nonnarcotic arrests, percent unmarried, percent non-White, aid to families with dependent children, general public assistance, food stamps, nondrug-related arrests, illegitimate births, homicides, total venereal disease, and percent of dwellings with average number of persons per room greater than or equal to one. There did not, however, appear to be any significant contribution when the ratings of neighborhood social pathology were entered into the regression models described above. It should also be noted that income level was included as a variable in all analyses, but no significant relationship was indicated.

The issue of race and gender with respect to treatment outcome is a complicated one in that the designations encompass a variety of genetic, behavioral, environmental, and social influences which fall beyond the scope of this study. The generality of the findings, however, need not be limited by considerations of causality. Awareness that specific characteristics may predict poorer treatment outcome within a given treatment modality allows for the introduction of enhanced/prohylactic measures for those at greater risk. This view is consistent with the notion of patient-treatment matching [12, 13] and with the optimal use of limited resources.

Our interest in behavioral and environmental influences on drug use behavior led to a second analysis of the data with race, age, and gender removed from the list of independent variables entered into the regression analysis. This analysis was conducted in order to derive indices which were related to behavioral and environmental characteristcs versus global physical/cultural identification. The average number of pretreatment opiate drug use episodes/weed (behavioral variable) and the frequency of pretreatment exposure to drug-related stimuli/events (environmental variable) emerged as two items which significantly and independently predicted the frequency of opiate use during the course of treatment. Removing the correlation overshadowing of race and gender allowed the behavioral/environmental factors to emerge from statistical obscurity and in doing so provide some support for behavioral models of substance us.

Efficacy of the 90-day methadone detoxification protocol was apparent in measures that reflected a dramatic reduction of opiate use compared with pretreatment levels of use. Of the 51 subjects who remained in treatment for longer than 4 weeks, the average dollar value of opiates consumed just prior to treatment entry averaged $345/week. A dramatic and sustained drop to an average of $23/week was noted during the stabilization period (Weeks 2 and 3). The dramatic decrease in the average dollar value of opiates consumed appears to have been maintained throughout treatment, as evidenced by the finding that the average value of opiates consumed during the 2-week period to treatment termination was only $35/week. This substantial during-treatment reduction of opiate use, presumably accompanied also by reductions in associated criminal activity, can be viewed as a benefit of detoxification treatment both to society and to the individual patients. The extent of improvement varied across individual patients. Opiate use patterns observed ranged from subjects who were able to abstain from opiate drug use for only a few days at a time to those who abstained from opiate drug use for up to 10 consecutive weeks perhaps an occasional lapse. The dramatic reduction of illicit drug use during detoxification treatment represents a legitimate but sometimes overlooked aspect of the efficacy of this type of treatment.

Anecdotal statements from the counseling staff and from the patients were universally positive with respect to the detoxification protocol. The patients often commented that they were pleased to find an alternative to the 21-day detoxification and that they did not wat maintenance because of the level of commitment it entailed as well as a frequently noted fear of addiction to methadone. The treatment staff were pleased by the quality of the therapeutic contact with the 90-day patients (as compared to 21-day methadone detoxification patients) in that a number of patients contined in some form of treatment (drug-free with the same counselor, Narcotics Anonymous, etc.) after leaving the detoxification program.

The data reported above rely heavily on the interview-based responses of the subjects regarding their behavior in the environment. We attempted to maximize the accuray/truthfulness of the self-reports by: 1) the absence of sanctions for drug use while in treatment, 2) using a semistructured interview with only a single interviewer, 3) TLC and EMIT urine tests X3/week to vefify drug use/nonuse and feddback to subjects regarding discrepancies, 4) interviewing on a weekly basis so as to minimize the need for recall and estimation, 5) paying a nominal sum as reimbursement for the time involved in order to encourage schedule compliance, and 6) working with the subjects to recreate activities during each day in the previous week as completely as possible in order to maximize accuracy. By requiring the subjects to answer numerous detailed questions about each drug use occasion, and by having the urine tests for the previous week available at the time of each interview, inconsistencies could be vigorously addressed.

Although average reported amounts and frequencies of opiate drug use during treatment was quite low, 37-65% of urine samples obtained from study patients after 1 week of treatment were opiate-positive, with frequent (3 times per week) urine testing. It should be noted, however, that urine test results do not provide a particularly reliable index of the quantitiy/frequency of recent drug use. Whether or not a particular episode of use is detected in urinalysis will depend on the amount used, the time elapsed between use and sample collection the drug metabolic and elimination profile of the particular subject, and the fluid balance of the subject prior to sample collection. Thus, a negative test does not preclude the possibility of recent use. On the other hand, a positive test indicates only that some recent use has occured but provides essentially no information about the recency or frequency of past use. Detection of two out of three opiate positive tests during a given week would be compatible with sporadic use (e.g., 1-2 times per week), particularly since it is possible for a single episode of use to result in more than one positive test during a frequent testing schedule.

In summary, this study has demonstrated some beneficial effects of detoxification treatment by showing dramatic decreases in rates and amounts of opiate drug use during treatment. The study has also identified predictors of treatment outcome in an opiate detoxification program. Race, gender, and the frequency of opiate use at the time of treatment entry all appeared to significantly influence treatment outcome. Identification of variables such as race and gender as significant determinants of treatment outcome may have implications for client to treatment matching and the optimal use of resources. As indicated above however, the variables of race and gender encompass a variety of genetic, behavioral, environmental, and social influences. The generality of our findings with regard to race and gender could be limited to the extent that the contribution of these underlying factors varies from region to region. The study also identified some behavioral and environmental influences on treatment outcome, specifically pretreatment frequency of opiate use and frequency of contact with drug-associated stimuli in the environment, that might profitably be included in future studies of treatment outcome predictors.

REFERENCES

[1] McLellan, A. T., Patient characteristics associated with outcome, in Research on the Treatment of Narcotic Addiction: State of the Art (J. R. Cooper el al., eds), National Institute on Drug Abuse Treatment Research Monograph Series, DHHS Publication No. (ADM) 87-1281, U.S. Government Printing Office, Washington, D.C., 1983, pp. 500-529.

[2] Babst, D., Chambers, C. D., and Warner, A., Characteristics predicting long-term retention in a methadone maintenance program, Br. J. Addict, 66:140-145 (1971).

[3] Szapocznic, J., and Ladner, R., Factors related to successful retention in methadone maintenance: A review, Inc. J. Addict. 12(8): 106-1985 (1977).

[4] Sechrest, D., and Crim, B., Methadone programs and crime reduction: A comparison of New York and California addicts, Int. J. Addict. 14:377-400 (1979).

[5] Stanton, M. D., The client as a family member, in Addicts and Aftercare (B. S. Brown, ed.), Sage, Beverly Hills, California, 1979, pp. 81-102.

[6] Krakowski, K., and Smart, R. G., Social and psychological characteristics of heroin addicts dropping out of treatment, Can. Psychiatr. Assoc. J. 19:41-47 (1974).

[7] McLeelan, A. T., Luborsky, L., O'Brien, C. P., et al., Predicitng response to alcohol and drug abuse treatments: Role of psychiatric symptoms, Arch. Gen. Psychiatry 40:620-628 (1983).

[8] Rounsaville, B. J., Kosten, T. R., Weissman, M. M., et al., Prognostic significance of psychiatric disorders in treated opiate addicts, Arch. Gen. Psychiatry 43:739-745 (1986).

[9] Barr, H. L., Problem drinking by drug addicts: Effects on treatment outcome, Issues Combined Alcohol Drug Abuse 1:413-440 (1978).

[10] Stitzer, M. L., McCaul, M. E., Bigelow, G. E., et al., Treatment outcome in methadone detoxification: Relationship to initial levels of illicit opiate use, Drug Alcohol Depend. 12:259-267 (1983).

[11] Nurco, D. N., Shaffer, J. W., and Cisin, I. H., An ecological analysis of the interrelationships among drug abuse and other indices of social pathology, Int. J. Addict. 19(4):441-145 (1984).

[12] McLellan, A. T., Druley, K. A., O'Brien, C. P., et al., Matching substance abuse patients to appropriate treatment: A conceptual and methodological approach, Drug Alcohol Depend. 3:189-195 (1980).

[13] McLellan, A. T., Luborsky, L., Woody, G. E., et al., Predicting response to alcohol and drug abuse treatments: Role of psychiatric severity, Arch. Gen. Psychiatry 40:620-625 (1983).

COPYRIGHT 1991 Taylor & Francis Ltd.
COPYRIGHT 2004 Gale Group