by Ben Goldacre
A policy of harm reduction tackles public health issues directly by seeking to reduce the personal and social costs of drug use. Abstinence is not regarded as a realistic short-term goal for most dependent users, and the principal ingredients of most programmes are syringe exchanges, educational and advisory services, and treatment and maintenance services (generally with methadone).
There is a hierarchy of achievable objectives, with non-users urged to abstain, and users advised to reduce doses, and to avoid the most potent drugs and riskier means of ingestion. Those who insist on injecting are offered advice on safer technique, and those who persist in sharing needles are even taught how to clean their equipment.
This policy has been vigorously opposed in some parts of the world, especially the USA, where drug-related mortality is almost twice that of the UK. Despite this, it has become the guiding principle behind UK drugs policy, along with the maintenance prescription of methadone.
Methadone
Methadone is an opioid receptor agonist with a half-life of approximately twenty-four hours, far longer than heroin. Drugs with longer half-lives tend to produce less acute withdrawal effects, a phenomenon which is utilised in the choice of anxiolytic drugs in psychiatric practice. Crucially, in comparison with heroin, methadone has a greatly reduced euphoric effect. The hope for methadone, therefore, is that it can contain the opiate cravings, on a once-daily oral dose, without providing so much of a ‘high’.
The aim of methadone maintenance is to stabilise and then to ‘cure’ the opiate user. This breaks down into such objectives as: improving the health of drug users, by providing clean drugs in measured doses under medical supervision; reducing drug-related crime by providing users with free legal opiates, thus reducing their need to steal to fund illicit heroin; improving the social situation of drug users (family relationships, finances, employment, housing and so on); persuading users to reduce their daily dose and ultimately take steps towards abstinence. This is in many ways an updated version of Rolleston’s rationale from 1926.
However, the policy of prescribing methadone may be criticised from many different angles, and to the best of my knowledge these criticisms have never been comprehensively considered in one article. Certainly there is no convenient meta-analysis of methadone programmes. I shall consider each criticism in detail, and later compare the use of methadone to the maintenance prescription of heroin, which still continues on a small scale in the UK, and has recently been reassessed in Switzerland and Australia.
Firstly, it is important to recognise that methadone is not a pleasant drug to take, causing nausea and vomiting, weight gain, profuse sweating, dysphoria and tooth decay. This is no major selling-point to a patient group clearly accustomed to making stringent aesthetic judgements about their drugs, and this, combined with the absence of the ‘buzz’ of heroin, means that the take-up rate amongst addicts is far lower than it ever was for heroin.
Hartnoll et al. (1980) found that only 29 per cent of those offered methadone in one DDU between 1972 and 1975 were still attending twelve months later. The reality of take-up rates for methadone prescription programmes amongst the general population of heroin addicts today is that only a small minority of addicts will attend methadone clinics, certainly less than 15 per cent, although specific statistics are hampered by the unknown quantity of the denominator, that is, the number of people in a population addicted to illicit drugs.
Treatment for drug dependency, to be successful and especially to have an impact at a community level, must have high take-up and retention rates amongst problem drug users who, unlike adults with right iliac fossa pain, may not spontaneously present themselves to healthcare professionals.
In order to be successful, therefore, a drug dependency unit must offer both treatment and the drug at a ‘price’ which the users are willing to pay: the prescriptions may be free, but the terms and conditions on which they are offered may act as a deterrent to some users, and the product offered (counselling services, advice, and possibly substitute drugs) must be appealing. Health economists have couched the problem in their own terminology: ‘For treatment to have a high take-up rate, it must sell … and be seen to sell … a good product at low cost.’
Retention in treatment, firstly, is an area where the philosophy guiding the work of a clinic may have as much of an impact as the nature of the drug it is offering. In a controlled study in Australia, heroin addicts were assessed and randomised to two clinics, one oriented to long-term methadone maintenance, and the other oriented to time-limited treatment, aiming primarily at abstinence from all drugs, including methadone. Both groups were urine-tested for heroin, and use of heroin outside the clinic was higher in the abstinence-oriented clinic.
An observational study in a different country showed that addicts were more likely to discharge themselves earlier from methadone clinics where the clinic staff scored highly on an ‘Abstinence Orientation Scale’, measuring their commitment to abstinence-oriented policies on heroin addiction. Other studies have shown that external compulsion to attend clinics, for example by law courts, is also associated with poor retention.
Conversely, a high re-attendance rate has been demonstrated at ‘user-friendly’ clinics where needle exchange and clean drugs are available, with no uninvited counselling. Experience has taught that regular and enduring contact with treatment services is a necessary precondition for successful treatment of addicts.
Finally, studies of drug users who present to rehabilitation programmes have shown that they are often in a poorer state of health than other heroin addicts in the population (of equally long standing) who have not chosen to present, and this is taken by some commentators to mean that addicts will only present as a last resort. Thus methadone programmes are by no means a universally attractive option to the addict population, and addicts often use their drug of choice to supplement their prescription.
Use of heroin outside the confines of a drug-rehabilitation programme (whilst ostensibly attending it) is, of course, associated with all of the risks of everyday heroin addiction: increased risk of intravenous drug use leading to infection, increased acquisitive crime, poor family relations. More importantly, the chaotic nature of the drug use means that the chances of abstinence after a period of regulated drug use are reduced. Thus use of heroin outside the clinic may be considered one of the definitively poor outcome measures.
However, methadone is also a dangerous drug in its own right: astonishingly, use of methadone has a higher mortality even than the use of illicit heroin, although to what extent is uncertain. For example, in 1992, there were 101 deaths from methadone, and forty from heroin; similarly, from 1982 to 1991 there were 349 methadone deaths and 243 heroin deaths: this is despite the fact that there are far more users of heroin, at every stratum of use, by a factor of at least 3:2, than of methadone.
However, to quantify the mortality requires an accurate denominator (the number of users for each drug), and this, as we have already discussed, can only be achieved indirectly for a covert and underground activity such as drug abuse. Estimates vary widely according to the denominator used, and authors are never so disingenuous as to claim pinpoint accuracy for their figures, but the most recent data to be analysed estimates the risk of methadone-related mortality at around four times that of heroin.
The dangers of methadone have long been recognised. Ghodse et al. (1985) analysed the patient records of notified addicts who died in the UK between 1967 and 1981, and found that among patients using heroin, three quarters of deaths were directly drug related, and ‘most deaths in which a drug was implicated were due to medically prescribed drugs’ (invariably methadone). A retrospective cohort study followed up 128 addicts who first presented in London in 1969, of whom twenty-eight had died, and reported similar findings.
Reasons for this high mortality have been ascribed to its long half-life: a large number of deaths occur in the first few days of treatment, and this may be due to the chronic accumulation of the m
ethadone in the bodies of addicts with reduced liver function. Other reasons proposed include black-market consumption, which is harder to quantify, and the co-administration of heroin and methadone, for which there is less evidence, albeit that death certificates provide notoriously poor data.
Clearly there is a paucity of mortality data in the literature on methadone prescription. In 1994, a review of the methodology of drug treatment evaluation found that only four out of seventeen UK studies had used mortality as an outcome measure. To neglect this most ‘ineffective’ of outcomes, in studies of a drug which is prescribed to 17,000 British addicts, in whom it has a demonstrably higher mortality than the drug it is substituted for, seems extraordinary.
Finally, and perhaps most bizarrely, it is generally recognised that methadone is a more addictive drug than heroin, with a more arduous withdrawal process, and this fact is recognised both among the drug-using subculture and in the scientific literature.
Heroin on Prescription
The current situation is that very little heroin is prescribed in the UK: it was estimated that 117 addicts were prescribed heroin in 1992, while 17,000 were prescribed methadone. Maintenance prescription of heroin, the ‘British System’ until the 1960s, is the ultimate extension of harm reductionist philosophy. There are many deductive arguments to support it, but little modern experimental data, and many criticisms that are laid against it. I shall consider these extensively, before examining the few studies of contemporary heroin maintenance programmes which have recently been published.
The philosophy behind the prescription of heroin owes a lot to the findings of the Rolleston Committee in 1926, is similar to the thinking behind methadone prescription, and is essentially as follows: addiction itself is not something that is readily amenable to medical intervention, and as such opiates are prescribed to the addict for as long as they remain addicted, in order to keep them in a state of good health and leading as normal (and crime-free) a life as possible.
Addiction has been famously characterised by Vaillant (1991) as a chronic relapsing condition with a spontaneous remission rate of 5 per cent per annum regardless of external intervention. This apparently flippant description is supported by empirical data on long-term follow-up of addicts which show that no external agency expedites the ending of addiction, not even major life events.
With drug addiction, we are often choosing between problems, rather than solutions, and so heroin maintenance, which is only ever offered to patients who have failed with other modalities of treatment, could be considered the best of a bad lot. ‘With readily available prescribed opiates, there is no need to commit acquisitive crime to buy drugs, to sell drugs to others to finance one’s own use, and to risk one’s own (and others’) health, not to mention life, with adulterated drugs of unknown strength.’ It is also likely to promote attendance at the clinic for intervention when deemed appropriate, and an important side effect is the denial to criminals of a lucrative source of income.
There are of course a number of criticisms of heroin prescription. The first is that it negates the deterrent effect of the criminal law. However, heroin addicts already resist the deterrent effects of arrest, imprisonment, beatings by gangsters, social isolation, and injury or death through adulterants and disease. It is hard to imagine any greater sanctions than these, and so for addicts of this nature the choice may not be between detoxification or prescribed heroin, but between heroin from the illicit market or heroin from a clinic.
The second criticism is the possibility that heroin prescription would increase drug use in the general population. However, there is good evidence that untreated addicts must indulge in low-level and aggressive marketing of heroin to provide themselves with a supply; that is, they push the drug in order to obtain it, thus promoting increased general consumption. It was partly the cessation of maintenance prescription in the 1960s that led to the arrival of an aggressive black market. Furthermore, it seems likely that the improved contact with family, friends and healthcare providers that comes with maintenance prescription improves the chances of a healthy productive lifestyle and ultimate abstention.
This criticism of increased general use is linked to the fear of leakage of prescribed heroin onto the black market. This problem is best addressed by the careful prescription of an exact dose by specialist prescribers, and the evidence from the one remaining Rolleston clinic in the UK, or rather from the local drugs squad (who undertook to examine all arrested addicts for evidence of drugs prescribed by the clinic), was that there was no leakage onto the black market.
A final criticism of heroin prescription is that it is expensive, costing up to ten times more per year, per addict, than methadone. Firstly it is important to recognise that the cost of any drug is not the sole factor in the running of a nationwide treatment and rehabilitation programme. Public expenditure on drug control was £500 million in 1995, and of this £60 million was spent on treatment and rehabilitation, while £350 million was spent on police and customs enforcement, deterrents and control (which prevents less than 15 per cent of drugs from arriving on the black market).
If we calculate that there were 20,000 recipients of £500 of methadone annually, that is £10 million from the treatment budget, which would be £100 million if heroin was prescribed to a similar number. It seems likely that take-up and retention rates in clinics would increase if heroin was prescribed. Thus it would seem that this is perhaps the most viable of all our criticisms, and local health authorities have criticised heroin prescription on grounds of cost.
However, it has been claimed that because only one company may distribute heroin for medicinal purposes in the UK (Evans Medical), a virtual monopoly has been created, to the point where heroin is overpriced by a factor of thirty. This monopoly position was addressed by the European Court of Justice in 1995, who ruled that the government would have to open up the market to competition, but as yet there are no plans to change the situation.
We must now consider the studies which have sought to compare heroin and methadone. Such data is extremely thin on the ground: there was one small randomised control trial in UCH from 1972–75; and one similar study in Switzerland in 1995; there is also poorly quantified data from the one Rolleston clinic in the UK which closed in 1995, and one small case-control study from Northern Ireland.
Hartnoll et al. (1980) studied ninety-six addicts randomised to oral methadone (OM) or heroin maintenance (HM). After twelve months, 74 per cent of the HM group were still attending, but only 29 per cent of the OM group had maintained contact with the clinic. For both groups, illicit heroin use decreased, in the HM group from 74 to 21 mg/day, but in the OM group from 74 mg to 37 mg/day. There was no difference between the two groups in employment status, but over the course of the year 32 per cent of the OM group had spent some time in prison, whereas 19 per cent of the HM group did so.
Perneger et al. (1998) studied fifty-one patients randomised similarly to heroin or methadone for six months in Geneva in 1995. At follow-up, there was a significant difference in use of street heroin, with 48 per cent of the OM group using street heroin on a daily basis, as opposed to 4 per cent of the HM group (p = 0.002). There was also a significant difference in the amount of money spent on drugs between the two groups, with the OM group spending approximately ten times the amount of the HM group (p = 0.039). Associated with this, the HM group were less likely to be charged with theft (p = 0.015) and less likely to be charged with drug dealing or possession (p = 0.067 and p = 0.008 respectively); overall, 57 per cent of the OM group were charged with any offence over the course of the six-month trial, whereas 19 per cent of the HM group were charged (p = 0.0004).
There was also a significant difference in mental-health status, with six suicide attempts in the OM group, against one in the HM group (p = 0.022). Finally, health-related quality of life was measured with the SF-36 scale, which found a greater improvement for the HM group in mental health (p = 0.025) and social functioning (p = 0.041). There was no d
ifferential improvement in employment status, housing situation, or somatic health between the two groups over six months.
McCusker and Davies (1996) found similar results at a clinic in Northern Ireland over six months. The HM group manifested lower levels of psychopathology and showed greater retention in treatment, criminal activity was significantly more reduced, as was illicit heroin use, and although there were again no differences in physical health, the OM group was the only one to report the sharing of used injecting equipment.
Thus our three trials demonstrated many advantages to the prescription of heroin, and none to methadone.
Conclusions
Clearly there are stout grounds for scepticism concerning the validity of prescribing methadone in the treatment of heroin addiction. It is also clear that there is a paucity of research in this field, a failure which must surely be redressed.
There are certain things of which we can be certain. Heroin is the most attractive drug for heroin addicts, and however we might wish them to behave, they continue to use illicit sources of the drug even if a substitute is prescribed. Methadone is undoubtedly a dangerous drug, and one that retards entry of addicts into the treatment programmes offered; it is also a drug whose effects have not been comprehensively researched. Heroin maintenance may well prove to be the best option we have.
Drug addiction is not a phenomenon that lends itself generously to empirical investigation. Even the outcome indicators are a subject for debate, and a viable study of what many would see as the ultimate index of success, abstention, would require a trial lasting more than a decade.
However, quantifiable indices of health status, social functioning, criminal behaviour, total opiate consumption, needle-sharing and so on are all viable and uncontroversial outcome measures, and should be comprehensively investigated for methadone and heroin. Furthermore, no indications have been found that prescribing heroin would inflict harm of a kind that might make such trials unacceptable.