That is a very important question, which is not easy to answer. However yes, of course, studies commissioned by industry—possibly even conducted by industry—can be perfectly valid. So we must be cautious and make decisions on a case-by-case analysis of the protocol for each study.
Having said that, I believe that a serious problem exists when we find results that are relatively unimportant. For example, as regards the significance of the risk, beyond the figure, often what would make a big difference is the way it is presented and interpreted. For example, for studies funded using public funds, that is what led the United States, at one time, to ask that at the end of a certain number of years, in other words once researchers had exploited their results, the data itself be made available for new analysis to see if other people would interpret the same results differently. That is undoubtedly the reason why Interphone is taking so long to appear. There is undoubtedly disagreement as to the interpretation of the results among the numerous co-investigators from the 13 countries: how much is due to a real effect, how much can be the result of bias?
When studies are partially funded by industry as well, as was the case for Interphone, a certain number of safeguards must be put in place that can be... as was done by Interphone, and the funds were managed by the International Union against Cancer. That way, there was not a direct link between the International Agency for Research in Cancer and the financiers, which gave researchers a little more latitude.
I believe a great deal of transparency is required. That way at least once everyone knows what is being done, we can get clearer results. But in the past, several studies have shown that in general, the source of funding, unfortunately, is an extremely important risk factor for the results one might expect, and this is not limited to electromagnetic fields.
Studies today are complex, especially when we look at the effect of low doses with anticipated risks that are generally low and that can be in the future. We must be as vigilant as possible, have groups overseeing the conduct of the studies, and a second analysis of the results by independent teams. Moreover, the same is true for public studies; they are sometimes analyzed again by industry. A great deal of transparency is required; I believe that is the best we can do.