16 November 2009

Does methadone cause QT problems or is it often viral?

HIV and Long QT syndrome - Cause or coincidence? Puri R, Roberts-Thomson KC, Young GD. International Journal of Cardiology 2009 133;1:e9-e10


Dear Colleagues,


This may be the first article to formally describe the link between HIV and torsade de pointes tachycardia and at the same time to question the role of methadone. Their single case report has much in common with others in the literature: the 36 year old female with HIV was taking long term methadone for dependency and presented with recurrent syncope. The dose was only 70mg daily at the time of the torsade but importantly, had been 190mg daily 18 months previously – at which time ECG showed the QTc to be normal - and there were no cardiac symptoms. The authors proposed that their patient did not have methadone induced QT changes, but HIV-induced long QT (LQT) syndrome. The QTc was 540ms around the time of the torsade.


These authors go on to discuss the effects of HIV on the heart. Up to 60% apparently have positive cardiac findings at autopsy and 30% of cases may have asymptomatic prolonged QTc, largely in the absence of ‘culprit’ medications (citing Kocheril 1997). “Minor repolarisation abnormalities in HIV infecteds may therefore become clinically overt in the setting of concomitant predisposing drug therapies.”


The authors state in their introduction: “Methadone use has been associated with prolongation of the QTc and an increased risk of sudden cardiac death.” In fact after 40 years of widespread use there has been no such association with sudden cardiac deaths. Yet this statement well exemplifies the current popular mythology around the subject. Since Krantz and colleagues wrote the original case series (but not the first case) in 2002, there have been no confirmed or strongly suspected deaths due to torsade tachycardia from my reading. The only 2 or 3 deaths were either remote from the period of the torsade and/or else were due to another reported cause such as myocardial infarction. French reports from over 30 years ago quote a mortality from ‘torsade de pointes’ of around 16%. Since this was before modern mobile resuscitation and pacing technologies were widespread, the survival rate of 84% might have increased to something over 95%.


It is now clear that ‘torsade de pointes’ tachycardia rarely if ever occurs in new entrants to methadone treatment. The 103 reports in the literature and Justo’s excellent summary of the field in the Addiction journal inform us that simple clinical features can highlight risk and indicate the need for ECG monitoring where appropriate. Almost 50% of the ~100 reported cases in the literature had HIV infection. The author of the original FDA report, Ellen Pearson, has postulated that QT prolongation and torsade are ‘threshold events’ with numerous contributors based on the known risk factors.


The risk factors (not in order) are:


(1) long term methadone maintenance for addiction
(2) female sex
(3) age over 40
(4) doses over 150mg daily
(5) HIV infection
(6) concomitant use of drugs which either increase methadone levels and/or prolong QT interval
(7) metabolic disturbance
(8) structural heart disease
(9) alcohol

When used with supervision and adequate supports methadone treatment for heroin addiction reduces mortality substantially. It should be used with confidence as the benefits far outweigh even the most pessimistic views of the possible side effects.


Comments by Andrew Byrne ..


Clinic web page: http://www.redfernclinic.com/c/
Refs:


Kocheril AG, Bokhari SAJ, Batsford WP, et al. Long QTc and torsades de pointes in human immunodeficiency virus disease. Pacing Clin Electrophysiol 1997 20:2810-6


Krantz MJ, Lewkowiez L, Hays H, Woodroffe MA, D. Robertson AD, Mehler PS. Torsade de Pointes Associated with Very-High-Dose Methadone. Ann Intern Med. 2002 137:501-504


Dessertenne PF. La tachycardie ventriculaire a deux foyers opposes variables. Arch Mal Coeur 1966 59:263-72


Pearson EC, Woosley RL. QT prolongation and torsades de pointes among methadone users: reports to the FDA spontaneous reporting system. Pharmcoepidemiol Drug Safety. 2005 14;11:747-753

Methadone safe in cancer patients - minor QT problems went away after 4-8 weeks methadone

The Effect of Oral Methadone on the QTc Interval in Advanced Cancer Patients: A Prospective Pilot Study. Reddy S, Hui D, El Osta B, de la Cruz M, Walker M, Palmer JL, Bruera E. Journal of Palliative Medicine, October 13, 2009 E-pub ahead.

Dear Colleagues,

These authors have done a great service by following serial ECGs prospectively on 100 patients who were being considered for methadone treatment for advanced cancer pain. ECG was ordered at baseline, 2, 4 and 8 weeks. Due to altered medication, hospice transfers, community discharges and one death (non-cardiac) in this palliative patient population the follow up results were available for 64, 41, and 27 patients at 2, 4 and 8 weeks.

Perhaps the most interesting and unexpected findings of this study were that even before starting the medication, over a quarter of patients (28%) had QT prolongation (>430 ms in males; >450 ms in females) and this dropped. The proportion of subjects was *lower* at each of the follow-up periods with only 8-11% of patients having QT prolongation.

At two weeks 11 patients (17%) had QTc>10% above baseline. However, by 4 and 8 weeks this had dropped to one single patient (3%). There was only one ECG in one single patient where QTc increased beyond 500ms (1.6%). This was asymptomatic and not associated with any tachycardia episode. Furthermore, that patient’s prolonged QT resolved spontaneously in subsequent ECG tracings. The authors express their surprise at these unexpected findings which they ascribed partly to the high baseline occurrence of QT prolongation and/or possibly a reduction in other drugs prescribed or improvements in electrolyte disturbances.

The doses of methadone were relatively low compared to the dose levels used for addiction (median at 2 weeks 23mg daily and maximum was 90mg daily).

The authors conclude: “clinically significant QTc prolongation rarely occurred … our preliminary findings are encouraging. … we believe that methadone should be prescribed without reservations … . For patients with significant risk factors … monitoring with ECGs at baseline and at subsequent intervals may be reasonable.”


This should give doctors and patients confidence that methadone is still a safe and effective analgesic and that concerns regarding cardiac side effects may have been exaggerated out of keeping with the literature. When I contacted the study’s author I was told that they had seen no cases of torsade tachycardia in relation to methadone treatment at the MD Anderson Cancer Center to date.

Comments by Andrew Byrne ..

1 November 2009

Safe and effective opioid prescribing in addiction treatment. Article written for UK psychiatry journal.

Commissioned by editor of 'Advances in Psychiatric Treatment', accepted, paid for but never published for reasons quoted as relating to 'legal matters'. 

Safe and effective opioid prescribing in addiction treatment.

Author Dr Andrew Byrne

Abstract:

A large body of research supports the prescribing of maintenance opioids for heroin addiction yet poor quality treatment in the UK has limited the potential benefits. This in turn has caused many to become disillusioned about addiction treatment generally. Inadequate dose levels without the necessary supervision and psychosocial supports have both contributed to this state of affairs in the UK. By failing to address this situation, the National Addiction Centre in London has actually perpetuated it. While some progress has been made in recent years, psychiatry trainees in the UK are ideally placed to help improve the quality of pharmacotherapies in line with other European countries in moving towards an evidence base.


Article:

The principles linking opioid maintenance treatment and behavioural therapies were defined in Dole and Nyswander’s classic paper which is now one of the most quoted in the medical literature [ref 1]. Psychiatrist Marie Nyswander had noted limited success treating heroin addicts in New York using psychoanalytic techniques alone. With Dr Dole, she reported a cohort of ‘hopeless’ New York street addicts responding favourably to a trial of strictly supervised, ‘high-dose’ methadone treatment (mean 100mg/d, range 15-180) with intensive psychosocial supports. They found dramatic reductions in illicit drugs use, excellent retention in treatment along with vocational, social and other demographic improvements. The trial was radical at the time as it placed social functioning as its primary goal, rather than abstinence from all opiates. Patients received daily supervised medication and their drug use was monitored by regular urine tests as part of treatment. Many rigorous studies since have further refined ‘best practice’ and also documented safety data. These were especially important in long-term patients, pregnancy and in those with coexisting mental illness.

Over four succeeding decades, methadone and other maintenance treatments have become just one component of a more complex therapeutic repertoire for addiction including the anti-craving drugs, mood altering medications, detoxification, brief interventions, CBT, formal psychotherapy and other strategies. These are all aimed primarily at reducing the harms from drug addiction while also encouraging engagement in normal social activities. Contrary to popular opinion, the natural history of opiate use, like smoking and alcoholism, in fact moves towards abstinence, with or without treatment [ref 2].

British perspective – historical background of opioid treatment in the UK.

Due to an unwillingness of the dependency establishment to accept methadone as a valid maintenance treatment, a majority of heroin users in treatment in the UK have been subjected to a ‘culture of abstinence’. This is akin to Nancy Reagan’s retort of “just say no to drugs!” Like smokers, drug addicts are generally well aware of the dangers they are taking. Even by 1989 when methadone maintenance was being introduced into many other countries, UK treatment practise was only for short term reduction prescribing. We knew then as now that this leads to relapse in over 90% of cases.

At a time when needle sharing was still common, this caused many otherwise preventable cases of HIV. To this day, many doctors in the UK will only condone short-term, low dose methadone. Others continue to implement a punitive policy of enforced dose reductions when drug use, even non-opiate drug use, is found on urine testing. Some NHS clinics refuse to readmit their own old discharged patients for arbitrary periods, raising further barriers for those most needing assistance.

In the 1980s two forward thinking doctors introduced a more evidence based type of maintenance treatment into Scotland using GP ‘shared care’. Pharmacists were instructed (and paid) to witness the administration of liquid doses and an emphasis on rehabilitation replaced a priority of dose reductions to abstinence [ref 3]. High quality, clinic-based services were also developed in some centres in England (eg. Sheffield, Portsmouth, Manchester). Nevertheless, a large proportion of methadone in the UK was still prescribed ‘on demand’ in general practice using doses which were often inadequate and ineffective, and in settings where there was no dose supervision, little urine testing and no check on compliance. The poor outcomes led predictably to a cycle of negative attitudes towards methadone treatment which persists to the present day [ref 4]. With the unrealistic goal of short term abstinence, it is not surprising that many informed citizens, parents, police and even health workers held little confidence in methadone treatment, despite its glowing record in public health circles when properly implemented.

Uniquely in the UK, methadone and other opioid prescriptions are at least theoretically available to all addicts through the NHS. This would utilise GPs and/or specialist clinics with other health workers giving counselling and psycho-social supports which are known to improve outcomes [ref 5]. Pharmacists or clinic nurses would administer (supervised doses) and dispense (give out medication for later consumption) the medication. Guidelines were finally introduced in 1999 which incorporated what Dole in New York and others around the world had been doing for decades [ref 6]. These advised maintenance treatment and also the use of supervised dosing for new and unstable patients as well as adequate dose schedules. Yet even four years later Strang et al report that most methadone is given without supervision and in doses which are still inadequate for most to curtail injecting behaviour [ref 7].

Some jurisdictions which introduced methadone maintenance propitiously have avoided the HIV epidemic almost completely in their injecting population (Hong Kong, Australia, New Zealand). Unfortunately, for reasons which are still being elucidated, this did not extend to hepatitis C which continues to spread even where new HIV cases had almost ceased.

What is opioid maintenance treatment? What can it do? What can’t it do?

Who needs treatment? When do they need it? The first dose.

Who is best placed to provide such treatments?

So what is needed for the future?


What is opioid maintenance treatment? What can it do? What can’t it do?

Opioid maintenance treatment involves the legal prescribing of a drug of dependence to an addicted patient within a defined therapeutic framework, involving goals, support, supervision and regular review. Short term opioid abstinence is usually considered secondary to other goals such as reduced risk taking behaviour, better general health, work, education and family responsibilities.

Many things change in an addict’s life when starting opioid maintenance treatment. Studies have shown mortality declines from over 2% per years to less than 0.5% [refs 8,9]. Since there is less injecting, viral disease is less likely to be passed on by those who are already infected. As well as less injecting, employment, legal and financial matters have all been shown to improve substantially for those in treatment (Ref 9b). And the longer treatment lasts, the greater these improvements. This is not to say that everyone needs treatment indefinitely and a large proportion do successfully withdraw from maintenance opioids [refs 2, 12].

Only a very small proportion of patients will successfully withdraw from the opioid treatment in the short term and still remain opiate abstinent [ref 9b]. This is probably less than 10% of the total, even though many more express a desire for such an outcome. Hence all opioid dependent patients should have access to continuing prescribed opioids and those who discontinue should be encouraged to seek supports which seem appropriate for the individual.


Who needs treatment? When do they need it? The first dose.

A careful assessment is essential in any patient presenting with drug or alcohol problems. This involves a thorough history, physical examination (pupils, mental state and injecting sites as a minimum) and usually a supervised urine test and blood tests including liver function, hepatitis B/C, HIV status, etc. As for other medical prescribing, the necessary minimum includes both a clear diagnosis and, usually, the failure of non-drug treatments. In practice, the need for opioid maintenance is relatively easy to establish, except in the very young or in those with concurrent medical or psychiatric illness. The diagnostic criteria for opiate dependence involve compulsive self administration, escalating doses, withdrawal effects and usually, documented adverse health and social consequences. It is important to document all of these clearly in the patient record before prescribing any medication. In addition, the patient’s identity and some aspects of their past treatment history needs to be confirmed.

To make our job easier, patients may have ‘self-selected’ by seeking out a doctor or clinic where dependency services are available. Some may not want methadone but seek other medications to assist with detoxification. Such patients should be informed of the benefits of maintenance therapies in case their detox episode is unsuccessful. All patients should be informed about self help groups including AA, NA and the new SMART Recovery movement [ref 10].

Most patients will have a substantial history of heroin or other opioid use, often by injection and with documented complications, end-organ damage, legal, financial and social consequences. Venous scarring is the most obvious sign of long term history. The drug use may take the form of injected heroin, black market methadone, codeine, morphine, opium or even poppy seeds in rare cases. As long as the use is consistent and compulsive with tolerance and withdrawal symptoms/signs the criteria of dependency are fulfilled. It is helpful to use the DSM-IV definition although it must be remembered that this was devised for use by private American psychiatrists and there may be occasional deviations in ‘normal’ countries.

In patients who are very young (under 18 years) or who have unstable mental illness, it is important to ascertain that opiate opioid maintenance therapy is indeed the most suitable option at the time. Some such patients may develop a mistaken notion that they need prescribed opiates opioids. They may also give a credible history of dependence. This always needs to be carefully corroborated with physical examination and urine testing. This is especially so if there are no venipunctures, no history of hepatitis C, overdose, financial, legal or other consequences of opiate dependence. In such sensitive cases it is prudent to seek a written opinion from a colleague to ensure that other forms of treatment may not be more appropriate. This may be the patient’s own GP or consultant who has been involved. In some jurisdictions parental permission may be required at this age. Health authorities or family services may also have to be involved in under-age cases, with details varying between jurisdictions.

As with other major treatment decisions, the patient should fully informed about its nature. This essential information should be given verbally, allowing for questions, as well as in writing. Various documents are available on the internet for patient education. Some documentation of consent should also be obtained in the patient records. Patients need to know that both methadone and buprenorphine have benefits and also certain side effects such as headache, constipation and sweating. The issue of cardiac conduction defects has never been shown to be a problem in patients being treated under dependency guidelines. However, for those taking higher doses (>150mg daily) or with other risk factors a cardiograph is a prudent step [ref 11].

Although many patients do attain opiate abstinence, methadone and buprenorphine treatments are not ‘cures’ for addiction. Patients should be aware that this is a treatment which requires regular attendance for medication, medical reviews, counselling and urine testing. They should also be informed that this treatment often lasts for months and sometimes for years. The myth of methadone being “for life” has been disproved by longitudinal studies with acronyms NTORS, ATOS, etc [ref 12]. Gossop points out that because clinics see successful patients less often than others, staff may develop the incorrect impression that few ever successfully withdraw from treatment.

Patients often arrive in distress and dismay, wanting to get into treatment urgently. It is still essential to ascertain who needs opioids and who may be more appropriate for detoxification services. Just because a patient says that they are in withdrawals does not mean that a doctor must write a prescription for opioids, although this should always be seriously considered as an option. Prescribing always has more predictable outcomes than detoxification. The doctor takes responsibility for the former and the patient the latter. Vincent Dole, the co-inventor of methadone treatment, said that “detoxification is an experiment in the life of the patient”
Who is best to provide such treatments? How is it done?

The delivery of methadone can occur in either the specialised clinic setting or in existing community facilities. There are advantages and disadvantages to both types, but ideally, new and unstable addiction cases would be treated in a specialised clinic. This allows close supervision for a period, after which stable and longer term patients could be referred back to GPs and pharmacists for community treatment. In practice there are usually more patients than services available so any treatment opportunity will have immediate applicants, most of whom are assessed as appropriate for maintenance treatment.

As with other acute presentations, one cannot do everything in the first consultation. However the basics need to be organized and a decision taken promptly as to whether or not the patient is to be prescribed opioids in a treatment ‘program’. At that point, one can afford to put off certain other matters until the patient feels better and has more confidence and familiarity with the staff and treatment setting. Another essential detail at this point is whether the patient has adequate housing considering they may be dispensed bottles of strong medicine. Also one needs to find out if there are children in the house and stress the importance of safe drug storage out of their reach.


The first dose.

The patient should usually be given a starting dose of 30mg with small increases in subsequent days up to the usual effective dose of 60 – 120mg. If given too quickly, drug accumulation can cause fatal toxicity so vigilance is needed in the first two weeks when this can occur. An additional 5 to 10mg every 2 to 3 days is usually a safe increase. On the other hand, if doses are kept too low, some patients will drop out while others may continue to use street drugs and/or alcohol. In some clinical settings it may be possible to give supplementary doses later the same day but only if the prescribing doctor has examined the patient 2 to 4 hours after the first dose. Where supplements are given, the second day’s dose should normally be the sum of the first day’s doses as long as there is no sign of toxicity.

Inductions onto buprenorphine are not as critical since early overdose is not a problem owing to the “ceiling effect” for respiratory depression. Most start with 4 to 8mg as a supervised sublingual dose, increasing only after 3 to 4 days when steady-state levels are achieved in this very long acting drug. Supplements may also be given, but these should be considered ‘loading doses’ and may only be needed in the first few days. The usual effective dose is 6 to 16mg daily with only a small proportion requiring more or less than this level. 32mg is the maximum daily dose.

Patients often know from previous experience how much they need and which drug suits them best. About a third of heroin addicts treated with buprenorphine will continue to feel drug cravings even when doses have been raised to the maximum of 32mg daily [ref 13]. Such patients usually do well on methadone using standard doses. For this and other reasons, methadone is probably still the best first line drug. A smaller proportion of methadone patients report unacceptable side effects such as sedation, sexual dysfunction, constipation or sweating and a transfer to buprenorphine can be very rewarding. However this can only be done ideally when the methadone dose has been reduced below 40mg daily due to the potential for a precipitated withdrawal episode as the partial opioid agonist buprenorphine replaces the full agonist methadone. This can be very unpleasant although it is usually short lived, in most cases less than one hour.

The first month of treatment is crucial to long term success. Hence it is essential to engage with the patient and establish a confident and professional relationship. This will involve all health care workers from reception staff to nursing, medical and pharmacists.

As with other conditions, management involves educating our patients, prescribing medications judiciously and supervising and monitoring progress. As with diabetes, depression or blood pressure, there is wide variation in views about how frequently patients may need to see a doctor, counsellor, pathology service or pharmacist. But the general principle is that new and unstable patients need more frequent and intensive involvement than long-term stable patients. Where there are psychologists, counsellors and other staff medical visits may be less frequent after the first month of treatment. There should be a formal interview each week until the patient shows signs of stability, then 2 to 4 weekly consultations should suffice for a year. Even very long term patients should probably see their prescriber every two months at a minimum.

Urine testing.

All patients who have come to the attention of dependency services should probably have urine testing at some frequency. This is essential at the initial assessment and twice yearly urine toxicology is probably a minimum for any person prescribed take-home doses of opioids, probably including pain management cases. Tests should be ‘supervised’ to some degree. The most useful tests for research or legal purposes will be directly witnessed and done at random. This is not always practical, nor is it necessary in most cases in clinical practice, unless the patient needs to prove their status for legal, family, sporting or sensitive employment matters. It is usually sufficient to ask for a urine test on a particular consultation day and have the staff test the temperature of the specimen. This may be done manually or using adherent temperature sensitive strips.

The interpretation of urine testing involves distinguishing non-specific ‘opiate’ positive tests from ‘morphine’ which is the breakdown product of heroin. One must take into account the half-lives of the cannabinoids, benzodiazepines, cocaine, amphetamine, etcetera. There should be no punitive outcomes from urine tests and these should only be used as a clinical indicator.


Dose supervision.

For new and unstable patients, as with other areas of medical practice, outcomes are directly related to compliance. The treatment of malaria, TB and HIV have each been shown to improve with directly observed treatment [refs 14,15]. Likewise with opioid therapies, witnessed doses improve outcomes and reduce the scope for diversion. In practice, most patients can be successfully treated by attending one to three times weekly depending on time in treatment, stability and dose level. The uniquely British practice of attendance at the pharmacy every one of two days to take bottles of medicine home is not based on any research and should be abandoned.

Cardinal rules for methadone:

The effective dose is generally 60-120mg daily with a small proportion needing more or less that this range due to unusual metabolism or tolerance. No more than 30mg should be given as a starting dose with increases of 5-10mg every 2-3 days, more rapidly only where close medical supervision is possible. Methadone should be avoided with fluvoxamine (inhibits metabolism), phenytoin or carbamazepine (induce metabolism) or pentazocine (may precipitate withdrawal, like buprenorphine). Special precautions are also necessary with various anti-HIV and TB drugs which may increase or decrease blood levels of methadone. Even grapefruit juice, with its effect on the cytochrome P450 enzymes can reduce methadone metabolism and raise levels. The principle is to carefully monitor any patient who is prescribed other drugs and be prepared to raise or lower the dose as appropriate – an examination 3-4 hours post-dose for signs of intoxication, and 24 hours afterwards for signs of withdrawal is generally more useful than measuring methadone blood levels [ref 16]. Patients should be warned not to drive, operate machinery or look after children until they are stable.

So what is needed for the future?

Treating addictions can be enormously rewarding and one does not have to wait for years to see the fruits of interventions. Many of these patients are ‘survivors’ who have enormous energy and resources which they often use to turn their lives around while in treatment. A ‘lapse’ back to drug use does not imply failure, but may mean that more attention needs to be paid to treatment. In cases of ‘relapse’, a second attempt at treatment is more likely to be successful than the first, especially if depression and anxiety are correctly dealt with.

All psychiatrists should be comfortable with treating dependency problems. There are some parallels between the management of nicotine, alcohol, opiate and stimulant addictions. Each has a behavioural and a chemical component. We should be aware of the differences and the similarities, each requiring appropriate interventions when required. Addictions are still inadequately covered in most undergraduate and family medicine training. Indeed, there was a time when some considered substance dependency not to be an area for doctors, nurses and pharmacists at all!

It is essential that consultant psychiatrists know how to set up and run a dependency unit within a community hospital setting. These will have the ability to take referrals with a view to assessments and a range of treatments, both medicated and non-medicated, based on rational, practical and cost-effective principles. As with general psychiatry, the great majority of such cases can be handled as out-patients but a small sub-set will need hospital admission. As with alcoholism, needs may vary from just brief respite care to acute care and intensive treatment. All the same principles of good medical practice should apply just as in every other medical specialty. While in treatment special attention needs to be paid to other areas of risk such as hepatitis C and other communicable diseases [ref 17].

Opioid maintenance treatment should be considered for all those who are addicted to either street heroin or pharmaceutical opioids and who are unable or unwilling to cease using such drugs. The same could be said for nicotine or, indeed, many medical situations where prescribing is only appropriate when non-drug approaches have failed or are inappropriate (eg. diabetes, hypertension, hyperlipidaemia).

For reasons which would be unacceptable in other fields, deficiencies in dependency treatments in the UK have undoubtedly contributed to the epidemics of HIV, hepatitis C, overdose and other consequences of addiction. It may take many years to turn these deficiencies around. Conceding them would be a great starting point. Methadone treatment has long been treated with great suspicion by the addiction ‘establishment’ in the UK. Indeed, Professor John Strang of the Maudsley Hospital has revealed his own misgivings about methadone by claiming that, despite its known benefits, it may have a ‘bitter final pathological twist’ (ref 18). Such personal reservations stand in stark contrast to 40 years of positive research findings, much published in the high rating journal, Addiction, of which he is an assistant editor.


References:
1. Dole VP, Nyswander ME. A medical treatment for diacetylmorphine (heroin) addiction. JAMA 1965 193:646-50
2. Thorley A. Longitudinal Studies of Drug Dependence. In: Drug Problems in Britain: A review of ten years. Eds: Edwards G, Busch C. 1981, Academic Press. p162
3. Greenwood J. Six years' experience of sharing the care of Edinburgh's drug users. Psychiatric Bulletin 1996 20:8-11
4. Stevenson RJ. Drug misusers are likely to abuse the system. BMJ 2007 335:317
5. McLellan AT, Arndt IO, Metzger DS, Woody GE, O'Brien CP. The Effects of Psychosocial Services in Substance Abuse Treatment. JAMA 1993 269:1953-1959.
6. Drug Misuse and Dependence - Guidelines on Clinical Management. 1999 HMSO Department of Health. Working Group Chair: Strang J.
7. The prescribing of methadone and other opioids to addicts: national survey of GPs in England and Wales. Strang J, Sheridan J, Hunt C, Kerr B, Gerada C, Pringle M. Brit J General Practice 2005 55;515:444-451
8. Caplehorn JRM, Dalton MSYN, Cluff MC, Petrenas A. Retention in methadone maintenance and heroin addicts' risk of death. Addiction 1994 89:203-7
9. Grönbladh L, Öhlund LS, Gunne LM. Mortality in heroin addiction: impact of methadone treatment. Acta Psychiatr Scand 1990 82:223-227
9b Gossop M, Marsden J, Stewart D, Treacy S. Outcomes after methadone maintenance and methadone reduction treatments: two-year follow-up results from the National Treatment Outcome Research Study. Drug and Alcohol Dep 2001 62;3:255-264
10. Self management and Recovery Training. http://www.smartrecovery.co.uk/ (accessed on 13/2/08).
11. Byrne A, Stimmel B. Methadone and QTc prolongation. Lancet 2007 369:366
12. Gossop M. The clinical fallacy and treatment outcomes. Addiction 2007 103:89
13. Kakko J, Grönbladh L, Svanborg KD, von Wachenfeldt J, Rück C, Rawlings B, Nilsson L-H, Heilig M. A Stepped Care Strategy Using Buprenorphine and Methadone Versus Conventional Methadone Maintenance in Heroin Dependence: A Randomized Controlled Trial. Am J Psychiatry 2007 164;5:797-803
14. Babudieri S, Aceti A, D'Offizi GP, Carbonara S, Starnini G. Directly Observed Therapy to Treat HIV Infection in Prisoners. JAMA 2000 284;2:179-180
15. Garner P, Volmink J. Directly observed treatment for tuberculosis. BMJ 2003 327:823-824
16. Hallinan R, Ray J, Byrne A, Agho K, Attia J. Therapeutic thresholds in methadone maintenance treatment: A receiver operating characteristic analysis. Drug Alc Dep 2006 81:129-136
17. Hallinan R, Byrne A, Dore G. Harm reduction, hepatitis C and opioid pharmacotherapy: an opportunity for integrated HCV-specific harm reduction. Drug Alc Rev 2007 26:437-443
18. Strang J. Looking beyond death: paying attention to other important consequences of heroin overdose. Addiction 2002 97:927-928

Declaration of interest:
Dr Byrne’s addiction clinic charges a fee for dispensing methadone and buprenorphine.