The announcement that Herceptin (the brand name of trastuzumab) has been approved for early stage breast cancer has also been accompanied by fears that it will cost the NHS about £100 million, on the assumption that it will be routinely prescribed in cases where women have early stage breast cancer. On a recent court case, a women took her health authority to court for refusing to prescribe the drug prior to its approval by NICE, the National Institute for Clinical Excellence. She was convinced that not receiving the drug was tantamount to a death sentence, and she made several emotive appeals to the press. The Secretary of State for Health intervene to direct that she should receive the drug.
This has been seen as an argument about rationing, finance and costs. I’m not convinced that it is about rationing at all. The first issue it raises concerns the process of approval. In some countries, doctors are able to prescribe drugs fairly freely. In France, prescriptions or licensed drugs are permitted unless the drug has been entered on a list of “références medicales opposables”. In the UK, the opposite is true; doctors are unable to prescribe drugs on the NHS until they have been approved for use. The rationale for limiting prescriptions reflects fears that the information available to doctors is partial, and unduly influenced by the pharmaceutical companies.
In the case of Herceptin, there have been some disturbingly misleading reports – including one in the New England Journal of Medicine, which really ought to know better. The article examining the use of the drug (1) suggested that using it in the early stages had reduced the recurrence of breast cancer by “approximately 50%”, and the journal editorial took that on trust. The numbers in the article are not clearly stated, and they seem to be different in different tables, but nearly 1700 women received trastuzumab for a year, and a roughly equal number did not. 127 women receiving the drug had a recurrence of their cancer, and 220 in the other group had recurring cancer – an improvement, on the face of the matter, for 42% of the treatment group, not “approximately 50%”. Crudely put, 93 people, or less than one person in 18, seemed to benefit. What also needs to be mentioned is that 84 patients receiving the medication were taken off it or withdrew because of ill effects, and that 29 people suffered symptomatic congestive heart failure. What we seem to have, then, is a drug which is potentially beneficial for a few people, potentially harmful for a few others, and makes little difference to most. This kind of profile is not particularly exceptional.
The key problem for the NHS rests in the finding that some people are significantly worse off as a result of receiving the drug. When the NHS approves a drug, it doesn’t just go to one person; it goes to hundreds, and sometimes thousands. What risk is acceptable to improve the circumstances of some people, at the cost of danger to others? This is not a simple question of mathematics, and there is no numerical answer. The moral responsibility of the NHS is to do as much as it can to ensure that the benefits go to the people who need it, and that the dangers for others are minimized. There are more people who benefit that who suffer, which is encouraging, but not good enough. The normal procedure would be to use the results of successive tests gradually to refine the definition of the potential recipient group, so that it is used most appropriately for people who stand to benefit, and avoided for those who are most at risk. That is what the procedures for testing and trial are supposed to do; and that is what the political intervention has stopped. The first duty of any medical service lies in he principle primum non nocere, “first do no harm”. This, not the money, is what is at risk.