Need for Advice

Abstract

Ms Schott and colleagues concluded in their study of the influence on funding of trials with medicinal products that publications from trials initiated by the pharmaceutical companies were often distorted. Also, such trials, they say, had positive results more often than non-pharmaceutically initiated trials. This was not explained by a difference in study quality. Their cited articles, however, were not based on insights from planning or conducting of trials but from their publication. Based on our working experience at an ethics committee of a medical faculty, we can state that advice is much more urgently required for non-industrially sponsored trials. Differences in planning clinical trials also seem to be responsible for the fact that more non-industrially funded trials do not take the desired course. It is of note, for example, that among 22 clinical drug trials with non-industry funders that received advice from the ethics committee at Cologne University in 2006–2008, 5 trials had substantial difficulties in recruiting subjects (fewer than half of the subjects were recruited during the planned recruitment period). This makes it doubtful whether such trials—especially on a background of limited funding—can ever be brought to a satisfactory conclusion. Problems in conducting trials will be reflected by the published results. By reverse conclusion, this should not lead to a situation where the pharmaceutical industry receives the blame for the successful conclusion, resulting in better publication opportunities. The authors did not sufficiently consider that different objectives exist for industry-sponsored and non-industry-sponsored trials, which in turn mean that for mere entrepreneurial reasons alone, industry sponsors will initiate only those trials that have a high chance of success. We agree with the authors’ reminder to take into consideration competing interests when interpreting study results. However, their comment that industry-sponsored drug trials are reported with a positive selection bias was not supported by reliable data. In order to state this, it would have been necessary to go further than looking at individual case reports and analyze the study quality comprehensively (including the planning and conducting of the trials, not only the publication of the results).