Full text of DOI:10.1111/j.1753-6405.2001.tb01832.x

Drugs and Alcohol
Withdrawal from methadone maintenance treatment:
prognosis and participant perspectives
Michael Lenné and Nick Lintzeris
Abstract
Turning Point Alcohol & Drug Centre Inc, Victoria
Objective: To determine the proportion of
clients engaged in methadone maintenance
treatment who have favourable prognosis
for withdrawal, and to examine client
perceptions and expectations of withdrawal.
Methods: A broad cross-section of 856
methadone clients was sampled across
Melbourne, Sydney and Brisbane. Self-
complete surveys were developed for the
clients, their clinic staff or pharmacists, and
methadone prescribers. The client survey
examined aspects of the clients’
Courtney Breen
National Drug & Alcohol Research Centre, New South Wales
Simon Harris
Alcohol & Drug Service, Peel Street Clinic, Queensland
Lynn Hawken
Alcohol, Tobacco & Other Drug Services, Gold Coast Hospital, Queensland
Richard Mattick
National Drug & Alcohol Research Centre, New South Wales
perspectives of withdrawal, and the surveys
for the service providers collected
information about each client’s current
treatment episode. Informed consent was
provided by clients to obtain information from
their clinic staff member or pharmacist, and
their methadone prescriber.
Alison Ritter
Turning Point Alcohol & Drug Centre Inc, Victoria
Results: Most clients (70%) were at least
very interested in methadone withdrawal.
Clients were also more optimistic about
their own post-withdrawal outcomes (in
terms of opioid use) than both their clinic
staff and prescribing doctors. Clinical
criteria indicated that 31% of clients had a
reasonable prognosis for withdrawal.
However, when considering all factors, 17%
had good withdrawal prognosis, were
interested in methadone withdrawal, and
believed it was very likely they would
remain opioid-free for three months post-
withdrawal.
he number of clients enrolled in examined client and clinician perspectives
methadone maintenance treatment of methadone withdrawal, although with-
T
(MMT) has progressively increased drawal has been identified by methadone
in Australia and other countries in the past clients in Melbourne as the major problem
decade. In Australia, the numbers enrolled with methadone treatment.2
in methadone programs have increased from
,000 in the mid-1980s to more than 25,000.1 drawal outcomes are generally poor, as evi-
A difficulty confronting methadone pro- denced by low rates of successful completion
Overseas research suggests that with-
3
grams is the capacity for the ongoing expan- of the withdrawal regime and high rates of
sion of treatment places within the context relapse to heroin use.3-7 Furthermore, the
of finite resources. Demand for methadone evidence regarding abstinence rates post
treatment continues to outstrip the supply of withdrawal suggests that only approximately
services, with reduced access to treatment one quarter of clients remain abstinent from
and less intensive treatment services as a heroin use for an extended period of time
Conclusions: Despite the likely continued
increase in client numbers in substitution
maintenance treatment, the majority of
methadone clients have a poor prognosis
for withdrawal and should not be
potential consequence.
(e.g. three months or more).3
One possible mechanism to address this
If we are to encourage clients to withdraw
problem is to encourage withdrawal from from MMT it is important to direct our in-
MMT. Hence, a better understanding regard- terventions to those clients who could be
ing successful withdrawal from methadone considered as having a better prognosis for
treatment is required, in particular, the per- successful withdrawal. It is possible to iden-
ceptions and expectations of methadone cli- tify better withdrawal outcomes in certain
ents and their treating clinicians. There has client groups from the literature. Milby re-
been no previousAustralian research that has viewedAmerican studies from 1970 to 1985
encouraged to cease treatment.
Implications: Clients who do not meet key
clinical criteria are likely to have poor
clinical outcomes regardless of how
withdrawal is attempted.
(Aust N Z J Public Health 2001; 25: 121-5)
Correspondence to:
Paper submitted: May 2000
Revision requested: November 2000
Accepted: January 2001
Dr Michael Lenné, Research Fellow, Accident Research Centre, Monash University, Victoria
3800. Fax: (03) 9905 4363; e-mail: Michael.Lenne@general.monash.edu.au
2001 VOL. 25 NO. 2
AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH
121

Lenné et al.
Article
that reported completion rates for clients withdrawing from MMT.3
Completion rates were considerably higher when withdrawal was
undertaken with the consent of treating staff (50%) than when
reductions were not approved by staff (19%). It is difficult to
operationalise the term ‘staff consent’, but in general this term
relates to agreement by the staff and client that an attempt at metha-
done withdrawal is warranted, given that the client has ceased
their use of heroin and other drugs, has a stable psychological
profile and a reasonable level of social functioning (such as em-
ployed, with non drug using supports). Withdrawal outcomes are
Queensland are treated in large clinic settings, facilitating the re-
cruitment of large numbers of survey participants in a short pe-
riod of time. The clinics chosen were the Peel Street Clinic and
Southside Clinic in Queensland, and the Langton Centre, Towers
and Coopers clinics in New South Wales. In Victoria, the vast
majority of clients receive their methadone through community
2
pharmacies. Clients in Victoria were recruited through 24 se-
lected pharmacies in metropolitan Melbourne.
All clients in participating community pharmacies (Victoria)
and clinics (New South Wales and Queensland) were approached.
The response rates for each State were: 224 of 611 (37%) in Vic-
toria; 237 of 513 (46%) in New South Wales; and 395 of 1064
(37%) in Queensland. This reflects an overall response rate of
39%.
8
also optimised following long-term maintenance and for clients
9
with positive beliefs about their withdrawal outcomes. Conversely,
clients with certain clinical conditions can be considered as poor
candidates for withdrawal from methadone, such as being preg-
nant, those with concomitant chronic pain conditions and other
conditions that benefit from continued maintenance treatment with
opioids, and those with serious medical or psychiatric conditions
The research team developed surveys for methadone clients,
clinic staff and pharmacists, and prescribing doctors. All surveys
were designed to be self-completed. It was a priority to match
each client’s record with information from their clinic staff mem-
ber and doctor. It was therefore necessary to obtain informed con-
sent from the clients to contact their service providers, with
assurances that the information they had provided would not be
available to their clinic staff or doctor.
(such as schizophrenia).
This study aimed to determine the proportion of clients in MMT
in Australia who have favourable prognosis for withdrawal with-
out relapse to heroin or other dependent drug use, or deteriora-
tion in their medical, psychiatric or social well-being. The study
also aimed to examine client perceptions and expectations of with-
drawal by assessing the level of interest among clients in with-
drawing from methadone, and the likely outcomes of attempts at
methadone withdrawal as perceived by clients and their treating
clinicians.
Operational criteria for positive withdrawal
prognosis
The research team identified and operationalised a number of
clinical criteria upon which to assess a client’s suitability for
attempting withdrawal from methadone. These criteria were
developed from previous research literature regarding metha-
done withdrawal outcomes and through consensus among ex-
perienced clinicians on the research team (see Table 1). On the
basis of the client, clinic staff, and prescriber data, clients who
were both suitable for and interested in methadone withdrawal
were identified.
Methods
Sampling technique
A broad cross-section of MMT clients across three Australian
States was sampled. There were differences in the sampling
techniques for each State, reflecting different treatment systems.
The majority of methadone clients in New South Wales and
Table 1: Operational criteria used to determine positive withdrawal prognosis.
Criteria
How measured
Rationale
No regular use of heroin
Use of heroin less than once a week
in the preceding three months: Urine
drug screen and self-report
High risk of relapse to regular heroin use
during or following methadone withdrawal
No regular use or abuse of other
substances such as alcohol,
benzodiazepines or psychostimulants
No dependence to other drugs (in
particular alcohol or benzodiazepines):
DSM-IV criteria
High risk of increased use of other drugs
during or following methadone withdrawal
No significant psychosocial dysfunction
Stable accommodation, employment,
supportive networks: Global Assessment
of Functioning Score (GAF) >60 14
Clients experiencing significant social
dysfunction are unlikely to cope with the
additional strain of methadone withdrawal
No significant medical or psychiatric
condition that either necessitates
continuation on an opioid, or complicates
withdrawal from methadone
Medical or psychiatric condition that may
be destabilised by their termination of
methadone treatment (such as chronic pain,
schizophrenia, bipolar affective disorder):
Clinical history and examination
Risk to the client of deterioration in physical
and/or mental health following withdrawal
from methadone
Not pregnant or breastfeeding
Clinical history and examination
Physical and psychological risks to the mother
and foetus of withdrawal during pregnancy
Client in MMT for more than six months
Confirmation of treatment records
Client outcomes are enhanced where
methadone treatment >6 months
1
22
AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH
2001 VOL. 25 NO. 2

Drugs and Alcohol
Withdrawal from methadone maintenance treatment
Results
Table 3: Client, clinic staff and doctor ratings of the
likelihood of clients remaining opioid free for three
months after methadone withdrawal (n=547). Data are
presented as percentages.
The number of client surveys returned was 856. Complete clinic
staff and doctor surveys were received for 547 (64%) of the cli-
ents surveyed. Analyses were conducted to examine if there were
differences between clients with complete doctor and clinic staff
data compared with the total client sample (t-tests and chi-square
analyses). There were no significant differences in client gender,
age, nor level of interest in withdrawal between these groups. All
subsequent results will focus on the complete dataset; that is, the
Likelihood of
remaining
Client
ratings
Clinic
staff
Doctor
ratings
opioid free
ratings
Not at all likely
12.4
17.0
19.4
24.0
27.2
21.3
38.1
23.7
13.3
3.6
23.7
31.6
25.2
16.1
3.4
Unlikely but possible
Moderately likely
Considerably likely
Extremely likely
5
47 clients for whom complete client, clinic staff, and doctor
datasets were obtained.
There was a significant difference in the age of the clients from
New SouthWales (mean = 33.19 years ±7.51), Queensland (mean
=
35.77 years ±7.22), and Victoria (mean = 36.86 years ±8.06)
(
F2,530 = 8.99, p<0.001). Post-hoc tests indicated that clients in
could stay off heroin and methadone for three months if he/she
withdrew from methadone? The distribution of responses is pre-
sented in Table 3.
New South Wales were younger than clients in Queensland and
Victoria (p<0.05). The slight majority of clients were male
(
56.8%), and there were no differences in the proportion of males
Approximately half (51.2%) the clients surveyed across the three
States indicated that it was at least considerably likely that they
would be able to remain opiate free for three months after metha-
done withdrawal. The distribution of service provider responses
is quite different from the client ratings. Clinic staff and doctors
indicated that only 16.9% and 19.5% of clients respectively would
be considerably or extremely likely to stay off opiates for three
months.The more optimistic ratings by the clients was highlighted
2
and females across the three states (χ (2) = 3.19, p>0.05).
The mean current methadone dose was 60.0 mg (±31.0), and
clients had been in their current MMT for an average of 32.8
months (±35.4). The mean GAF score was 73.3 (±13.3), indicat-
ing a reasonably high level of psychosocial functioning. The
majority of clients (85.1%) were also eligible for take away doses
in accordance with the criteria used in each State. Under half of
the clients (42.6%) were judged by their doctors not to be engag-
ing in any parenteral opioid use, and only 15.9% were thought to
do so more than once a week. Significant proportions of clients
were currently experiencing a severe medical or psychiatric con-
dition (15.7%) or were dependent upon other substances (20.9%).
Fewer clients experienced chronic pain (8.2%) or were pregnant
or lactating (2.3%).
10
by analysis using the Kendall Coefficient of Concordance, which
confirmed that there was not significant agreement between the
2
three groups of ratings (χ (2) = 214.19, p<0.001). This high level
of client optimism is consistent with the extremely high levels of
client interest in withdrawal.
Proportion of clients with good withdrawal
prognosis
Client interest in withdrawal
An estimate of the proportion of clients that would be clini-
cally suitable for methadone withdrawal was made in accordance
with the criteria established inTable 1. A total of 171 clients from
the total of 547 (31.3%) satisfied these criteria and could be con-
sidered to have a reasonable prognosis for withdrawal from metha-
done. The reasons why the remaining 376 clients (68.7%) were
not suitable for withdrawal are illustrated in Table 4. Unstable
drug use was clearly the main reason why clients were not suit-
able for withdrawal. The majority of clients (80%) who were not
suitable did not meet two or more of the criteria.
The clients were asked How interested are you currently in with-
drawing from methadone? The client interest in methadone with-
drawal was very high, with 70% of clients in the survey indicating
very high or extreme interest in withdrawing from methadone
(
see Table 2).
Perceived withdrawal outcomes
Clients were asked to rate on a Likert scale How likely is it that
you could stay off heroin and methadone for three months if you
withdrew from methadone? Their clinic staff and prescribing
doctors were also asked to rate How likely is it that this client
From the group of 171 clients who had good prognosis for with-
drawal, only 92 clients (16.8% of total sample, 53.8% of those
with good prognosis) were also considerably or extremely inter-
ested in withdrawal, and believed that it was considerably or
extremely likely they would remain opioid-free for three months
post-withdrawal.
Table 2: Client interest in methadone withdrawal (n=547).
Level of interest
Number
Per cent
Not at all interested
Slightly interested
Moderately interested
Very interested
56
38
10.2
7.0
68
12.4
21.0
49.4
Discussion
The key findings from this study are that more than 70% of
clients report considerable interest in withdrawing from MMT,
115
270
Extremely interested
2001 VOL. 25 NO. 2
AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH
123

Lenné et al.
Article
Table 4:The proportion of clients (who were classified as
not suitable for withdrawal) who did not meet each
criteria (n=376). Note: these are not mutually exclusive
categories.
client’s medical or psychiatric condition. It is worth noting, how-
ever, that clinician estimates of heroin use in the preceding few
months in this study were comparable to self-reported heroin use
of methadone clients in previously reportedAustralian studies.2,11
Operational criteria
Percentage of
clients who
did not meet
each criteria
There may also be some dispute regarding the validity of the
criteria used to adjudge prognosis for methadone withdrawal.
While the clinical criteria described in Table 1 can be supported
by (limited) research evidence and expert clinical consensus, well-
conducted research has not clearly established these criteria as
predicting outcomes following withdrawal, and it is possible that
the criteria may be over or under-inclusive. Counter to this con-
cern is the finding that most clients (80%) were considered to
have poor prognosis for withdrawal due to a combination of rea-
sons, with only a minority being ‘excluded’ for only one reason,
indicating that minor alteration of the prognostic criteria would
not significantly alter the findings.
Unstable parenteral opioid use
53.7 (n=202)
30.9 (n=116)
29.8 (n=112)
Dependence to other substances
Some level of psychosocial dysfunction
Experiencing a medical or psychiatric condition 23.3 (n=88)
Less than six months in MMT
14.6 (n=55)
12.3 (n=46)
Experiencing chronic pain
(requiring continued MMT)
Pregnant or lactating
3.4 (n=13)
Despite the potential shortcomings of this study, there are a
number of clear trends that emerge. There is a high level of inter-
est among clients to withdraw from methadone, which may be
the result of several factors.The demands of participation in MMT
on a long-term basis are not insignificant for clients. Clients must
attend regularly for dosing and treatment reviews; a proportion
will experience unwanted side effects to methadone; many will
experience stigma from family, friends and the broader commu-
nity for being ‘on methadone’; and there is considerable finan-
cial cost involved for some clients (e.g. treatment costs account
for an average 20% of methadone client’s legitimate income in
and that there is considerable discrepancy between clients and
their treatment providers regarding the perceived outcomes fol-
lowing withdrawal, with the majority of clients (51.2%) report-
ing that withdrawal would be unlikely to lead to relapse to heroin
use. In contrast, treating clinicians and doctors estimated that only
a minority of clients (16.9% and 19.4% respectively) would be
unlikely to relapse within three months of ceasing MMT. Client
prognostic factors indicated that only a minority of clients (31%)
had favourable prognosis for methadone withdrawal, and further,
only 17% had all three factors: good prognosis, interest in with-
drawal, and positive post-withdrawal expectations regarding opioid
use.
2
Victoria ). However, the high level of interest in methadone with-
drawal should not necessarily be interpreted as methadone treat-
ment being a ‘poor’ treatment modality. Other areas of medicine
reveal that many people with chronic medical conditions (such as
diabetes, asthma, Crohn’s disease) have ambivalent feelings
towards their treatment and often adhere poorly with treatment
regimes.13
Before considering the implications of this research, it is nec-
essary to examine a number of potential limitations of the study.
One issue is the extent to which this sample of methadone clients
is representative of the broader population of Australian metha-
done clients. The response rate of 39%, similar across the three
States, raises questions about the extent to which there may be
differences between those clients who completed the survey and
those that did not. Furthermore, only urban clients from clinic-
based treatment settings were recruited to the study, and it is pos-
sible that sampling of rural clients may yield different findings.
Nonetheless, comparison between this sample of methadone cli-
ents and previously published accounts of Australian methadone
Another trend emerging from the results is the greater client
optimism regarding their ability to successfully withdraw from
methadone and remain opiate free than their treatment providers.
This may reflect undue optimism by clients and/or undue pessi-
mism by clinicians. Several factors suggest the former. First, pre-
vious research literature suggests that only a minority of clients
can successfully withdraw from methadone and remain opiate free.
Second, approximately 50% of clients indicated that they were
‘considerably’or ‘extremely’likely to remain opiate free for three
clients indicates that there are no marked differences on variables
such as age, gender, duration on program, and methadone dose.2,11
months following methadone withdrawal; whereas, a much smaller
proportion of clients in this sample (31%) met favourable prog-
nostic criteria for withdrawal.
Another potential limitation is the validity of the data collected
from the treating clinicians for the purpose of estimating with-
drawal prognosis. The uncertainty here is the extent to which the
treating clinicians were actually aware of their client’s drug use,
medical conditions or psychosocial functioning. Many clients
conceal their drug use from their methadone providers (for exam-
ple in order to retain takeaway dose privileges). A considerable
It is possible to extrapolate from these results and to consider
policy implications regarding the capacity for methadone clients
to successfully ‘leave the treatment system’. Although 31% of
clients had good prognosis for withdrawal, this does not take into
account client interest and expectancies about withdrawal.
A smaller proportion of clients (17%) had good prognosis for
withdrawal, were very interested in methadone withdrawal, and
believed it was very likely that they would remain opiate free
proportion of clients do not utilise their treating methadone pre-
scribers for their general health care, in Victoria at least,12 and as
such, methadone prescribers may be unaware of aspects of the
1
24
AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH
2001 VOL. 25 NO. 2

Drugs and Alcohol
Withdrawal from methadone maintenance treatment
post-withdrawal. However, of these, approximately half would
relapse into heroin use within months of completing their with-
users to treatment, particularly among those individuals previ-
ously unwilling to consider MMT. The likely continued expan-
sion in treatment numbers requires that program administrators
consider the relevant resource implications of providing treatment
to increasing numbers of heroin users.
3
drawal, suggesting that less than 10% of all existing methadone
clients would have a favourable outcome (ceasing methadone treat-
ment and not relapsing to opiates) using conventional approaches
to methadone withdrawal. It would seem therefore that the vast
majority of methadone clients in Australia (85% to 90%) will
benefit from continued methadone treatment. Calls from various
sectors in the community and in drug treatment services to
encourage clients off methadone programs or to limit the dura-
tion of treatment programs are poorly founded, with only a
minority of clients likely to benefit from such a policy shift.
There is likely to continue to be an expansion of client num-
bers in substitution maintenance treatment in Australia. As stated
previously, the majority of methadone clients have a poor prog-
nosis for withdrawal and should not be encouraged to cease main-
tenance treatment. Even if new techniques are developed that
improve rates of successful withdrawal completion, clients who
continue to frequently use heroin, are dependent on other drugs,
have medical or psychiatric complications, or severe psychoso-
cial dysfunction are likely to have poor clinical outcomes,
regardless of how withdrawal is attempted. Findings from this
study suggest that it is only a minority of clients who are likely to
benefit from improved withdrawal techniques, and this number is
unlikely to consistently match the continued increase in metha-
done treatment numbers in Australia (in recent years numbers in
Australian methadone programs have increased by 15% to 20%
In summary, there is considerable interest and optimism among
most methadone clients about ceasing methadone treatment,
despite the findings of this study that suggest otherwise.
References
1
.
National Drug Strategy. National Policy on Methadone Treatment. Canberra:
AGPS; 1997.
2
.
Ezard N, Lintzeris N, Odgers P, et al. An evaluation of community methadone
services in Victoria, Australia: Results of a client survey. Drug Alcohol Rev
1999;18:417-23.
3
.
Milby JB. Methadone maintenance to abstinence: How many make it? J Nerv
Ment Dis 1988;176:409-22.
4. Des Jarlais DC, Joseph H, Dole VP. Long term outcomes after termination
from methadone maintenance treatment. Ann N Y Acad Sci 1981;362:231-8.
5
6
.
.
Kleber H, Riordan C, Rounsaville B, et al. Clonidine in outpatient detoxifi-
cation from methadone maintenance. Arch Gen Psychiatry 1985;42:391-4.
Johnson RE, Jaffe JH, Fudala PJ. A controlled trial of buprenorphine treat-
ment for opioid dependence. J Am Med Assoc 1992;267:2750-5.
7. Senay EC, Dorus W, Goldberg F, Thornton W. Withdrawal from methadone
maintenance: Rate of withdrawal and expectation. Arch Gen Psychiatry
1
977;34:361-7.
8
9
1
1
.
.
Ward J, Mattick RP, Hall W. Methadone Maintenance Treatment and Other
Opioid Replacement Therapies. Amsterdam: Harwood Academic; 1998.
Milby JB, Hohmann AA, Gentile M, et al. Methadone maintenance outcome
as a function of detoxification phobia. Am J Psychiatry 1994;151:1031-7.
0. Siegel S, Castellan NJ. Nonparametric Statistics for the Behavioral Sciences.
Singapore: McGraw-Hill; 1988.
1. Bell J, Ward J, Mattick RP, Hay A, Chan J, Hall W. An Evaluation on Private
Methadone Clinics. Canberra: AGPS; 1995. National Drug Strategy Report
Series No.:4.
1
each year ). The prospect of improved techniques that facilitate
1
2. Lintzeris N, Koutroulis G, Odgers P, et al. Report on The Evaluation of Com-
munity Methadone Services in Victoria. Victoria: Turning Point Alcohol &
Drug Centre; 1996.
withdrawal from methadone will have obvious benefits to indi-
vidual clients, but are unlikely to result in a reduction in client
numbers in maintenance programs. Furthermore, the introduc-
tion of new maintenance pharmacotherapies in Australia (such as
buprenorphine and LAAM) is likely to result in an influx of heroin
1
1
3. O’Brien CP, McLellan AT. Myths about the Treatment of Addiction. Lancet
1996;347:237-40.
4. Jones SH, Thronicroft G, Coffey M, Dunn G. A brief mental health outcome
scale-reliability and validity of the Global Assessment of Functioning (GAF).
Br J Psychiatry 1995;166:654-9.
2001 VOL. 25 NO. 2
AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH
125