Fact Checking Ron Herberman's Alert

At this writing, Google News has a list of some 900 articles on the cell phone health alert issued by the University of Pittsburgh a couple of days ago. The Post-Gazette, the hometown paper, broke the story on the same day (it got an advance copy), and though some newspapers like the Baltimore Sun ran their own write-ups, the vast majority relied on the Associated Press for their coverage.

Unfortunately, the AP reporters made a hash of it. Their story suffers from a number of serious errors as well as misplaced emphases, which made it seem as if the actions of Ronald Herberman, the director of the University's Cancer Institute, were misguided and inconsistent with the published literature.

Here are some corrections and clarifications:

• The AP reporters cite a 2008 University of Utah meta-analysis of nine published studies as finding no brain tumor risks among cell phone users. In fact, it did point to a 25% increase among long-term users, that is, those who had used a mobile phone for ten or more years. This increase reached (just) statistical significance.

• The AP states that Herberman is relying on the as-yet unpublished Interphone study, a 13-country effort to investigate possible tumor risks from cell phone use. It's true that the overall Interphone results have not yet been public —it's now close to three years behind schedule— but a number of the participating countries, either individually or in groups, have reported elevated incidences of three different types of tumors: glioma (brain tumors), acoustic neuroma, and parotid gland tumors among long-term users. These findings have been published in leading peer-reviewed journals.

• The AP states that the Interphone study suffers from selection bias and casts doubt on the reliability of its risk estimates. This remains an open and highly contested issue among members of the project team and is a major reason for the delay in the release of the final results. While the NRC's recent report cites selection bias as a possible confounder, it also details other reasons as to why the Interphone study may underestimate the risks. It is far too early to toss out the project's final results as unreliable. Shouldn't we at least wait for it to be published before trashing ten years of work?

• While the French Interphone study does not show significant elevated risks, it does point to increases. These were sufficiently strong to prompt the French Ministry of Health to issue an advisory reiterating the recommendation that children be discouraged from using mobile phones.

• It's true that most studies have not indicated a tumor risk, but most of these have only looked at short-term users. For instance, the Muscat study included only 17 cases which had used a phone for four or more years. The NCI study also had very few long-term users: 22 who had used a cell phone for five years or more. Neither discloses how many had used a phone for at least ten years —maybe none.

• It's also true that the Danish study showed no increased risks. But as a cohort study it could not provide any information on the side of the head the phone was used, which, not surprisingly, turns out to be a key variable. The study also excluded corporate accounts, which tended to have the heaviest users.

To be sure, the jury is still out on cell phone health risks. That said, it will be a long time before we know what the true risks may be. It seems that the American Cancer Society is waiting for conclusive data showing elevated tumor rates before it is willing to advise caution. And some like physicist Lennart Hardell, that point to long-term health risks. Surely it is time to have an open discussion on what these data mean and how we should protect the most vulnerable in our society. After all, there are now 260 million regular users in the U.S. and more than two billion worldwide, and the wireless phone industry, still looking to grow, is marketing its phones to younger and younger children. This is exactly what Herberman has achieved with his alert.

They say journalism is the first draft of history, but in this case the AP story was a rough draft at best. But we should also understand that writing a story on a complicated subject under a tight deadline is far from easy. Reporters must rely on their sources to point them to the facts. In this case, they were misled by those who want to play down the risks so that our love affair with cell phones can continue unchallenged.