The
Medicines Act 1968 set up a licensing authority that grants a marketing
authorisation (product licence) for a medicinal product only if it is effective
and safe and of good quality. Once licensed, a drug can usually be prescribed by
any doctor under the NHS. But general use of a newly licensed drug may be
undesirable. Firstly, the licensing process cannot define uncommon adverse
effects. It is easier to measure common therapeutic benefits than rare, but
important, reactions. The numerical problem is daunting. If n patients
have been treated, and none has suffered a particular adverse effect, then we can
be 95% sure that the true incidence of that adverse effect is between O/n
and 3/n(1). Licensing decisions are based on trials involving on average
around 1500 patients,(2) so at the time of licensing, a serious reaction that
affects as many as 1 in 500 patients could be undetected, and undetectable.
Britain's Committee on Safety of Medicines asks for "yellow card" reports of any
reactions to newly licensed medicines, marked with an inverted black triangle. A
post marketing surveillance scheme, which monitors prescriptions and adverse
events, exists only in general practice. Both schemes rely on the good will of
prescribers rather than systematic study, and only a fraction of all important
reactions is notified.
Secondly, relative efficacy plays no part in licensing
decisions, though the licensing authority presumably considers relative safety.
Most early studies of new medicines are performed against placebo rather than an
established active agent. This makes it difficult to be sure of a new drug's true
utility.
Thirdly, prescribers are not constrained to use drugs rationally and
cost effectively. Rational prescribing should consider both the benefit and, in
its broadest sense, the cost of a treatment. The cost of profligate use of new
antibacterial agents, for example, is not simply the money wastefully spent but
also the cost of increasing bacterial resistance.(3) Local mechanisms, such as
practice and hospital formularies, and drugs and therapeutics committees, can
have some effect, but the contributors to different formularies are likely to
differ in expertise and in freedom from external influences, and they will not be
privy to the information on which licensing decisions are based.
Licensing
decisions are now made both by the UK Licensing Authority through the Medicines
Control Agency and by the European Medicines Evaluation Agency, which runs in
parallel with national agencies. The licensing authority is not a drugs
wholesaler. It would anyway be unreasonable to hold a Dutch auction, in which a
pharmaceutical company reduced the price of its product until the licensing
authority would take it. But the NHS is interested in money, and it should see
that newly licensed medicines be prescribed at NHS expense only if there are
proportionate benefits.
Interferon beta-1b, which has recently been licensed
through the European Medicines Evaluation Agency, is an example (see pp 1159,
1195). It probably reduces hospital admission by one day every three years on
average in selected patients with multiple sclerosis but has no demonstrable
effect on disability.(4) The published data concern just 124 patients receiving
the 8 million unit dose of the drug. Treatment for one patient costs around £10
000 per year. The NHS might reasonably ask for evidence that the money should be
spent on the drug, rather than on other services for patients with multiple
sclerosis, or other patients. Several new agents for multiple sclerosis, such as
copolymer 1, will pose similar problems.
The newer antiepileptic drugs
provide another example where licensed drugs might have been treated more
circumspectly by the NHS. Two papers (pp 1169, 1184) and an editorial (p 1158) in
this week's BMJ cast considerable doubt on their long term value.(5-7) As Marson
et al say, "the gold standard to determine future use of new drugs will be
actively controlled studies"(5): trials of relative efficacy by another name.
One answer would be to introduce a form of probation for newly licensed
medicines, in which they are subject to careful scrutiny before they become
available for all doctors to prescribe. A possible way to do this in Britain
would be to allow the licensing authority to operate as before, but to decide
separately whether a drug should be available for prescription within the NHS.
That decision would necessarily require the manufacturer to show, on the basis of
randomised clinical trials, that a drug was at least as effective as standard
treatment. It would also permit the NHS to conduct its own trials into the costs
and benefits of a newly licensed treatment, within the framework of NHS
research.(8) This would need some integration, since cost effectiveness is a
matter for the health service, and safety a matter for the licensing authority,
but that should be possible. Prescribing outside the trials would be prohibited
or discouraged by a ban on general prescription within the NHS. The
Pharmaceutical Benefits Advisory Committee, which advises the Australian Minister
of Health about which drugs should be available under the national pharmaceutical
benefits scheme, is a useful model.(9)
The NHS might also wish to see
evidence of cost effectiveness before agreeing to support major changes in the
use of established licensed drugs. For example, using lipid lowering agents such
as pravastatin in the primary prevention of coronary heart disease(10) could be
examined.(11)
The Committee on Safety of Medicines has a good record of
protecting the public from frankly dangerous medicines, while allowing
potentially useful drugs to be marketed. More careful monitoring after a drug has
been marketed would make it easier to detect "rogue" drugs like benoxaprofen
(Opren). Local controls on prescribing have been less successful in ensuring that
drugs are used rationally. The NHS should not be obliged to pay for new drugs
unless they are at least as good as older ones, nor for expensive drugs whose
benefits are uncertain. A good starting point would be a trial of the costs and
benefits of interferon beta-1b before patients are exposed haphazardly to unknown
risks, and before large sums of money are spent for poorly quantified benefits.
R E FERNER Consultant physician
West Midlands Centre for Adverse
Drug Reaction Reporting,
City Hospital,
Birmingham B18 7QH
References
1 Eypasch F, Lefering R, Kum C K, Troidl H. Probability of adverse events that
have not yet occurred: a statistical reminder. BMJ 1995;311:619-20.
2 Rawlins M D. Pharmacovigilance: paradise lost, regained, or postponed? The
William Withering Lecture 1994. J R Coll Physicians London
1995;29: 41-9.
3 Davey, P G, Bax R P, Newey J, Reeves D, Rutherford D, Slack R, et al Growth in
the use of antibiotics in England and Scotland in 1980-93. BMJ
1996;312:613.
4 The IFNB multiple sclerosis study group.
Interferon-beta-1b is effective in relapsing-remitting multiple sclerosis. 1.
Neurology 1993;43:655-61.
5 Marson A G, Kadir Z A, Chadwick D W. New antiepileptic
drugs: a systematic review of their efficacy and tolerability. BMJ
1996 313:1160-740.
6 Walker M C, Li L M, Sander J W A S. Long term use of lamotrigine and vigibatrin
in severe refractory epilepsy: audit of outcome. BMJ
1996;313:1184-5.
7 Mingot G. Drug trials in epilepsy. BMJ
1996;353:1158.
8 Pecknam M, Smith R. Health care: what's science got to do with it? BMJ
1995;310:208.
9 Anonymous. Guidelines for the pharmaceutical industry on preparation of
submissions involving economic analyses. Canberra: Commonwealth Department of
Human Services and Health, 1995 .
10 Shepherd J, Cobbe S M, Ford I, Isles C G, Lorimer A R, MacFarlane P W, et
al. Prevention of coronary heart disease with pravastatin in men
with hypercholesterolemia. N Engl J Med 1995;333:1301-7.
11 Phoroah P D P, Hollingworth W. Cost effectiveness of
lowering cholesterol in patients with and without pre-existing heart
disease. BMJ 1996;312:1443-8.
