10 October 2003

Cost effectiveness of buprenorphine versus methadone. Both 'winners'.

Buprenorphine versus methadone maintenance: a cost-effectiveness analysis. Doran CM, Shanahan M, Mattick RP, Ali R, White J, Bell J. Drug Alc Dependence 2003 71;3:295-302

Dear Colleagues,

This monumental study was one of the steps required in the licensing of buprenorphine in Australia. It is a credit to the authors who went against almost all previous studies of comparative opioid addiction treatments in keeping the randomised interventions 'naturalistic' rather than arbitrary, fixed or even placebo. Doses were not capped and the initial period was double blind. Pregnant women, dual diagnosis folk and those currently in agonist treatment were exluded.

This analysis is of the costs versus benefits and has a brief but important literature review of the subject. The summaries are worth quoting: "The most comprehensive cost-benefit analysis to date … by Gerstein [CALDATA study] … found that for each modality of treatment the summed benefits significantly exceeded the cost of delivering the episode of care. For residential treatment the ratio of benefits to costs was 4.8. Comparable ratios were obtained for social model treatment and for continuing methadone. Much higher ratios of 11 to 1 and 12.6 to 1 were estimated for outpatient and discharged methadone participants."

[In this study:] "Treatment with methadone was found to be both less expensive and more effective than treatment with buprenorphine, which suggests methadone dominates buprenorphine. However, [this may not be] statistically significant. The results of this study . indicate that buprenorphine provides a viable alternative to methadone in the treatment of opioid dependence."

The authors could not directly test the cost of those opioid dependent patients who do not do well on methadone treatment. Other studies indicate that they are more likely to die and to be involved in high risk behaviours. Hence an alternative as effective as methadone must be a boon and would increase the cost-efficiency even further, even in the case of slightly higher costs of buprenorphine and its delivery.

Comments by Andrew Byrne ..

Reference: Gerstein DR, Johnson RA, Harwood HJ, Fountain D, Suter N, Mallow K. 1994 Evaluating Recovery Services: The California Drug and Alcohol Treatment Assessment (CALDATA). Report by NORC at University of Chicago and Lewin-VHI Inc, Fairfax Virginia.

Heavy drinking and illicit drug use in London methadone patients.

Heavy drinking and illicit drug use common in London methadone patients - were doses adequate? Research summary by Dr Richard Hallinan.

Excessive alcohol consumption and drinking expectations among clients in methadone maintenance. Hillebrand J, Marsden J, Finch E, Strang J. Journal of Substance Abuse Treatment 2001 21;3:155-60

Dear Colleagues

This study investigated the prevalence and psychological determinants of alcohol consumption in 66 MMT clients in South London in 1999, using the framework of the Theory of Reasoned Behaviour (Ajzen and Fishbein, 1980). By a sensitive criterion of 3 or more of 8 DSM-IV markers of alcohol dependence, 54% of surveyed clients were alcohol dependent in the previous 12 months. In some cases this period could have preceded the time in MMT, which together with the sensitive criterion of alcohol dependence may have given a liberal estimate of this problem in MMT clients, however the results are consistent with many previous studies showing a high prevalence of alcohol dependency in MMT clients in the US and UK.

The study participants had been in treatment at least a month (mean 28.4 months), and were dosed at community pharmacies with a mean dose of 49mg (SD 27mg). There were also high levels of recent use of heroin (62%), stimulants (47%) and benzodiazepines (32%).

The researchers examined the clients' perceived functions for alcohol use, and found 84% used it to relax, 68% to relieve boredom, 66% to improve low mood, 64% to forget problems, 54% to help them sleep and 30% to increase the effect of methadone; 28% to get going in the morning; 24% to stop feeling sick in the morning. 24% used it to calm down after using other drugs.

The strongest predictor of clients' expectation of change in their own drinking was 'subjective norms', suggesting their susceptibility to the views of 'important others' around them, potentially including MMT staff. The use of alcohol to perform particular functions was also a strong predictor of low expectation of change, suggesting that if these specific functions could be served in other ways, clients may be more susceptible of change in their drinking behaviour. The dose of methadone was significantly and negatively related to expectation of change in drinking.

The mechanism for determination of methadone doses in this client group is not specified in the report; in particular it is not apparent whether there were actual or perceived ceilings to methadone dosing.

The MMT clients reported in this study apparently received a considerable range of doses but the mean dose was lower than the minimum 60mg daily recommended by the US National Institutes of Health consensus panel guidelines (1997). Only 35.5% of MMT clients in the US were receiving doses less than 60mg daily in 2000, down from 79.5% in 1988 (D'Aunno and Pollack 2000). Ball and Ross (1991) found that prevalence of heroin use fell to under 10% only with MMT doses higher than 50mg daily. In a study of 211 MMT patients Eap et al (2000) found a trough R,S- methadone level of 400 ng/ml was a significant therapeutic threshold with an 81% specificity for absence of heroin use, a level achievable with a mean dose of circa 80mg daily (Wolff et al 1991).

Hillebrand et al's study should thus be examined through the prism of possible underdosing, which is strongly suggested by the high prevalence of continuing heroin use. Most of the specific perceived functions of alcohol use which they identified (viz to relax, relieve low mood, get to sleep, get going or stop feeling sick in the morning, to augment the effect of methadone) could be interpreted as self medication of symptoms of abstinence. Borg et al 1995 found methadone levels under 150 ng/ml in 9 of 10 patients with such symptoms, and reported "Early opioid withdrawal, requiring a higher dose of methadone, is often difficult to diagnose because many of the symptoms are also symptoms of other syndromes common in the methadone maintenance population."

The authors identified a negative correlation of dose of methadone with expectation of change in drinking behaviour, which they speculate may reflect a greater 'attachment' to their drinking behaviour in higher dose clients. This finding needs scrutiny considering the limited information given about dosing practices. On the figures given, at least 95% of the clients had doses less than 103 mg daily (ie mean+2xSD). In a situation where there is considerable dosing flexibility within actual or perceived limits, higher dose clients are more likely to include those limited by a 'ceiling'. Thus, paradoxically, higher doses could be associated with a greater likelihood of underdosing, explaining the greater 'attachment' of some higher dose clients to their drinking. The fact that the association of high dose with lower expectation of change was independent of perceived functions of alcohol, need only imply that the function of alcohol in alleviating symptoms of abstinence is not always perceived as such.

The strongest positive predictor of expectations was found to be subjective norms, evidence of the importance of the MMT environment in addressing problematic alcohol consumption. The power of subjective norms also suggests a mechanism by which clients might limit their own doses to 'acceptable' limits and fail to perceive the extent to which alcohol is used to alleviate symptoms.

Clients who felt they were using alcohol for specific functions were less likely to expect change in their drinking. The simplest alternative to alcohol for many of the functions identified by the researchers might be methadone dose titration.

In contrast to the good evidence that adequate doses of methadone can reduce cocaine use, there is only a modest literature of observational studies suggesting the same for alcohol and benzodiazepines Byrne 1998 Maxwell and Shinderman 1999, Tennant 1987, Stimmel et al 1982, Gelkopf et al 1999. Recently Lubrano et al 2002 et al have described high levels of craving for alcohol in MMT patients receiving low doses. Hillebrand et al's simple and elegant study highlights the need for further research to evaluate the issue of adequacy of methadone dose in dealing with the important problem of alcohol use in MMT.

Comments by Dr Richard Hallinan, 75 Redfern St, Redfern, 2016


Ajzen, I., and Fishbein, M. (1980) Understanding attitudes and predicting social behaviour. London: Prentice-Hall International.

Ball, J.C., and Ross, A. The effectiveness of Methadone Maintenance Treatment. New York: Springer-Verlag, 1991.

Borg L, Ho A, Peters JE, Kreek MJ. Availability of reliable serum methadone
determination for management of symptomatic patients. J Addict Dis 1995;14(3):83-96

Byrne, A. Use of serum levels for optimising doses in methadone maintenance treatment. J Main Addict 1998; 1: 13-4.

D'Aunno T, Pollack HA. Changes in methadone treatment practices: results from a national panel study, 1988-2000. JAMA. 2002 Aug 21;288(7):850-6.

Eap CB, Bourquin M, Martin J, Spagnoli J, Livoti S, Powell K, Baumann P, Deglon J. Plasma concentrations of the enantiomers of methadone and therapeutic response in methadone maintenance treatment. Drug Alcohol Depend. 2000 Dec 22;61(1):47-54.

Gelkopf M, Bleich A, Hayward R, Bodner G, Adelson M. Characteristics of benzodiazepine abuse in methadone maintenance treatment patients: a 1 year prospective study in an Israeli clinic. Drug Alcohol Depend. 1999 Jun 1;55(1-2):63-

Maxwell S, Shinderman M. Optimizing response to methadone maintenance treatment: use of higher-dose methadone. J Psychoactive Drugs. 1999 Apr-Jun;31(2):95-102.

Stimmel B, Hanbury R, Sturiano V, Korts D, Jackson G, Cohen M. Alcoholism as a risk factor in methadone maintenance. A randomized controlled trial. Am J Med 1982 Nov;73(5):631-6

Tennant FS Jr. Inadequate plasma concentrations in some high-dose methadone maintenance patients. Am J Psychiatry. 1987 Oct;144(10):1349-50.

Hillebrand J, Marsden J, Finch E, Strang J. Excessive alcohol consumption and drinking expectations among clients in methadone maintenance. J Subst Abuse Treat. (2001) 21(3):155-60

National Addiction Centre Institute of Psychiatry, Maudsley Hospital 4, Windsor Walk, London SE5 8AF, UK.

1 October 2003

Pharmacokinetics of high-dose buprenorphine

Drug Alc Depend (2003) 72; 1:75-83

Pharmacokinetics of high-dose buprenorphine following single administration of sublingual tablet formulations in opioid naïve healthy male volunteers under a naltrexone block. McAleer SD, Mills RJ, Polack T, Hussain T, Rolan PE, Gibbs AD, Mullins FGP, Hussein Z.

Dear Colleagues,

At last we have some real data on buprenorphine half lives, absorption and effects from healthy volunteer studies. These researchers, who were working for the manufacturers, have taken 35 healthy males and given substantial doses of buprenorphine after 50-150mg naltrexone 'block'. They then measured clinical and blood parameters at regular intervals for up to 3 days in an in-patient setting. Techniques for measuring blood levels of buprenorphine are still being developed and are not generally available in clinical practice. This makes patient history and clinical observation even more important than otherwise. A liquid chromatographic tandem mass spectrometric (LC-MS/MS) assay was developed by these researchers and it was validated for the measurement of buprenorphine and nor-buprenorphine, its metabolite, in blood.

Mean half life was found to be 26 hours with a wide range from 9 to 69. Interestingly, these authors have confirmed some observations in clinical practice including highly variable half lives and the 'bi-exponential' decay sometimes reported by patients. Some had a 'secondary peak' at about 10 hours from dosing. The authors report that some had higher levels following meals and propose 'entero-hepatic recirculation' in some cases. Maximum or 'peak' levels occurred between 30 minutes and 3 hours.

The addition of naloxone in the sublingual preparation made no difference to blood levels. The time taken for the tablets to dissolve were similar for all dose levels from 2 to 16mg (range 6-12 min) and were no different for the combination product. It is worrying that the American FDA has approved the marketing of the combination product even before research of this nature had been established. One wonders if they have different standards for drugs used in the treatment of addiction.

The authors confusingly claim to have shown that the two formulations of buprenorphine - naloxone were 'bioequivalent'. This should not be read as their being bioequivalent to the pure product which for some reason they did not test or else did not report if they did. Also, and importantly, these patients had already been given a large dose of naltrexone, a close chemical cousin of naloxone. Unsurprisingly, they found approximately half the blood levels when half the buprenorphine dose was administered. To demonstrate bioequivelence between the old pure sublingual product and a new formulation will take specific research which is yet to be done to my knowledge. The manufacturers have sponsored numerous experts who have stated that for practical clinical purposes the pure product is bio-equivalent yet this is not based on any research to my knowledge and is also very unlikely based on the small amount of research there is out there.

Thus a major weakness of the present study is that it did not examine levels in on-going treatment, but just individual single doses. Thus is it more relevant to the initiation period, which is still a major problem for some patients and is thus a helpful contribution to the scientific literature. There were no opioid effects using the naltrexone block at between 50 and 150mg doses 4 hours prior to opioid administration. This is also helpful information for those using blocking therapies, despite their limited place in normal clinical practice.

comments by Andrew Byrne ..