SELF-REGULATION OF DOSE IN METHADONE MAINTENANCE WITH CONTINGENT PRIVILEGES BARBARA HAVASSY* and WILLIAM A. HARGREAVES University
of California.
San Francisco
and LEE DE Marin
Treatment
BARROS
Center.
San Rafael. California
Abstract-The effect of giving methadone maintenance clients the opportunity to rcgulatc their dosage and of offering take-home doses as an incentive for dose reduction was studied. Subjects (116) were randomly assigned to one of three conditions: self-regulation of dose (SR-I ): sclf-regulation of dose with incentive for reduction (SR-2) and standard treatment (control): Dcpcndent variables were dosage and use of illicit drugs (measured by urinalyses). Results for the first four months show SR-I subjects increased dose and maintained themselves at dosages significantly greater than the SR-2 or control groups while having significantly less use of illicit opiates during certain time periods. SR-2 subjects were not diflerent from controls. Findings indicate subjects behaved responsibly under self-regulation and that this regimen is clinically feasible. Nevertheless. an increase in take-home privileges was an insufficient incentive to yield major reductions in dosage in SR-2 as a group.
.
Methadone maintenance. a treatment for heroin addicts that was initially highly acclaimed. has more recently come under much criticism. Thought by many to have the potential ability to attract and retain the majority of heroin addicts in rehabilitative treatment. it has not realized this potential. In a recent article, two pioneers of this form of treatment, Dole & Nyswander (1976), have argued that the limited impact of methadone maintenance is not attributable to the drug methadone, but to the delivery of the treatment. They assert that the critical problems stem from state and Federal regulations regarding treatment, such as restrictions on maximum dosage and frequent required clinic attendance. Improvement of treatment was argued to be unlikely under the conditions resulting from current regulations. This paper reports a clinical experiment that manipulated two of these regulated aspects of treatment. dosage and take-home privileges. These are sources of much concern and friction for clients and staff alike. The present study examined the effects of client regulation of dosage and also an increased number of take-home doses as an incentive for subjects to decrease their methadone dose.
SELF-REGULATION
OF
DOSE
Previous experience in allowing maintenance clients to regulate their daily dose of methadone has suggested that this can be a useful therapeutic technique. Although the procedure has been used infrequently, encouraging outcomes have been reported by Angle et al. (1973). and Goldstein et al. (1975). One interesting finding has been that the self-regulation procedure did not produce extreme dose increases, as some clinicians have feared. Goldstein er al. found that subjects increased their median dose only 10mg in 6 months (from 40 to 50 mg), however, they could not increase beyond 50 mg without losing their take-home doses. Of equal importance is a second finding that some patients used self-regulation to lower their dose and continued to remain at doses lower than study-start dose. A third observation was that among self-regulation clients. those who increased their dose also showed a decrease in the incidence of urines *Reprint requests should be sent to: Barbara of Psychiatry. San Francisco General Hospital.
Havassy. University of California, San Francisco. Department 1001 Potrero Avenue. San Francisco. California 941 IO. 31
32
BAWISAKAHAVASSY. WILLIAM A. HAKCXI~AVES and Lrirc DIGBARROS
containing illicit drugs. It appeared that clients preferred to increase their methadone dose rather than supplement it with illicit drugs. PERCEPTION
OF
CONTROL
Theoretical support for the efficacy of this kind of dosage regimen comes from psychological research on the perception of control. The literature indicates that when subjects feel a sense of control over impinging stimuli, effects are produced that are different from situations in which subjects feel their behavior has been forced by the circumstances. Glass & Singer (1972) Bowers (1968) Corah & Botfa (1970) Geer et al. (1970) and Staub et al. (1971) all have reported positive effects on task performance, tolerance to pain. and stress and anxiety indicators when subjects have control over aversive stimuli. Pain-producing stimuli are less painful and disruptive to subjects who can control and predict them. Predictable and/or controllable painful stimuli are preferred to equally painful stimuli that cannot be controlled or predicted. In some research involving social interactions (Langer. 1975; Langer & Rodin. 1976) subjects who had the illusion of control felt more confident and were more likely to take risks than subjects not induced to feel control. Similarly, aged residents of a nursing home receiving treatment emphasizing their responsibility for themselves and their freedom to make choices showed a significant improvement on ratings of alertness. active participation and sense of well-being over residents receiving treatment that stressed the staffs responsibility for them (Langer & Rodin, 1976). TAKE-HOME
DOSE
AS A BEHAVIORAL
INCENTIVE
Several reports have suggested that take-home doses can be powerful reinforcers. In a survey of client preferences. Yen (1974) found that the most preferred reward was additional take-home doses. Goldstein rr al. (1975) found that when the choice to increase dosage was coupled with the consequence of losing take-home privileges many clients preferred to retain take-home privileges by foregoing the increased dose. Bigelow et al. (1976) report a series of case studies suggesting that take-home doses could be used to shape abstinence from illicit drugs. Stitzer er al. (1977) also found that weekend take-homes delivered contingently for attendance at counseling sessions significantly increased attendance. These findings notwithstanding, we have found no report of a randomized clinical trial of a procedure for self-regulation of dose comparing its clinical effectiveness to “standard” treatment. Nor have we found reports of the effectiveness of take-home doses as an incentive for clients to reduce dose. In the clinical trial reported here, clients were randomly assigned to one of two self-regulation groups or to a control group (standard methadone treatment) in which all clinic procedures remained the same. One of the self-regulation groups was offered increased take-home privileges as an incentive for lowering their dose. The key dependent variables were dosage and illicit drug use (as reflected in urinalyses). It was predicted that subjects in both of the self-regulation groups would not raise their doses to levels significantly higher than those of the control group. It was also predicted that the self-regulation group with take-home incentives would achieve a lower dose over the experimental period than either the control or the other self-regulation group. With regard to illicit drug use. it was hypothesized that neither self-regulation group would have a higher rate of abuse than the control group. It was predicted that subjects in the self-regulation groups who substantially raised their dose would show a concomitant decrease in illicit drug use. Whether or not subjects who can determine their own dosage will in fact reduce it, the question of self-regulation is an important practical one. If subjects do not use self-regulation to ingest massive doses of methadone: if they do not create medical, behavioral, and/or administrative problems; and if there is no increased rate of drug abuse, it would indicate that some of the problematic restrictions of methadone treatment are not needed.
Self-regulation
of dose in methadone
maintcnancc
33
METHOD
The study was conducted at two methadone maintenance clinics in San Francisco. Clinic 1 was relatively large, with about 250 clients. while Clinic 2 had about 125 clients. Subjects consisted of maintenance clients who volunteered and who had already earned the privilege of at least one-take-home dose per week under Federal and State of California regulations. There were 116 persons from the two clinics who began the study, 72 males and 44 females. Of the 116, 85 were from clinic I and 3 1 were from Clinic 2. Clinic 1 subjects had a mean age of 35.3 yr; had been at that clinic an average of 27.7 months. and had a mean dose of 46.3 mg. Clinic 2 subjects had a mean age of 34.7 yr, had been at that clinic an average of 24.9 months and had a mean dose of 55.2 mg.
Medical charts were examined to provide a population of eligible subjects at each clinic. Flyers describing the experiment as one offering the possibility of greater control over dose and the chance to earn extra take-home doses were posted at the clinics prior to client contact. Eligible volunteers participated in an informed consent procedure. Subjects were stratified by sex and by whether they described themselves during the recruiting interview as undergoing gradual detoxification from methadone (tapering) and were randomly assigned to one of the three groups described below. Self-regulation of dose (SR-I). These subjects knew their dose level and decided how much methadone they were to receive, within the following limits: (1) only one dose change was allowed each week; (2) individual dose changes were limited to a range of 2-10mg for subjects at or above 30 mg; l-5 for those on a lower dose: (3) doses could be increased up to the Federal maximum of 120 mg/day; and (4) subjects above 100 mg could not receive take-home doses. The usual awarding of increased take-home privileges as specified by Federal and State of California regulations was not forfeited as long as the subject did not exceed the 100 mg level. Self-rryulation of dose with incrrlticr for duction (SR-2). Subjects in this condition had all the privileges and constraints of those in the previous group (SR-1). In addition, subjects who decreased their methadone dose below the study-start dose gained additional take-home privileges according to an incentive schedule. The greatest take-home privilege that could be earned was 2 weeks (13 days) of take-home doses. A subject who increased dosage after earning additional take-home privileges for a decrease forfeited the additional take-home doses. Each subject had an individually calculated dosage incentive schedule that was based on the subject’s starting maintenance dose and take-home privileges earned prior to the study. The same sequence of privileges was available to all subjects-what varied was the number of milligrams by which a subject needed to decrease his/her dose in order to earn the next privilege step. The schedule indicated the number of take-home doses that would be earned by decreasing dosage to levels less than study-start dose. The schedule was explained and given to each subject at entry to the study. Standard trratnwzt control. These subjects received no experimental treatment intervention. There was no clinic policy to prohibit clients knowing their dosage so that clients generally knew their dosage at all times. Clients could request a change of dosage that would be considered by clinic staff in relation to their perception of the client’s needs. The general orientation of clinic staff was that clients should be maintained on as low a dose as possible (without causing a relapse to regular heroin use). Therefore. dosage decreases were often more readily obtainable than increases. Dependent
cariables
The primary dependent variables were dosage and illicit drug use. Subjects’ dosage was obtained from clinic pharmacy records of actual dose dispensed. AR.
4 I-_(
daily
34
BAHI~AKAHAVASSY. WILLIAM A. HAKGKI:AVU and Lr.r. Dr! BARROS
Data on illicit drug use were obtained from results of urinalyses. At least one specimen a week was provided by subjects on randomly selected days. Specimens were provided at the clinic under observation by staff. A minimum of one specimen per subject per week was analyzed for the presence of opiates. Tests for the presence of amphetamines. barbiturates, and methadone were performed at least once a month per subject. Subjects earning extended take-home privileges, i.e. clinic attendance of once a week, once in 10 days or once every 2 weeks, had their urine tested for all substances (opiates. amphetamines, barbiturates. methadone) at each clinic visit. All urine specimens were sent for analysis to the laboratory used by maintenance clinics in San Francisco. the Chemistry Laboratory. Department of Public Health, San Francisco General Hospital. The laboratory is licensed as a methadone drug analysis laboratory by the State of California and had a rating of 90% or better for the duration of the study on monthly state-conducted external proficiency tests (required for retention of state license). This score indicates at least 900/, of the analyses were correct (i.e. not false positive or false negative). The laboratory screened the specimens using a homogeneous enzyme immunoassay system (the EMIT system). The EMIT system can detect a 0.5 pg/ml of morphine in urine with greater than 95:/, confidence. Specimens detected as positive by EMIT were confirmed by thin-layer chromatography.
RESULTS
These results are the findings of the first 16 weeks of the intervention at each of the two clinics. They consist of the analyses of dosage, illicit drug use and client retention. With regard to retention, the number of subjects leaving the study was small enough not to warrant examination by treatment group. At Clinic 1. 7 subjects or 8.27; of the Clinic sample left during the 16 weeks: 3 resigned from the study; 2 were removed for medical reasons; and 2 detoxified from methadone and terminated treatment. At Clinic 2. 5 subjects or 16”/, of the Clinic 2 sample left: 3 resigned and 2 terminated treatment. The 7 subjects in Clinic 1 who left during the first 16 weeks were omitted from the analyses reported in this paper, leaving a total II of 78. There were 26 subjects in SR-1. 27 in SR-2 and 25 in the control group. Of the 31 subjects in Clinic 2, 10 were in SR-1. 11 in SR-2 and 10 in the control group. Dosage
The success of the random assignment in equating initial dose was confirmed by an analysis of variance (ANOVA) conducted on dose at baseline by treatment assignment. There were no significant differences on baseline dose between treatment groups at either clinic, nor were there sex differences on dose at baseline. There were no interactions of treatment, sex, and taper status on baseline dose at either clinic. Dosage results for the first 16 weeks of the treatment period are shown in Fig. 1. The effect of treatment assignment on dose at Clinic 1 consisted of an initial rise in dose in the SR-1 group that was retained throughout the 16 week period. No major effects of treatment assignment of dose were observed at Clinic 2. Dose at baseline was highly correlated with dosage levels during the subsequent 16 weeks. Consequently, analyses of covariance (ANCOVA). using dose at baseline as the covariate, were conducted on dosage for weeks 1. 4. 8. 12 and 16. Separate ANCOVAs for treatment by sex and treatment by taper status were done since the U’S were not large enough for a 3-way analysis. Clir7ic 1. Treatment by sex results for the 16 weeks are reported in Table 1. Treatment assignment has a significant main effect on dose for weeks 1, 4. 8, and 12. At week 16, the treatment effect approaches significance. The effect is the same each time period: SR-1 subjects have increased their dose the most while SR-2 and the control group are at significantly lower and roughly equal dose levels, as seen in Fig. 1.
Self-regulation
75,_
CLINIC
of dose in methadone
r
I
35
maintcnancc
CLINIC
2 B-B_
Control SR- 1
.. .. .... . 33-2
Oh+i+d60?6 Week Fig. I. Mean
methadone
dose by treatment
groups for baseline and 16 weeks of study
The phenomena of the overall increase in dose of SR-I is attributable to the increases by SR-I subjects who began the study at doses of 4665 mg. (See Fig. 2). Except for these subjects. dose changes. irrespective of treatment assignment. are small. This interaction of dose at baseline with treatment assignment was detected by using dose at baseline as a four-category blocking variable (less than 30mg, 3&45 mg, 46-65 mg, and 66 mg or more), rather than as a covariate in ANOVAs on dose with treatment as the other main factor. The dose increase of the SR-I 46-65 mg subjects is reflected in treatment by baseline dose interactions at week 4 (F(6,77) = 2.56, P = 0.03); week 8 (F(6,75) = 2.83, P = 0.02); and week 12 (F(6.71) = 1.90. P = 0.10). At these times, SR-I 46-65 mg subjects had significantly larger doses than subjects at the same starting dose assigned to either SR-2 (P = 0.05 by Newman-Keuls) or the control group (P = 0.05 by Newman-Keuls). CIirzic ,7. There were no effects of treatment assignment on dose during the first 16 weeks of the study at Clinic 2. There were also no effects of sex or taper status. nor were there any interaction effects.
Data on illicit use of opiates, amphetamines. and barbiturates. during the intervention and also for a baseline period of 6 months prior to the start of the study, were examined. Results are for Clinic 1 only. Clinic 2. due to administrative problems, did not have sufficient data for analysis. Opiutc use. Illicit opiate use is expressed as the proportion of urine specimens collected that contained opiates. Subjects refusing to leave specimens were counted as positive Table
I. Analyses
of covariance
of dose with dose at baseline
I Source of variance Treatment (A) Sex (B) AxB Error d.f.
d.f. 2
1 2
.‘The f’s for the covariate in the tdbte. * P $ 0. IO. ** P I 0.05.
F 3.57** 0.62 0.85 71
Week 8
4 F 3.1x** 0.16 0.26 70
are all significant
F 4.63** 0.32 0.48 69
as the covariate”
I2
F 2.99* 0.67 1.55 65
16
F 2.27 2.29 1.94 63
at P 5 0.001. and are not reported
BAWISAKAHAVASSY. WILLIAM A. HAHGKEAVESand Lw Dr: BAKROS
36
Starting
-
E o
-u
65-1’
0
\
L----J Control ---SR_, . . . . . . . . . SR_2
\
I
\
Starting
50
Fig. 2. Mean
methadone
2 4 6 Weeks
Dose: >65mg
Dose: 46-65mg
8 IO 12 14 16 in Study
dose over 16 weeks for subjects
at different
study-start
doses
for opiates. During the baseline period there were no major differences related to treatment assignment. sex, taper status, or baseline methadone dose on illicit opiate use. For the 16 weeks of the intervention. however. SR-I tends to have the smallest proportion of opiate positive urine specimens (see Fig. 3). Results of ANOVAs on opiate use data with treatment and baseline opiate use as main factors are displayed in Table 2. The dependent variable is the proportion of total specimens positive for opiates cumulated over four successive 4-week time periods. There are significant main effects of treatment on opiate use during weeks 9-12. but none for other time periods. Post hoc analyses of the week 9-12 treatment effect indicate that SR-1 has significantly less opiate use (mean proportion of positive specimens = 0.10)
Control w-w. SR-1 . . . . . . . . . . 33-2
[,i,
, Baseline
l-4
,
,
,
5-8
9-12
13-16
Weeks Fig. 3. Means
by treatment
group
of proportion
of urines
positive
for opiates.
Self-regulation
Table
of dose in methadone
2. Analyses of variance
on proportion
of
37
maintenance
urines positive for opiates Weeks
1-4 Source of variance Treatment (A) Baseline opiate use” (B)
AxB Error d.f. “Subjects were divided * P 5 0.10. ** P I 0.05. *** P 2 0.01.
d.f. 2
I 2
5-8
F
F
0.55 8.87***
0.88 9.68***
I .02 72
2.74* 72
9-12
13-16
3.:2** 5.49**
1.04 6.60**
F
I.17 71
0.36 67
into two groups at the median.
than either the control group (x = 0.32) or SR-2 (x = 0.27) (P = 0.05, Newman-Keuls). Baseline opiate use exhibited a major relationship to later opiate use as seen in Table 2. Subjects below the median in baseline opiate use have significantly less opiate use than subjects above the median at baseline (0.15 vs 0.32). Taper status, sex, and methadone dose at baseline were not observed to affect opiate use during the 16 week period. Barbiturate and amphetamine use. These data were examined by chi-square tests of association (number of positive specimens by treatment group) as they were not suitable for ANOVA. Each drug was examined separately for each clinic for the 4 week time periods and also summed over the 16 weeks. There were no significant x2 tests at Clinic 1 and insufficient data at Clinic 2. When the data for barbiturates and amphetamines for the 16 weeks at Clinic 1 were combined (an index of non-opiate substance abuse) the xZ test was not significant. DISCUSSION
The dosage results at Clinic 1 for the first 16 weeks do not support our hypotheses or earlier reports of self-regulation regimens. It is important to note, however, that other self-regulation regimens in which instances of dose lowering were reported did not employ a control group against which to compare self-regulation subjects. What these reports indicated is that subjects who were allowed to regulate their dosage did not markedly increase dose and some subjects lowered their dose. Our experience with self-regulation, both with and without an incentive for dose lowering, has not been different: subjects did not greatly increase dose and some decreased their dose substantially. In addition, what we found was that take-home doses are apparently a weaker incentive to reduce dose than hypothesized. Either subjects wish to remain on methadone, or the perceived consequences associated with lower doses are thought to be more aversive than regular attendance requirements. One factor not to be ignored is that admissions into the methadone program became severely limited during the period of the study and clients might have felt if they detoxified they might not be able to reenter the program. Nevertheless, the incentive of take-home doses in SR-2 does appear to have counteracted the tendency toward dose increase seen in the SR-1 group. The dose increase observed among the SR-1 subjects starting the study in the 4&65 mg range is clinically understandable. These subjects were at a dose range in which increases would be physically noticeable and also possible without exceeding maximum dosage allowed. They had little apparent reason to decrease dosage since they tended not to be tapering, nor did they have any experimental incentive to lower their dose. These subjects were different in this particular combination of factors from all other SR-1 subjects. SR-I subjects at higher starting doses did not have any real ability to increase their dose noticeably and SR-1 subjects at lower starting doses tended to be tapering. In contrast to the dosage results. the findings regarding opiate use cpnfirm the hypothesis that neither self-regulation group would show a higher abuse rate than the control group. In fact. during weeks 9-12, SR-1 subjects had significantly less opiate use than
38
BAHIIAKA HAVASSI.. WILLIAM A. HAKGHI~AVI.S and LI.I; DI: BAKKOS
the other two treatment groups. This particular outcome was not predicted and we have no ready explanation of it. We expect that examination of the opiate use data from the latter part of the experiment (weeks 1748) will indicate whether this outcome was a chance occurrence or part of a time-related pattern of differences among the treatment groups. The other prediction about drug abuse, that self-regulation subjects who raised their dose would show concomitant decrease in illicit drug use, was not supported. In summary, the results of the first 16 weeks do not support hypotheses regarding the dosage behavior of subjects on the two self-regulation regimens. Nevertheless. both of these regimens proved to be clinically feasible. Though the self-regulation subjects did exercise their option to change dosage. they behaved responsibly. an experience consistent with the reports of Goldstein et cl/. (1975). Self-regulation subjects also did not have a higher opiate abuse rate than “standard” treatment subjects. These results do not provide any support for the view that dosage and take-home privileges need to be as highly regulated and restricted as at present. Although a self-regulation regimen may not lead clients to decrease their dosage. it is possible to utilize self-regulation to dispense methadone doses and take-home doses in a systematic, safe, and organized way. d~lir~o~~lct/~c,,~lc,rrs--The rcscarch rcportcd in this article was supported by NIDA Grant No. DA4RGOl 1. The authors wish to thank the clinical staff of the San Francisco Methadone Maintenance Treatment Program and Drs Jcrrold Wittcrt and Wcslcy Johnson for their assistance. REFERENCES Angle. H. V.. Strclingcr. R. W.. & Parawatikar. S. Self-prescription of methadone: Inpatient and outpatient addict populations. P,occc&qs qf rllti F#/I Ntrriomrl Coc$wv~c~~ m Mct/~t~/ow Tr~~arr~~~r~t. 1973. I. 681 689. Bigclow. G.. Lawrcncc. C.. Stit7cr. M.. & Wells. D. Behavioral treatments during outpatient methadone maintcnunce: A controlled evaluation. Paper presented at American Psychological Association meeting. Washington. D.C.. September. 1976. Bowers. K. Pain. anxiety. and pcrccivcd control. Joffrr~trl o/’ Corlsdrirlg cud Clirtictrl Ps~d~o/o~~~~. 1968. 32. 596-602. Corah. N.. & Boffa. J. Perceived control. self-observation, and response to aversive stimulation. Journal $ Pcrsonulitr ontl Social Psycholo~~y. 1970. 16( 1). l-4. Dole V. P.. & Nyswander. M. E. Methadone maintenance treatment: A ten year perspective. Journul of the Auwrictrn Met/i& Associtrriorl. 1976. 235 (19). 2117-21 19. Gccr. J.. Davison. G.. & Gatchcl. R. Reduction of stress in humans through nonveridical perceived control of aversive stimulation. Jo~rr~fal y/’ Pcrsorur/ir~ trrd Socitrl Ps~ddog~. 1970. 16. 73 I-738. Glass. D.. & Singer. J. L’rhan Srrc,ss. New York: Academic Press. 1972. Goldstein. A.. Hansteen. R. W.. & Horns. W. H. Control of methadone dosage by patients. In E. Senay. V. Shorty & H. Alksne (Edsl. D~,rc,/op~nc*nt.sirl r/w Fir/t/ of‘ Dmg A/me. pp. 373.-381. Massachusetts: Schenkman. 1975. Langer. E. J. The illusion of control Jourr~trl of Posorltrlir!, tr~ltl So&r/ Ps~d~ology. 1975. 32. 31 I --328. Langer. E. J. and Rodin J. The effects of choice and enhanced personal responsibility for the aged: A field experiment in an institutional setting. Jormwl qf‘ Pcrxmrliry ctrd Socitrl Ps,~d~o/og,~~. 1976. 34(Z). I9 I-I 98. Staub. E.. Tursky. B.. & Schwartz. G. E. Self-control and predictability: Their effects on reactions to aversive stimulation. Jorrrrztrl of Pmwri~~lir~~ trml So&r/ Psycho/oy~. 197 I. 18. 157-l 62. Stitzer. M.. Bigelow. G.. Lawrence. C.. Cohen. V.. D’Lugoff. B.. & Hawthorne. J. Medication take-home ;IS a reinforcer in a methadone maintenance program. Atltlicrilr Bhwiors. 1977. 2(l). 9-14. Yen. S. Availability of activity reinforcers in a drug abuse clinic. P.s~ddoyiccr/ Rcporrs. 1974. 34(3). 1021-1022.
of California.
San Francisco
and LEE DE Marin
Treatment
BARROS
Center.
San Rafael. California
Abstract-The effect of giving methadone maintenance clients the opportunity to rcgulatc their dosage and of offering take-home doses as an incentive for dose reduction was studied. Subjects (116) were randomly assigned to one of three conditions: self-regulation of dose (SR-I ): sclf-regulation of dose with incentive for reduction (SR-2) and standard treatment (control): Dcpcndent variables were dosage and use of illicit drugs (measured by urinalyses). Results for the first four months show SR-I subjects increased dose and maintained themselves at dosages significantly greater than the SR-2 or control groups while having significantly less use of illicit opiates during certain time periods. SR-2 subjects were not diflerent from controls. Findings indicate subjects behaved responsibly under self-regulation and that this regimen is clinically feasible. Nevertheless. an increase in take-home privileges was an insufficient incentive to yield major reductions in dosage in SR-2 as a group.
.
Methadone maintenance. a treatment for heroin addicts that was initially highly acclaimed. has more recently come under much criticism. Thought by many to have the potential ability to attract and retain the majority of heroin addicts in rehabilitative treatment. it has not realized this potential. In a recent article, two pioneers of this form of treatment, Dole & Nyswander (1976), have argued that the limited impact of methadone maintenance is not attributable to the drug methadone, but to the delivery of the treatment. They assert that the critical problems stem from state and Federal regulations regarding treatment, such as restrictions on maximum dosage and frequent required clinic attendance. Improvement of treatment was argued to be unlikely under the conditions resulting from current regulations. This paper reports a clinical experiment that manipulated two of these regulated aspects of treatment. dosage and take-home privileges. These are sources of much concern and friction for clients and staff alike. The present study examined the effects of client regulation of dosage and also an increased number of take-home doses as an incentive for subjects to decrease their methadone dose.
SELF-REGULATION
OF
DOSE
Previous experience in allowing maintenance clients to regulate their daily dose of methadone has suggested that this can be a useful therapeutic technique. Although the procedure has been used infrequently, encouraging outcomes have been reported by Angle et al. (1973). and Goldstein et al. (1975). One interesting finding has been that the self-regulation procedure did not produce extreme dose increases, as some clinicians have feared. Goldstein er al. found that subjects increased their median dose only 10mg in 6 months (from 40 to 50 mg), however, they could not increase beyond 50 mg without losing their take-home doses. Of equal importance is a second finding that some patients used self-regulation to lower their dose and continued to remain at doses lower than study-start dose. A third observation was that among self-regulation clients. those who increased their dose also showed a decrease in the incidence of urines *Reprint requests should be sent to: Barbara of Psychiatry. San Francisco General Hospital.
Havassy. University of California, San Francisco. Department 1001 Potrero Avenue. San Francisco. California 941 IO. 31
32
BAWISAKAHAVASSY. WILLIAM A. HAKCXI~AVES and Lrirc DIGBARROS
containing illicit drugs. It appeared that clients preferred to increase their methadone dose rather than supplement it with illicit drugs. PERCEPTION
OF
CONTROL
Theoretical support for the efficacy of this kind of dosage regimen comes from psychological research on the perception of control. The literature indicates that when subjects feel a sense of control over impinging stimuli, effects are produced that are different from situations in which subjects feel their behavior has been forced by the circumstances. Glass & Singer (1972) Bowers (1968) Corah & Botfa (1970) Geer et al. (1970) and Staub et al. (1971) all have reported positive effects on task performance, tolerance to pain. and stress and anxiety indicators when subjects have control over aversive stimuli. Pain-producing stimuli are less painful and disruptive to subjects who can control and predict them. Predictable and/or controllable painful stimuli are preferred to equally painful stimuli that cannot be controlled or predicted. In some research involving social interactions (Langer. 1975; Langer & Rodin. 1976) subjects who had the illusion of control felt more confident and were more likely to take risks than subjects not induced to feel control. Similarly, aged residents of a nursing home receiving treatment emphasizing their responsibility for themselves and their freedom to make choices showed a significant improvement on ratings of alertness. active participation and sense of well-being over residents receiving treatment that stressed the staffs responsibility for them (Langer & Rodin, 1976). TAKE-HOME
DOSE
AS A BEHAVIORAL
INCENTIVE
Several reports have suggested that take-home doses can be powerful reinforcers. In a survey of client preferences. Yen (1974) found that the most preferred reward was additional take-home doses. Goldstein rr al. (1975) found that when the choice to increase dosage was coupled with the consequence of losing take-home privileges many clients preferred to retain take-home privileges by foregoing the increased dose. Bigelow et al. (1976) report a series of case studies suggesting that take-home doses could be used to shape abstinence from illicit drugs. Stitzer er al. (1977) also found that weekend take-homes delivered contingently for attendance at counseling sessions significantly increased attendance. These findings notwithstanding, we have found no report of a randomized clinical trial of a procedure for self-regulation of dose comparing its clinical effectiveness to “standard” treatment. Nor have we found reports of the effectiveness of take-home doses as an incentive for clients to reduce dose. In the clinical trial reported here, clients were randomly assigned to one of two self-regulation groups or to a control group (standard methadone treatment) in which all clinic procedures remained the same. One of the self-regulation groups was offered increased take-home privileges as an incentive for lowering their dose. The key dependent variables were dosage and illicit drug use (as reflected in urinalyses). It was predicted that subjects in both of the self-regulation groups would not raise their doses to levels significantly higher than those of the control group. It was also predicted that the self-regulation group with take-home incentives would achieve a lower dose over the experimental period than either the control or the other self-regulation group. With regard to illicit drug use. it was hypothesized that neither self-regulation group would have a higher rate of abuse than the control group. It was predicted that subjects in the self-regulation groups who substantially raised their dose would show a concomitant decrease in illicit drug use. Whether or not subjects who can determine their own dosage will in fact reduce it, the question of self-regulation is an important practical one. If subjects do not use self-regulation to ingest massive doses of methadone: if they do not create medical, behavioral, and/or administrative problems; and if there is no increased rate of drug abuse, it would indicate that some of the problematic restrictions of methadone treatment are not needed.
Self-regulation
of dose in methadone
maintcnancc
33
METHOD
The study was conducted at two methadone maintenance clinics in San Francisco. Clinic 1 was relatively large, with about 250 clients. while Clinic 2 had about 125 clients. Subjects consisted of maintenance clients who volunteered and who had already earned the privilege of at least one-take-home dose per week under Federal and State of California regulations. There were 116 persons from the two clinics who began the study, 72 males and 44 females. Of the 116, 85 were from clinic I and 3 1 were from Clinic 2. Clinic 1 subjects had a mean age of 35.3 yr; had been at that clinic an average of 27.7 months. and had a mean dose of 46.3 mg. Clinic 2 subjects had a mean age of 34.7 yr, had been at that clinic an average of 24.9 months and had a mean dose of 55.2 mg.
Medical charts were examined to provide a population of eligible subjects at each clinic. Flyers describing the experiment as one offering the possibility of greater control over dose and the chance to earn extra take-home doses were posted at the clinics prior to client contact. Eligible volunteers participated in an informed consent procedure. Subjects were stratified by sex and by whether they described themselves during the recruiting interview as undergoing gradual detoxification from methadone (tapering) and were randomly assigned to one of the three groups described below. Self-regulation of dose (SR-I). These subjects knew their dose level and decided how much methadone they were to receive, within the following limits: (1) only one dose change was allowed each week; (2) individual dose changes were limited to a range of 2-10mg for subjects at or above 30 mg; l-5 for those on a lower dose: (3) doses could be increased up to the Federal maximum of 120 mg/day; and (4) subjects above 100 mg could not receive take-home doses. The usual awarding of increased take-home privileges as specified by Federal and State of California regulations was not forfeited as long as the subject did not exceed the 100 mg level. Self-rryulation of dose with incrrlticr for duction (SR-2). Subjects in this condition had all the privileges and constraints of those in the previous group (SR-1). In addition, subjects who decreased their methadone dose below the study-start dose gained additional take-home privileges according to an incentive schedule. The greatest take-home privilege that could be earned was 2 weeks (13 days) of take-home doses. A subject who increased dosage after earning additional take-home privileges for a decrease forfeited the additional take-home doses. Each subject had an individually calculated dosage incentive schedule that was based on the subject’s starting maintenance dose and take-home privileges earned prior to the study. The same sequence of privileges was available to all subjects-what varied was the number of milligrams by which a subject needed to decrease his/her dose in order to earn the next privilege step. The schedule indicated the number of take-home doses that would be earned by decreasing dosage to levels less than study-start dose. The schedule was explained and given to each subject at entry to the study. Standard trratnwzt control. These subjects received no experimental treatment intervention. There was no clinic policy to prohibit clients knowing their dosage so that clients generally knew their dosage at all times. Clients could request a change of dosage that would be considered by clinic staff in relation to their perception of the client’s needs. The general orientation of clinic staff was that clients should be maintained on as low a dose as possible (without causing a relapse to regular heroin use). Therefore. dosage decreases were often more readily obtainable than increases. Dependent
cariables
The primary dependent variables were dosage and illicit drug use. Subjects’ dosage was obtained from clinic pharmacy records of actual dose dispensed. AR.
4 I-_(
daily
34
BAHI~AKAHAVASSY. WILLIAM A. HAKGKI:AVU and Lr.r. Dr! BARROS
Data on illicit drug use were obtained from results of urinalyses. At least one specimen a week was provided by subjects on randomly selected days. Specimens were provided at the clinic under observation by staff. A minimum of one specimen per subject per week was analyzed for the presence of opiates. Tests for the presence of amphetamines. barbiturates, and methadone were performed at least once a month per subject. Subjects earning extended take-home privileges, i.e. clinic attendance of once a week, once in 10 days or once every 2 weeks, had their urine tested for all substances (opiates. amphetamines, barbiturates. methadone) at each clinic visit. All urine specimens were sent for analysis to the laboratory used by maintenance clinics in San Francisco. the Chemistry Laboratory. Department of Public Health, San Francisco General Hospital. The laboratory is licensed as a methadone drug analysis laboratory by the State of California and had a rating of 90% or better for the duration of the study on monthly state-conducted external proficiency tests (required for retention of state license). This score indicates at least 900/, of the analyses were correct (i.e. not false positive or false negative). The laboratory screened the specimens using a homogeneous enzyme immunoassay system (the EMIT system). The EMIT system can detect a 0.5 pg/ml of morphine in urine with greater than 95:/, confidence. Specimens detected as positive by EMIT were confirmed by thin-layer chromatography.
RESULTS
These results are the findings of the first 16 weeks of the intervention at each of the two clinics. They consist of the analyses of dosage, illicit drug use and client retention. With regard to retention, the number of subjects leaving the study was small enough not to warrant examination by treatment group. At Clinic 1. 7 subjects or 8.27; of the Clinic sample left during the 16 weeks: 3 resigned from the study; 2 were removed for medical reasons; and 2 detoxified from methadone and terminated treatment. At Clinic 2. 5 subjects or 16”/, of the Clinic 2 sample left: 3 resigned and 2 terminated treatment. The 7 subjects in Clinic 1 who left during the first 16 weeks were omitted from the analyses reported in this paper, leaving a total II of 78. There were 26 subjects in SR-1. 27 in SR-2 and 25 in the control group. Of the 31 subjects in Clinic 2, 10 were in SR-1. 11 in SR-2 and 10 in the control group. Dosage
The success of the random assignment in equating initial dose was confirmed by an analysis of variance (ANOVA) conducted on dose at baseline by treatment assignment. There were no significant differences on baseline dose between treatment groups at either clinic, nor were there sex differences on dose at baseline. There were no interactions of treatment, sex, and taper status on baseline dose at either clinic. Dosage results for the first 16 weeks of the treatment period are shown in Fig. 1. The effect of treatment assignment on dose at Clinic 1 consisted of an initial rise in dose in the SR-1 group that was retained throughout the 16 week period. No major effects of treatment assignment of dose were observed at Clinic 2. Dose at baseline was highly correlated with dosage levels during the subsequent 16 weeks. Consequently, analyses of covariance (ANCOVA). using dose at baseline as the covariate, were conducted on dosage for weeks 1. 4. 8. 12 and 16. Separate ANCOVAs for treatment by sex and treatment by taper status were done since the U’S were not large enough for a 3-way analysis. Clir7ic 1. Treatment by sex results for the 16 weeks are reported in Table 1. Treatment assignment has a significant main effect on dose for weeks 1, 4. 8, and 12. At week 16, the treatment effect approaches significance. The effect is the same each time period: SR-1 subjects have increased their dose the most while SR-2 and the control group are at significantly lower and roughly equal dose levels, as seen in Fig. 1.
Self-regulation
75,_
CLINIC
of dose in methadone
r
I
35
maintcnancc
CLINIC
2 B-B_
Control SR- 1
.. .. .... . 33-2
Oh+i+d60?6 Week Fig. I. Mean
methadone
dose by treatment
groups for baseline and 16 weeks of study
The phenomena of the overall increase in dose of SR-I is attributable to the increases by SR-I subjects who began the study at doses of 4665 mg. (See Fig. 2). Except for these subjects. dose changes. irrespective of treatment assignment. are small. This interaction of dose at baseline with treatment assignment was detected by using dose at baseline as a four-category blocking variable (less than 30mg, 3&45 mg, 46-65 mg, and 66 mg or more), rather than as a covariate in ANOVAs on dose with treatment as the other main factor. The dose increase of the SR-I 46-65 mg subjects is reflected in treatment by baseline dose interactions at week 4 (F(6,77) = 2.56, P = 0.03); week 8 (F(6,75) = 2.83, P = 0.02); and week 12 (F(6.71) = 1.90. P = 0.10). At these times, SR-I 46-65 mg subjects had significantly larger doses than subjects at the same starting dose assigned to either SR-2 (P = 0.05 by Newman-Keuls) or the control group (P = 0.05 by Newman-Keuls). CIirzic ,7. There were no effects of treatment assignment on dose during the first 16 weeks of the study at Clinic 2. There were also no effects of sex or taper status. nor were there any interaction effects.
Data on illicit use of opiates, amphetamines. and barbiturates. during the intervention and also for a baseline period of 6 months prior to the start of the study, were examined. Results are for Clinic 1 only. Clinic 2. due to administrative problems, did not have sufficient data for analysis. Opiutc use. Illicit opiate use is expressed as the proportion of urine specimens collected that contained opiates. Subjects refusing to leave specimens were counted as positive Table
I. Analyses
of covariance
of dose with dose at baseline
I Source of variance Treatment (A) Sex (B) AxB Error d.f.
d.f. 2
1 2
.‘The f’s for the covariate in the tdbte. * P $ 0. IO. ** P I 0.05.
F 3.57** 0.62 0.85 71
Week 8
4 F 3.1x** 0.16 0.26 70
are all significant
F 4.63** 0.32 0.48 69
as the covariate”
I2
F 2.99* 0.67 1.55 65
16
F 2.27 2.29 1.94 63
at P 5 0.001. and are not reported
BAWISAKAHAVASSY. WILLIAM A. HAHGKEAVESand Lw Dr: BAKROS
36
Starting
-
E o
-u
65-1’
0
\
L----J Control ---SR_, . . . . . . . . . SR_2
\
I
\
Starting
50
Fig. 2. Mean
methadone
2 4 6 Weeks
Dose: >65mg
Dose: 46-65mg
8 IO 12 14 16 in Study
dose over 16 weeks for subjects
at different
study-start
doses
for opiates. During the baseline period there were no major differences related to treatment assignment. sex, taper status, or baseline methadone dose on illicit opiate use. For the 16 weeks of the intervention. however. SR-I tends to have the smallest proportion of opiate positive urine specimens (see Fig. 3). Results of ANOVAs on opiate use data with treatment and baseline opiate use as main factors are displayed in Table 2. The dependent variable is the proportion of total specimens positive for opiates cumulated over four successive 4-week time periods. There are significant main effects of treatment on opiate use during weeks 9-12. but none for other time periods. Post hoc analyses of the week 9-12 treatment effect indicate that SR-1 has significantly less opiate use (mean proportion of positive specimens = 0.10)
Control w-w. SR-1 . . . . . . . . . . 33-2
[,i,
, Baseline
l-4
,
,
,
5-8
9-12
13-16
Weeks Fig. 3. Means
by treatment
group
of proportion
of urines
positive
for opiates.
Self-regulation
Table
of dose in methadone
2. Analyses of variance
on proportion
of
37
maintenance
urines positive for opiates Weeks
1-4 Source of variance Treatment (A) Baseline opiate use” (B)
AxB Error d.f. “Subjects were divided * P 5 0.10. ** P I 0.05. *** P 2 0.01.
d.f. 2
I 2
5-8
F
F
0.55 8.87***
0.88 9.68***
I .02 72
2.74* 72
9-12
13-16
3.:2** 5.49**
1.04 6.60**
F
I.17 71
0.36 67
into two groups at the median.
than either the control group (x = 0.32) or SR-2 (x = 0.27) (P = 0.05, Newman-Keuls). Baseline opiate use exhibited a major relationship to later opiate use as seen in Table 2. Subjects below the median in baseline opiate use have significantly less opiate use than subjects above the median at baseline (0.15 vs 0.32). Taper status, sex, and methadone dose at baseline were not observed to affect opiate use during the 16 week period. Barbiturate and amphetamine use. These data were examined by chi-square tests of association (number of positive specimens by treatment group) as they were not suitable for ANOVA. Each drug was examined separately for each clinic for the 4 week time periods and also summed over the 16 weeks. There were no significant x2 tests at Clinic 1 and insufficient data at Clinic 2. When the data for barbiturates and amphetamines for the 16 weeks at Clinic 1 were combined (an index of non-opiate substance abuse) the xZ test was not significant. DISCUSSION
The dosage results at Clinic 1 for the first 16 weeks do not support our hypotheses or earlier reports of self-regulation regimens. It is important to note, however, that other self-regulation regimens in which instances of dose lowering were reported did not employ a control group against which to compare self-regulation subjects. What these reports indicated is that subjects who were allowed to regulate their dosage did not markedly increase dose and some subjects lowered their dose. Our experience with self-regulation, both with and without an incentive for dose lowering, has not been different: subjects did not greatly increase dose and some decreased their dose substantially. In addition, what we found was that take-home doses are apparently a weaker incentive to reduce dose than hypothesized. Either subjects wish to remain on methadone, or the perceived consequences associated with lower doses are thought to be more aversive than regular attendance requirements. One factor not to be ignored is that admissions into the methadone program became severely limited during the period of the study and clients might have felt if they detoxified they might not be able to reenter the program. Nevertheless, the incentive of take-home doses in SR-2 does appear to have counteracted the tendency toward dose increase seen in the SR-1 group. The dose increase observed among the SR-1 subjects starting the study in the 4&65 mg range is clinically understandable. These subjects were at a dose range in which increases would be physically noticeable and also possible without exceeding maximum dosage allowed. They had little apparent reason to decrease dosage since they tended not to be tapering, nor did they have any experimental incentive to lower their dose. These subjects were different in this particular combination of factors from all other SR-1 subjects. SR-I subjects at higher starting doses did not have any real ability to increase their dose noticeably and SR-1 subjects at lower starting doses tended to be tapering. In contrast to the dosage results. the findings regarding opiate use cpnfirm the hypothesis that neither self-regulation group would show a higher abuse rate than the control group. In fact. during weeks 9-12, SR-1 subjects had significantly less opiate use than
38
BAHIIAKA HAVASSI.. WILLIAM A. HAKGHI~AVI.S and LI.I; DI: BAKKOS
the other two treatment groups. This particular outcome was not predicted and we have no ready explanation of it. We expect that examination of the opiate use data from the latter part of the experiment (weeks 1748) will indicate whether this outcome was a chance occurrence or part of a time-related pattern of differences among the treatment groups. The other prediction about drug abuse, that self-regulation subjects who raised their dose would show concomitant decrease in illicit drug use, was not supported. In summary, the results of the first 16 weeks do not support hypotheses regarding the dosage behavior of subjects on the two self-regulation regimens. Nevertheless. both of these regimens proved to be clinically feasible. Though the self-regulation subjects did exercise their option to change dosage. they behaved responsibly. an experience consistent with the reports of Goldstein et cl/. (1975). Self-regulation subjects also did not have a higher opiate abuse rate than “standard” treatment subjects. These results do not provide any support for the view that dosage and take-home privileges need to be as highly regulated and restricted as at present. Although a self-regulation regimen may not lead clients to decrease their dosage. it is possible to utilize self-regulation to dispense methadone doses and take-home doses in a systematic, safe, and organized way. d~lir~o~~lct/~c,,~lc,rrs--The rcscarch rcportcd in this article was supported by NIDA Grant No. DA4RGOl 1. The authors wish to thank the clinical staff of the San Francisco Methadone Maintenance Treatment Program and Drs Jcrrold Wittcrt and Wcslcy Johnson for their assistance. REFERENCES Angle. H. V.. Strclingcr. R. W.. & Parawatikar. S. Self-prescription of methadone: Inpatient and outpatient addict populations. P,occc&qs qf rllti F#/I Ntrriomrl Coc$wv~c~~ m Mct/~t~/ow Tr~~arr~~~r~t. 1973. I. 681 689. Bigclow. G.. Lawrcncc. C.. Stit7cr. M.. & Wells. D. Behavioral treatments during outpatient methadone maintcnunce: A controlled evaluation. Paper presented at American Psychological Association meeting. Washington. D.C.. September. 1976. Bowers. K. Pain. anxiety. and pcrccivcd control. Joffrr~trl o/’ Corlsdrirlg cud Clirtictrl Ps~d~o/o~~~~. 1968. 32. 596-602. Corah. N.. & Boffa. J. Perceived control. self-observation, and response to aversive stimulation. Journal $ Pcrsonulitr ontl Social Psycholo~~y. 1970. 16( 1). l-4. Dole V. P.. & Nyswander. M. E. Methadone maintenance treatment: A ten year perspective. Journul of the Auwrictrn Met/i& Associtrriorl. 1976. 235 (19). 2117-21 19. Gccr. J.. Davison. G.. & Gatchcl. R. Reduction of stress in humans through nonveridical perceived control of aversive stimulation. Jo~rr~fal y/’ Pcrsorur/ir~ trrd Socitrl Ps~ddog~. 1970. 16. 73 I-738. Glass. D.. & Singer. J. L’rhan Srrc,ss. New York: Academic Press. 1972. Goldstein. A.. Hansteen. R. W.. & Horns. W. H. Control of methadone dosage by patients. In E. Senay. V. Shorty & H. Alksne (Edsl. D~,rc,/op~nc*nt.sirl r/w Fir/t/ of‘ Dmg A/me. pp. 373.-381. Massachusetts: Schenkman. 1975. Langer. E. J. The illusion of control Jourr~trl of Posorltrlir!, tr~ltl So&r/ Ps~d~ology. 1975. 32. 31 I --328. Langer. E. J. and Rodin J. The effects of choice and enhanced personal responsibility for the aged: A field experiment in an institutional setting. Jormwl qf‘ Pcrxmrliry ctrd Socitrl Ps,~d~o/og,~~. 1976. 34(Z). I9 I-I 98. Staub. E.. Tursky. B.. & Schwartz. G. E. Self-control and predictability: Their effects on reactions to aversive stimulation. Jorrrrztrl of Pmwri~~lir~~ trml So&r/ Psycho/oy~. 197 I. 18. 157-l 62. Stitzer. M.. Bigelow. G.. Lawrence. C.. Cohen. V.. D’Lugoff. B.. & Hawthorne. J. Medication take-home ;IS a reinforcer in a methadone maintenance program. Atltlicrilr Bhwiors. 1977. 2(l). 9-14. Yen. S. Availability of activity reinforcers in a drug abuse clinic. P.s~ddoyiccr/ Rcporrs. 1974. 34(3). 1021-1022.
