Is it ethical to ask a patient to enter a clinical trial and then ignore the evidence from such trials? In ovarian cancer this has been the situation in the UK since April 1998.

By establishing the National Institute for Clinical Excellence, the government is showing a commitment to evidence-based medicine. But at the moment, the pursuit of this goal is competing with cost-containment, forcing oncologists to ignore the evidence and be economical with the truth in their dealings with patients.

In the 1980s the prognosis for the vast majority of solid tumours stood as it had for decades. But the 1990s have seen a very exciting increase in the range and number of new anti-cancer drugs, many showing both survival benefits and improvements in quality of life.

The new generation of anti-cancer agents are more expensive than the last. It was ever so. But even for first line treatment of ovarian cancer - where the combination of a platinum analogue and paclitaxel (TaxolTM) has been shown to have a statistically significant survival benefit in two randomised controlled trials - there is still no agreement about the funding of this drug.

Yet in health economic terms, paclitaxel for ovarian cancer comes out at£6,500-£7,500 per quality adjusted life year - good value by any criteria.

In February 1998, a group of UK oncologists with a particular interest in ovarian cancer drew up a consensus statement. Subsequently, the Joint Council for Clinical Oncology - representing clinical and medical oncologists - made the unprecedented move of publishing guidelines for ovarian cancer which included the recommendation that paclitaxel-platinum was the optimal firstline treatment.

The first randomised controlled trial, the GOG 111 trial in the US, closed in April 1998, showing an overall survival benefit of 13 months.

A month later a second, confirmatory trial - the intergroup study (OV10) - was presented at the annual American Society of Clinical Oncology meeting, showing an overall survival benefit of 10 months.

The ultimate scientific test - the randomised control trial - had shown significant survival benefits, not once but twice and with comparable results. In the summer of 1998, Cancer BACUP surveyed health authorities and found only 20 per cent were funding paclitaxel.

With the new data many of us, not just clinicians, expected the situation to change. But nearly a year later, postcode prescribing still prevails. Since then the plot has thickened - data from the ICON 3 trial has now been presented. The researchers have described the results as preliminary, yet some are concluding that the trial is negative.

But the data is not yet sufficiently mature to draw any firm conclusions, particularly as in contrast to the previous two randomised trials, one-third of patients had early stage disease (stage 1) or a good prognosis, requiring a longer follow-up.

Findings from the group of patients comparable with the previous two randomised trials - those with advanced stage disease and poor prognosis - are consistent with those of the first two trials.

With the UK's ovarian cancer survivals among the worst in Europe, how do oncologists answer patients who discover from the Internet that every other country in the West now funds this drug as part of first-line treatment?

The UK's spend of£68m on chemotherapy hardly budged last year. This is one-third of the figure per head in France and must be set against£44m on acne,£70m on laxatives and over£600m on anti-ulcerants.

Worst of all, these figures do not reflect evidence-based practice. I'm not unsympathetic to the purchasers' dilemma - they have not been given a directive such as the one for Viagra - nor is there any pot of gold to cover this leap in the cost of chemotherapy.

What is needed is central action. In advance of any guidance from NICE, will somebody please endorse evidence-based practice and enable us to treat our patients with cost-effective therapy that will extend their lives and may even cure some of them?

We should feel shame that middle-aged women - more than 5,000 a year - are being denied an extra year of life so that their teenagers will not be spotty or their elderly dependants constipated.

Hilary Thomas is professor of oncology, Surrey University and the Royal Surrey County Hospitals trust.