They believe it would lessen the risk of wishful thinking – an unconscious tendency by doctors to expect new drugs to perform better than the older ones they are being tested against.
This perception of a drug’s effectiveness was named ‘wish bias', when it was first reported more than a decade ago. In a report published in the current issue of Annals of Oncology* the research team says it may help to explain why the response rate of tumours to a new drug decreased over time.
Dr Roldano Fossati and colleagues from the Mario Negri Institute in Milan combed through the results of 29 randomised trials that took place between 1975 and 1999. They analysed the response rate of 2,234 women with advanced breast cancer who had been in the arms of the trials using the anti-cancer drug, doxorubicin (Adriamycin). They extracted the global (i.e. complete plus partial) response rate from all the trials.
"We found that every five years there was a 11% relative decrease in the odds of a global response and most of this decrease was in the partial response rate, where evaluation involves more subjective judgement than is involved in determining complete response," said Dr Fossati.
Of the 29 studies they investigated, only one had used a double blind** approach, and in two studies patients’ records were just externally audited.
"In these situations a bias due to financial and academic conflicts of interest or more subtle forms of ‘wish bias’ could easily arise and account for an impression that when a drug is new it does better," said Dr Fossati.
The team carried out the assessment of the 29 studies to test their hypothesis that doctors’ unconscious favourable attitudes to new treatments may result in a tendency to overestimate their efficacy. They chose doxorubicin because it had been used alone or in combination since the early 1970s and was still widely used as a comparison against which to evaluate new breast cancer drugs.
"Under the effect of the biases we describe we would expect to observe a declining response rate to doxorubicin that was still detectable after taking into account changes in disease management over time," said Dr Fossati.
This was what the research team found, although Dr Fossati stressed that results should be interpreted within the context of the limitations to the study, which was that they looked at only one drug (doxorubicin) in one clinical context (advanced breast cancer).
According to Dr Fossati, blinding is uncommon in these types of clinical trials. In an earlier meta-analysis by the Milan team of 189 trials involving chemotherapy and hormone therapy for metastatic breast cancer, only six involved blind evaluation and only 23 trials had involved independent or extramural assessment. There was no reason to suppose that these were not typical of cancer drug trials.
"Since indirect evidence for the existence of a wish bias emerged from our analysis, we believe that blinding should be strongly recommended for any subjective endpoint assessment such as response, time to progression of disease and so on, and the search for this kind of bias should be unremitting in order to push the medical research community to guarantee objectivity and maintain public confidence."
* Does a drug do better when it’s new? R. Fossati et al. Annals of Oncology. Vol. 13. No. 3. pp 470-473. ** A double-blind trial is one where neither doctors nor patients know which treatment the patients are receiving.
Notes:
* Annals of Oncology is the monthly journal of the European Society for Medical Oncology. Please acknowledge the journal as the source in any reports.
* A complete copy of the paper should be available 13 March on website: http://www.annonc.oupjournals.org
* An embargoed PDF copy of the paper is available immediately from Margaret Willson.
Contact:
Margaret Willson (media relations officer)
Tel: +44 (0)1536 772181
Fax: +44 (0)1536 772191
Mobile: +44 (0)7973 853347
Home tel: +44 (0)1536 770851
Email: m.willson@mwcommunications.org.uk
Journal
Annals of Oncology