Drug and Alcohol Dependence, 18 (1986) 341-348 341

Elsevier Scientific Publishers Ireland Ltd.

EFFECT OF METHADONE DOSE CONTINGENCIES ON

URINALYSIS TEST RESULTS OF POLYDRUG-ABUSING
METHADONE-MAINTENANCE PATIENTS*

MAXINE L. STITZER”,**, WARREN K. BICKEL”. GEORGE E. BIGELOW” and

IRA A. LIEBSON”

“Department of Psychiatry and Behavioral Biology. The Johns Hopkins lJnil1ersit.v School of
Medicine/Key Medical Center and hDepartmcnt of Psychiatry. Albert Einstein College of
Medicine, Baltimore, MD (U.S.A.)

(Received August 8th. 1986)

SUMMARY

Drug abuse outcomes were examined during 2 contingency management

procedures in which the size of the methadone dose was determined by re-
cent urinalysis test results. Twenty polydrug-abusing methadone-mainte-
nance patients were exposed to one of two specific altered dose
consequences: a positive incentive procedure in which dose could only in-
crease above baseline levels as a result of drug-free urines and a negative
incentive procedure in which dose could only decrease below original base-
line levels as a result of drug-positive urines. About 13% of urinalysis test
results were drug-free during a lo-week pre-study baseline period while
about 40-50% of urines were drug-free for both treatment groups during an
18-week intervention period. About half of the subjects in each study condi-
tion showed marked improvement during the intervention, while the other
half failed to improve their urine test results. Treatment failures in the dose
decrease as compared with the dose increase condition were more likely to
drop out of the study. The study showed that size of the methadone dose can
be effectively used as a consequence to influence illicit drug use during
treatment. The study suggests that positive reinforcement as compared with
aversive control procedures can produce an equivalent number of successful
cases while avoiding dropout among patients who fail to respond to treat-
ment.

*Supported under USPHS research grants DA01472 and DA04104, training grant T32 DA07209
and research scientist development award DA00050 from the National Institute on Drug Abuse.
**To whom correspondence should be addressed at: Behavioral Pharmacology Research Unit.
D-5-West, Francis Scott Key Medical Center, 4940 Eastern Ave. Baltimore, MD 21224. U.S.il.

0376-8716/86/$03.50
(‘8 Elsevier Scientific Publishers Ireland Ltd
Printed and Published in Ireland
342

Key words: Methadone dose contingencies - Urinalysis test - Polydrug-

abusing methadone-maintenance patients

INTRODUCTION

Continuing drug abuse during methadone-maintenance treatment is a sig-

nificant clinical problem. Both opiate and non-opiate drugs may be abused,
with one common pattern being abuse of an assortment of sedative and ben-
zodiazepine tranquilizer drugs [ 1,2]. Concurrent abuse of cocaine can also
be anticipated among methadone-maintenance patients based on recent na-
tionwide trends in availability and abuse of this drug. Patients who abuse
illicit drugs during treatment pose a management challenge for clinical
staff, a challenge that is most often met by applying aversive control proce-
dures. Patients are typically denied take-home and other program privileges
while counselors attempt to dissuade them from continuing their illicit drug
use. If drug use does continue, a treatment termination contract may be pre-
sented which specifies a high standard of improvement needed for continued
treatment participation, with treatment termination the result of failure to
meet the contract requirements.
Recent studies have shown that treatment termination contracting can
be an effective strategy for eliminating illicit drug supplementation among
methadone patients [3,4]. However, other strategies that do not withdraw
treatment as a consequence of on-going drug abuse might also be effective.
For example, previous studies from this laboratory have shown that take-
home privileges can promote reduced drug use [5] and improved clinic per-
formance [6] when they are offered contingent upon objective evidence of
behavior change.
Methadone dose alterations might also be effectively used to promote re-
duced illicit drug use among maintenance patients. Since the daily metha-
done dose acts as a mild reinforcer [7], partial dose reductions could be
scheduled as an aversive consequence of continued drug use. Alternatively,
the opportunity to receive dose increases, which also function as reinforcers
for methadone patients [8,9] might be used as a positive incentive for evi-
dence of improved behavior. The purpose of the present study was to com-
pare the effects of a positive and a negative dose incentive procedure on
measures of illicit drug use among a group of chronic polydrug abusers. In
the positive incentive procedure, subjects were offered the chance to raise
their methadone dose over its original stable level by providing drug-free
urine samples, but the dose could not decrease below its original stable
level. In the negative incentive procedure, drug-positive urine samples re-
sulted in partial dose decrements. Drug-free samples could result in restora-
tion of the original dose, but the dose could not increase above its original
stable level. Results of the present study may shed light on the relative
efficacy of positive versus negative incentive procedures. Further, since the
dose alteration procedures employed could be feasibly implemented in any
treatment clinic as a consequence of urine test results, the present study
may have practical importance for developing optimal treatment procedures
at methadone clinics.

METHODS

Subjects
All patients enrolled in our treatment research clinic signed informed
consent at program entry which stated that incentive programs would at
times constitute an important part of their treatment plan and that during
these programs, clinic privileges and treatment components including the
size of their methadone dose might be determined by their drug use and
other behaviors. The 20 patients selected for participation in this study de-
livered more than 50% drug-positive urine samples during a lo-week base-
line evaluation period, with benzodiazepine tranquilizers being the most
frequently detected drug type (67% of drug positive tests), followed by seda-
tive drugs, primarily phenothiazines (19% of positive tests), opiates (11.7”~
of positive tests) and cocaine (2.4% of positive tests). Six subjects were fe-
male and 14 were male. Average age was 33.3 years (range 23-43). Extensive
histories of methadone treatment were common, with an average of 3 prior
treatment admissions. Subjects had been enrolled at the present clinic for
an average of 11.8 months (range 2-23) prior to the start of the study inter-
vention with an average pre-study daily methadone dose of 51.8 mg (range
4&60 mg).

General procedures
The study began in August, 1985 with 17 subjects. Three additional sub-
jects (PO, GP, LT) were started in November, 1985. Throughout. the study.
urine samples were collected 3 times weekly, on Monday, Wednesday and
Friday. Samples were obtained under staff observation and temperature was
tested to ensure the veracity of samples [lo]. All samples were analyzed at
an outside testing laboratory using thin layer chromatography which dr-
tects a wide variety of opiate and nonopiate drugs including morphine,
codeine, hydromorphone, propoxyphene, diazepam, lorazepam, oxazepam
barbiturates, phenothiazines, hydroxyzine pamoate (Vistaril ” ), ethchlorvynol
(Placidyl ” ), and amitriptyline (Elavil ” ). Following a lo-week baseline
evaluation period, subjects were randomly assigned to one of the two-dose
incentive procedures described below and previously approved by the FSK
Institutional Review Board for human research. The procedures were
implemented after a 2-week warning period, and subjects were generalI\
followed for 18 weeks of intervention. At week 13, five subjects originally
assigned to the dose increase condition who were judged to be treatment
failures at that time (JP, MJ, WK. DD, JT) were switched to the dose
decrease condition.
344

Dose incentive procedures

Urine test results were received from the outside testing laboratory by
Thursday afternoon for samples collected on Monday of that week and on
Friday and Wednesday of the previous week. Doses were altered each Mon-
day based on these three most recent test results (Wed, Fri, Mon). Thus, the
delay between sample collection and the altered dose consequence was l-1.5
weeks. Each urine sample that contained detectable quantities of any drug
besides methadone (quinine positives excluded) resulted in a 5-mg dose de-
crease, while each sample free of all supplemental drugs resulted in a 5-mg
dose increase. Shown below are the dose changes that could be earned each
week by all study participants.

Drug-free Altered dose calculation

specimens
Drug-positive Net dose change (mg)
specimens

0 3 - 15
1 2 -5
2 1 f5
3 0 + 15

Subjects in the positive incentive condition could raise their dose to 160%
of its original value by providing drug-free urines. Dose increases were lost
according to the above dose calculation schedule if too many drug-positive
specimens were obtained, but the dose could not decrease below its original
stable value. For subjects in the negative incentive condition, the dose
could be reduced to 40% of its original stable value if drug-positive urines
were provided. The dose could be restored according to the above dose cal-
culation schedule if a sufficient number of drug-free specimens were ob-
tained, but could not increase above its original stable level.

Data analysis
Percent of drug-free urines was determined for each subject during suc-
cessive 2-week blocks of pre- and post-intervention time. An average percent
of drug-free urines was also determined for each subject during the entire
lo-week pre-intervention and 1Sweek post-intervention periods. Missing
data during the intervention for early study dropouts was replaced by aver-
age percent of drug-free urines obtained for that subject during the pre-in-
tervention baseline period. Missing data for reassigned dose increase
subjects during weeks 13-18 was replaced by their average percent of drug-
free urines during intervention weeks 1-12. Effects of the study intervention
were assessed with a repeated measures analysis of variance. The average
percent of drug-free samples obtained for each subject during baseline and
during the intervention was subject to an arcsine transformation before be-
ing entered into data analysis.

RESULTS

As shown in Fig. 1, the average percentage of drug-free samples during

baseline was similar for the 2 study groups and did not exceed 200/o drug-
free samples during any 2-week pre-study baseline period.
Both interventions resulted in improved rates of drug-free sample deliv-
ery, with the average percent of drug-free samples falling generally between
40% and 50% during the intervention. Figure 1 shows that improvements
occurred early in the incentive program, with maximal effects apparent by
study weeks 3-4 and that effects were generally sustained throughout the
first 18 weeks of the intervention period. Repeated measures analysis of
variance using average scores for the baseline and intervention periods t-e-
vealed a significant treatment effect (F(l,lB) = 55.3, P < 0.001). It is evident
from Fig. 1 that there was no difference between the two incentive proce-
dures as far as the overall extent of improvement in drug use was concerned
(F(l,lB) = 0.12, NS).
Individual subject outcomes are presented in Table I. About half of the
subjects in each experimental condition showed substantial improvement
and delivered 50% or more drug-free urine specimens on average during the

100

I
a
5
rn
a,
60

60 I Baseline Intervention

E
e
a”

Study Weeks

Fig. 1. Average percent of drug-free urine samples is shown over successive 2-week blocks of
study time. A IO-week pre-intervention baseline period is shown on the left. Data obtained fol-
lowing announcement of the upcoming study intervention is shown at the point labeled - 2. On
the right are data from 18 post-intervention weeks. Average percent of drug-free samples is
shown separately for subjects exposed to a dose increase (N = 10) and a dose decrease (N =: 10)
incentive procedure. Missing data during the post-intervention period has been replaced by
average percent of drug-free samples observed for that subject during the pre-intervention base-
line period (dose decrease study dropouts) or during the initial weeks of intervention
(reassigned dose increase treatment failures).
346

TABLE I

PERCENTAGE OF DRUG-FREE URINE SAMPLES

Dose decrease Dose increase

Subjects Baseline Intervention Subjects Baseline Intervention

LC 14.6 88.9 JR 32.0 94.4

PO 18.2 86.3 TM 27.6 81.4
DL 0.0 69.8 KL 14.0 79.6
MB 23.8 63.6 WD 0.0 66.7
JD 0.0 54.9 JP 20.6 44.3
WB” 29.0 31.3 MJ 6.6 19.5
LW” 30.6 30.6 WK 5.0 14.5
SP” 10.0 7.8 DD 6.5 5.5
GP 10.0 5.8 LT 3.4 1.9
NS” 0.0 0.0 JT 0.0 0.0
Average 13.6 43.9 Average 11.6 40.8

“Dropped out by intervention week 5.

intervention period (subjects LC, PO, DL, MB and JD in dose decrease; sub-
jects JR, TM, KL, and WD in dose increase). The remaining subjects showed
no substantial improvement during the intervention. Nor did dose increase
treatment failures show improvement when switched to the dose decrease
condition,
A qualitative difference in outcomes for study failures in the 2 experi-
mental groups was also apparent. Four subjects exposed to the dose de-
crease condition (WB, LW, SP, NS) dropped out of the study by the fifth
intervention week either by arranging transfer to another clinic (N = 2) in-
curring a disciplinary detox action at this clinic (N = 1) or enrolling tempo-
rarily in an inpatient sedative detox program (N = 1). No subjects assigned
to the positive incentive dose increase condition dropped out.

DISCUSSION

This study showed that polydrug abuse among methadone-maintenance

patients was reduced from pre-intervention levels by a contingency manage-
ment program in which the size of the daily methadone dose was determined
by recent urinalysis test results. Average percentage of drug-free urine re-
ports increased about 30% from 12.6% during baseline to 42.4% during the
period when the altered dose consequence was in effect. These results, ob-
tained with a group of chronic polydrug-abusing patients, suggest that
dosage contingencies might be usefully applied to more heterogeneous
groups of methadone patients enrolled in non-research clinics as a strategy
for suppressing illicit drug use during treatment.
 347

The present study provided a unique opportunity to compare the effects

of a positive and a negative incentive procedure, both involving methadone
dose alteration. The positive (dose increase) and negative (dose decrease)
consequences produced a similar number of successful cases, suggesting
equal efficacy. Equivalence of the 2 procedures is further supported by the
fact that dose increase treatment failures did not improve when switched to
the dose decrease condition. However, the positive and negative incentive
procedures had differential effects on patient retention. Specifically, the
negative incentive procedure involving dose decreases appeared to promote
dropout from the study and from treatment at this clinic. Although a no-
change assumption was used to substitute missing data for study dropouts
in the present study, patients who drop out of methadone treatment and do
not transfer to other programs are in fact likely to have worse outcomes on
measures of drug use, crime and mortality than patients who remain in
methadone treatment. This suggests that positive reinforcement procedures
may have an advantage as compared with negative incentive procedures
since an equivalent number of successful cases may be obtained, while treat-
ment dropout among non-responders may be avoided. The generality of
treatment retention effects will have to be tested in larger samples and
across interventions involving different types of reinforcers and punishers
since the study dropout effect observed in the present study could be specifi-
cally related to the fact that methadone dose decreases were involved in the
intervention procedure.
Important individual differences in treatment performance were observed
during this study. About half the subjects in each group showed marked im-
provement, delivering 50% or more drug-free urines during the intervention,
while the remaining subjects either left the clinic or showed only marginal
improvement at best. These individual differences in outcome are consistent
with other recent reports concerning effects of contingency management in-
terventions including treatment termination contracting [ 31 and take-home
incentive procedures [6]. The fact that individual patients appear to respond
to treatment interventions in an all-or-none fashion suggests that different
approaches will be needed for non-responders. For example, shaping proce-
dures that reward smaller amounts of behavior change might be effectively
employed.
Baseline data for the present study subjects highlights the poor perfor-
mance that can occur when no consequences are attached to urine results.
while the improvement noted during the study intervention suggests that
systematic consequences attached to urine results have desirable therapeu-
tic consequences. While the study shows that dosage consequences can be
effective, it does not distinguish between the specific effects of the proce-
dures employed and the non-specific effects of starting any new treatment
intervention. It is possible that the specific content of the intervention used
may be less relevant than the fact that some consequence is attached to uri-
nalysis test results. Consistent with this notion is a recent report by
348

McCarthy and Borders [4] where improved treatment outcomes were ob-
tained using a very simple set of structured consequences that required pa-
tients to be drug-free for at least 1 out of every 4 treatment months. Similar
effects of the positive and negative incentive procedures noted in the
present study is also consistent with this notion.
It is also possible, however, that some reinforcers and punishers available
at the methadone clinic for use in contingency management interventions
are more potent than others as agents of behavior change. Previous re-
search from this laboratory suggests that methadone dose alteration may in
fact be a relatively weak consequence for methadone-maintenance patients
as compared with take-home opportunities [5] or menus offering a choice of
reinforcer consequences [6]. More research is needed to determine whether
different specific types of treatment procedures (e.g. contingent vs. non-con-
tingent) or different types of consequences used in contingency management
procedures can have differential effects on outcome measures during drug
abuse treatment.

REFERENCES

1 B. Stimmel, M. Cohen and R. Hanbury, Ann. N.Y. Acad. Sci., 311 (1978) 99.
2 M.L. Stitzer et al., Drug Alcohol Depend., 8 (1981) 189.
3 M.P. Dolan et al., J. Consult. Clin. Psychol., 53 (1985) 549.
4 J.J. McCarthy and O.T. Borders, Am. J. Psychiatry, 142 (1985) 1419.
5 M.L. Stitzer, G.E. Bigelow and I. Liebson, Addict. Behav., 4 (1979) 245.
6 M.L. Stitzer et al., J. Appl. Behav. Anal., 15 (1982) 493.
7 M.E. McCaul et al., Clin. Pharmacol. Ther., 31 (1982) 753.
8 M.L. Stitzer et al., Clin. Pharmacol. Ther., 34 (1983) 29.
9 W.K. Bickel, S.T. Higgins and M.L. Stitzer, Drug Alcohol Depend., 18 (1986) 165.
10 B.A. Judson, D.U. Himmelberger and A. Goldstein, Am. J. Drug Alcohol Abuse, 6 (1979) 197.