HOOKED contains a considerable discussion of the evidence that pharmaceutical trials funded by industry are substantially more likely to favor the company's drug than studies funded by neutral parties (by about 4-fold). However, if people are going to prove it some more, I'm game, especially if they help us see how it's done.
A team of investigators hailing from Spain, Mexico, and Wisconsin reviewed 275 studies of inhaled corticosteroids that were company-funded and compared 229 studies of the same class of drugs that were non-industry-funded. The specific question they explored was how likely the study was to report significant adverse effects.
They discovered that only 34.5% of the industry-funded studies reported any adverse effects compared to 65.1% of the neutral studies. Once they adjusted for important features of the study design, the difference was no longer significant. This in turn indicated that the way industry-sponsored studies underreport adverse effects is by being deliberately designed with trial methods that are likely not to detect such events--such as by using the study drug in very low doses.
HOOKED explains that according to the best thinking in today's drug industry, research is a marketing tool--in fact, if the company CEO allows the scientists to design trials without the marketing depatrtment looking over their shoulder at every step of the way, the CEO should be fired. Here we see just one more example of this philosophy in action.
Nieto A, Mazon A, Pamies R, et al. Adverse effects of inhaled corticosteroids in funded and nonfunded studies. Arch Intern Med 167:2047-53, Oct. 22, 2007.