Sir

Your Naturejobs Special Report “Trial blazers” ( Nature 442, 480–481; 2006) highlights the demand for investigators adept at planning, running and managing clinical trials for the pharmaceutical industry.

The industry's share of total biomedical research increased from nearly a third in 1980 to nearly two-thirds in 2000. Never-theless, a report documenting combined data from more than 1,100 studies showed that industry-sponsored clinical trials are “significantly more likely to reach conclusions that were favorable to the sponsor than were non-industry studies”, possibly because of publication bias or selection of an inappropriate comparator to the drug being evaluated (J. E. Bekelman, Y. Li and C. P. Gross J. Am. Med. Assoc. 289, 454–465; 2003).

Efforts being made by some academic institutions and peer-reviewed journals to control financial conflicts include the disclosure of financial interests by researchers in academia and as authors in peer-reviewed journals; statistical analyses performed by independent academic researchers; and research protocols and results from clinical trials made fully available in registries accessible to the public.

Given that the trend towards industry's increased share of biomedical research will undoubtedly continue, further evaluation of differential bias between investigator- and industry-initiated contracts for clinical trials within academic medical centres is needed. Of particular note in this respect is the fact that some investigator-initiated trials are funded entirely by industry, whereas others are funded via peer-reviewed grants or foundations.

Within academic institutions, for example, trials funded by industry but initiated by the investigators typically have 10–20% lower overhead costs. They tend to restrict industry sponsors to ensuring adherence to good clinical-practice guidelines and to checking data for serious adverse events, notation errors and omissions. Hence industry has very limited influence on the conduct of the trial. In contrast, trials that are both initiated and funded by industry within academic institutions have higher overhead costs, usually in the range of 30–40%, paid to the academic institution. Such trials allow greater involvement by the industry funder, for example in writing the protocol, training and assisting investigators, and helping in data collection and analysis.

Academic investigators should be encouraged to lead investigator-initiated clinical trials, thereby limiting industry involvement and reducing any bias that may creep into trial protocols, conduct and results. This should result in better patient care via increased knowledge about the safety and efficacy of the drug therapy under evaluation.