THE HEROIN PRESCRIBING DEBATE: INTEGRATING SCIENCE AND POLITICS

Heroin is abused in almost all countries. It is estimated that about 8
million people (0.14% of the world's population) use heroin each year
(1). Of the illegal drugs, it is associated with the highest mortality
and most emergency room episodes, and so is arguably the most
problematic from a health perspective (1). Along with prevention and
law enforcement strategies, treatment is an essential tool for
reducing illicit heroin use and its resulting problems.

The ultimate goal of treatment is to help those affected overcome
dependence and be fully reintegrated into society. Although dependent
users can proceed directly to detoxification and then strategies to
prevent relapse, this fails for a large proportion. Nonetheless, many
of these individuals may achieve new stability by daily oral
administration of methadone, which acts at the same cell surface
receptor as heroin. Methadone maintenance greatly reduces illicit drug
use and other criminal activities and improves health and social
behavior (2). It is the most widely practiced treatment for heroin
dependence in the developed world (1). But some fail to benefit from
methadone and other treatments, so that alternatives are needed for
these resistant individuals. Consequently, heroin itself has been
reconsidered as a treatment option.

The best information about the prescription of heroin as a maintenance
drug comes from decades of experience in the United Kingdom and from
recent Swiss cohort studies. Clinical trials have also commenced in
the Netherlands (3), and there is growing debate about initiating
trials in other countries. This debate should be informed by
scientific evidence and should address political, social, clinical,
and scientific concerns.

What Does Existing Evidence Tell Us?

In the United Kingdom, heroin prescription has existed within an
overall policy of prohibition for decades. The 1926 Rolleston
Committee (4) established the right of medical practitioners to
prescribe regular supplies of an opioid drug, including heroin, if
this would allow patients to lead "a useful and normal life" that
could not otherwise be achieved.

Heroin dependence in youths only became significant in the 1960s, and
thus heroin began to be prescribed more widely and in higher doses,
especially in the London area. Overprescribing by a small number of
practitioners also created a substantial black market (5). This was
rectified in 1968 partly through restriction of the right to prescribe
heroin to specially licensed doctors and the establishment of clinics
(6). There are, however, widely varying interpretations of the
effectiveness of heroin prescription in the UK. Although a randomized
controlled trial in which oral methadone was compared with injectable
heroin showed that neither was clearly superior (7), consensus among
clinicians led to a shift away from the prescribing of injectable
heroin to injectable and oral methadone in the 1970s.

Since then, heroin has only been prescribed for those with long
histories of dependence for whom other treatments have not been
effective. Currently around 300 people (between 1 and 2% of those
receiving a prescription for the treatment of opioid dependence)
receive pharmaceutical heroin (8), usually provided as take-home
supplies. Of the 109 doctors with licenses to prescribe heroin, fewer
than 50 currently use them and about 20 of these account for the bulk
of prescriptions (9). In a survey of physicians who treat people
dependent on heroin, half of the 105 respondents thought heroin
prescription was justified sometimes or often (10). A recent study
that gave 58 long-term dependent users the choice of injecting heroin
or methadone showed improvements in health and social functioning and
reductions in criminal behavior (11).

In Switzerland, a cohort study undertaken from 1994 to 1996 showed
that a system of supervised heroin administration at clinics with
restricted operating hours was feasible as well as politically and
socially acceptable. This result was documented in 17 clinics where a
total of 1035 individuals with chronic and treatment-refractory heroin
dependence were accepted for maintenance with supervised injectable
pharmaceutical heroin (often in combination with take-home oral
methadone) in the framework of a comprehensive assessment and care
program. Concerns about doses escalating out of control proved to be
unfounded, and most participants achieved stable doses in 2 to 4
months. Randomized studies showed that injectable heroin was superior
to both injectable morphine and injectable methadone in attracting the
target group, preventing premature treatment dropout, and reducing
illegal drug use (12).

Participants in this study showed substantial improvements in health
and well-being and very pronounced reductions in crime (12). Crime
reduction was verified by examination of police records and the
central criminal register. Similar results were found in a randomized
controlled study in Geneva (13). The results for the cohort study
participants were compared with those for 121 newly admitted patients
on methadone maintenance who received comparable psychosocial support.
Those in the cohort study showed significant reductions in illegal
heroin, cocaine, and nonprescribed benzodiazepine use, whereas the
methadone patients showed smaller reductions in illegal heroin use.
Both groups showed similar improvements in social integration. Thus,
heroin prescription can be helpful for those on methadone maintenance
treatment who continue to use illegal heroin regularly, as well as for
those who have dropped out of existing treatments (14).

Requirements for New Trials

If the pharmaceutical in question were not heroin and the disease were
not heroin dependence, the next step--a double-blind, double-dummy
Phase III clinical trial in which the new treatment would be compared
to the current gold standard (in this case methadone)--would be
straightforward. But pharmaceutical heroin is not simply a replacement
for methadone. The British experience and Swiss studies have shown
that to achieve stabilization lasting 24 hours, injectable heroin
(which is short-acting) is most often combined with a low dose of oral
methadone (which is long-acting). In addition, doses of both drugs are
tailored to individual requirements. Thus, fixed-dose comparisons
between heroin and methadone make little sense. In addition, unlike a
standard clinical trial, participants would be experienced users of
both heroin and methadone, so that double-blind, double-dummy studies
would not work. For trials targeting those who have failed methadone
treatment, the possibility of random assignment to methadone treatment
may limit those prepared to participate. A variety of trial designs,
each addressing a different issue about heroin prescription, is
therefore warranted. No one trial will provide a definitive answer,
but interlocking trials will allow decisions about the long-term value
of heroin prescription to be reached.

Defining the target population also raises challenges. There are three
target populations: those who have not been helped by existing
treatments, those currently in treatment who continue to use
substantial amounts of illegal heroin, and those who refuse to try
currently available treatment options. However, these definitions are
problematic. It is common for people to have several treatment
attempts before they are successful; continued use of illegal heroin
during treatment may be a reflection of too short a time in treatment
or of the treatment's inadequacy. Also, most dependent heroin users
are reluctant to enter treatment until compelled to do so by social,
legal, or economic crises. It can be argued that something other than
heroin prescription might be effective. A resolution would be to
include a third "arm," testing an alternative therapy, in each
clinical trial.

Risks

An Australian feasibility study (15) identified strategies to deal
with individual and social risks associated with trials of
pharmaceutical heroin. Because there is so little empirical evidence
about heroin prescription and because the issue is highly politicized,
it is difficult to estimate the likely magnitude of the potential
risks. There are five risks of overriding concern:

1. Heroin prescription might be linked with more permissive attitudes
to illegal drug use, encouraging use especially among young people.
This was the reason given by the Australian government for blocking a
proposal for a clinical trial of heroin prescription in 1997. However,
heroin prescription and permissiveness are not inevitably linked, as
the British experience and Swiss studies show.

2. There might be an influx of dependent users to the trial city. This
"honey pot" effect can be minimized by enforcement of strict residency
criteria, by limiting the number of trial participants, and by close
cooperation with the local police (16).

3. Heroin prescription may reduce the proportion of participants who
become abstinent. There has been little research into the achievement
of abstinence because of the long-term nature and expense of the
necessary investigations, hence the available figures are limited and
dated (17). Critics of the Swiss cohort study argue that it has failed
because only around 8% of participants moved into drug-free treatments
within 18 months (12). Yet, over this short time frame, the Swiss
results are consistent with existing evidence for chronically
dependent people (17) and may instead show that heroin prescription
does not reduce the rate of achievement of abstinence.

4. The introduction of heroin prescription may undermine the
attractiveness and effectiveness of other treatments. There is little
evidence on which to assess this risk.

5. Heroin treatment may be unaffordable, especially as ever-increasing
health costs are a concern of many governments. Results from the Swiss
cohort study, however, indicate significant overall savings (SF45 net
per person per day) (12).

Conclusions

Assessment of the effectiveness of heroin prescription for the
treatment of heroin dependence requires that standard clinical trials
be set up. However, the nature of the condition, problems with
consent, difficulties in running a double-blind trial, and more than
one outcome measure are major problems, although they are not
insurmountable.

Is the testing of heroin prescription worth the effort? Research
trials will be deemed unnecessary and inappropriate by various
parties--by some dependent heroin users and their advocates who
believe the benefits are self-evident, and by some who find such an
approach offensive and incompatible with the principles of medical
practice. But the debates about heroin prescription and the potential
hope it offers the chronically dependent cannot be resolved without
high-quality empirical evidence.

Where will it all lead? Heroin will not replace oral methadone as the
treatment of first choice for stabilization. Its short-acting nature
and expense (including the necessary social safeguards) preclude its
widespread introduction. The clinical trials are important to
determine whether heroin has a role as an adjunct to methadone
maintenance--to improve treatment success for those who have failed
existing treatments. Finally, the prescribing of heroin is about
medicalization, not legalization. The 1961 United Nations Single
Convention on Narcotic Drugs places effective constraints on
pharmaceutical heroin availability. These "limit exclusively to
medical and scientific purposes the production, manufacture, export,
import, distribution of, trade in, use and possession of drugs" (18).
Heroin prescription does not challenge the fundamentals of
prohibition. Indeed, the debate about heroin prescription should
promote continuing assessment of the scientific evidence underpinning
current treatment, law enforcement, and prevention policies, as well
as stimulating well-designed empirical investigations to find more
effective strategies.

References

1.United Nations International Drug Control Programme, World Drug
Report (Oxford Univ. Press, Oxford, 1997).
2.J. Ward, R. P. Mattick, W. Hall, Eds., Methadone Maintenance
Treatment and Other Opioid Replacement Therapies (Harwood,
Amsterdam, 1998).
3.Investigating the Medical Prescription of Heroin. A Randomized
Trial to Evaluate the Effectiveness of Medically Co-Prescribed
Heroin and Oral Methadone, Compared to Oral Methadone Alone in
Chronic, Treatment-Refractory Heroin Addicts (Central Committee
on the Treatment of Heroin Addicts, Utrecht, 1997).
4.Rolleston Report of the Departmental Committee on Morphine and
Heroin Addiction (Ministry of Health, London, 1926).
5.H. B. Spear, in Heroin Addiction and Drug Policy: The British
System, J. Strang and M. Gossop, Eds. (Oxford Univ. Press,
Oxford, 1994), pp. 3-28.
6.Interdepartmental Committee on Drug Addiction (Brain Committee),
Drug Addiction, Second Report (Her Majesty's Stationery Office,
London, 1965).
7.R. L. Hartnoll et al., Arch. Gen. Psychiatry 37, 877 (1980).
8.J. Strang and J. Sheridan, Drug Alcohol Rev. 16, 7 (1997).
9.P. M. Fleming, in Feasibility Research into the Controlled
Availability of Opioids, G. Bammer, Ed. (Stage 2 Working Paper
No. 14, National Centre for Epidemiology and Population Health,
Australian National University and Australian Institute of
Criminology, Canberra, 1997), pp. 1-8.
10.L. Sell, M. Farrell, P. Robson, Drug Alcohol Rev. 16, 221 (1997).
11.N. Metrebian, W. Shanahan, B. Wells, G. Stimson,
Med. J. Aust. 168, 596 (1998).
12.A. Uchtenhagen, F. Gutzwiller, A. Dobler-Mikola, T. Steffen,
Eur. Addict. Res. 3, 160 (1997); ____, M. Rihs-Middel, Eds.,
Prescription of Narcotics to Heroin Addicts. Main Results of the
Swiss National Cohort Study (Karger, Basel, in press).
13.T. V. Perneger, F. Giner, M. del Rio, A. Mino, Br. Med. J. 317,
13 (1998).
14.A. Dobler-Mikola, S. Pfifer, V. Muler, A. Uchtenhagen, Heroin and
Methadone Maintenance: A Comparative Study (Karger, Basel, in
press).
15.G. Bammer, Report and Recommendations of Stage 2 Feasibility
Research into the Controlled Availability of Opioids (National
Centre for Epidemiology and Population Health, Australian
National University and Australian Institute of Criminology,
Canberra, 1995), report and working papers available at
http://nceph.anu.edu.au/pub/opioids/opioids.htm; ------ and
R. M. Douglas, Med. J. Aust. 16, 690 (1996); G. Bammer, CNS Drugs
11, 253 (1999).
16.G. Bammer, D. Tunnicliff, J. Chadwick-Masters, Feasibility
Research into the Controlled Availability of Opioids (Stage 2
Working Paper No. 9, National Centre for Epidemiology and
Population Health, Australian National University and Australian
Institute of Criminology, Canberra, 1994).
17.A. Thorley, in Drug Problems in Britain: A Review of Ten Years,
G. Edwards and C. Rush, Eds. (Academic Press, London, 1981),
pp. 117-169.
18.Single Convention on Narcotic Drugs, 1961, as Amended by the 1972
Protocol Amending the Single Convention on Narcotic Drugs, 1961
(United Nations, New York, 1977), p. 19.

G. Bammer is with the National Centre for Epidemiology and Population
Health, Australian National University, Canberra ACT 0200, Australia.
E-mail: Gabriele.Bammer@anu.edu.au. A. Dobler-Mikola is with the
Addiction Research Institute, Konradstrasse 32, CH-8005 Zurich,
Switzerland. E-mail: isfdomi@isf.unizh.ch. P. M. Fleming is at the
Portsmouth City Drugs and Alcohol Service, 130 Elm Grove, Southsea,
Portsmouth, Hampshire PO5 1LR, UK. E-mail: PhilipFleming@compuserve.com.
J. Strang is at the National Addiction Centre, Institute of
Psychiatry, London SE5 8AF, UK. E-mail: j.strang@iop.kcl.ac.uk. A.
Uchtenhagen is at the Addiction Research Institute, CH-8005 Zurich,
Switzerland. E-mail: uchtenha@isf.unizh.ch.