Send in the Frowns

Nature Magazine, owned by London-based Macmillan Publishers Ltd., is considered one of the world's leading science journals. As a first-time, non-academic author, I should be honored that they chose "The $800 Million Pill" for its lead book review in this week's issue. I've really been looking forward to my first review by someone who comes from the world of science and understands the interplay of scientific and economic realities discussed in the book.

But one has to wonder about what has happened to standards in the scientific publishing world when they asked a "scholar" from the American Enterprise Institute -- an avowedly conservative think tank in Washington, DC that takes considerable funding from the drug industry -- to write the review. Funding for John Calfee's own research, which he discloses at the end of the review, comes in part from pharmaceutical firms.

The editors must have known there no possibility that the review would be positive when they sent it out to Calfee. As they say in England, that hardly seems cricket.

Despite the many positive things Calfee says about my writing and understanding of science and medicine, the review was largely a stale rehash of industry arguments that I debunk in the book. I won't repeat those discussions here.

Rather allow me to point out that Calfee breaks the cardinal rule of fairness in book reviewing. You may quarrel with a person's argument, but you should at least accurately present it at the outset of the review. He failed this basic test by misrepresenting my analysis of the $800 million cost-per-drug figure put out by industry-paid researchers.

He says my analysis largely rests on a Public Citizen analysis and a study by the Global Alliance for TB Drug Development. My narrative does go over those studies and the arguments they've engendered (the industry trade group hired the accounting firm of Ernst and Young to debunk the Public Citizen study). But my own analysis relies on the fact that at least half of industry R&D is wasted on drugs that add nothing to physicians' armamentarium for fighting disease, and thus can more properly be categorized as corporate waste. He also says I "defy sound economic reasoning" by ignoring the "opportunity cost" question of R&D expenditures. In fact, I discuss that at length (I argue it's an expense funded by current drug consumers) and quote academic experts to bolster my argument.

Calfee doesn't mention either of these discussions. He describes my arguments as rehashes of others' studies. Indeed, he ignores the fact that I specifically state at the end of the chapter that if you take these factors into consideration, the $800 million industry figure would be "similar to the Global Alliance" figure. That's hardly "relying" on or "accepting" the results of their study.

There are other points in the review where Calfee makes basic mistakes (he calls Gleevec a biotechnology product) and inaccurately describes what I wrote. But suffice it to say here that when reading reviews by partisans in a debate -- even in the pages of a prestigious journal -- consider the source.

Here's the relevant paragraph from the June 24, 2004 Nature review:

"Relying mainly on non-scholarly critics such as Public Citizen (a lobby
group in Washington DC founded by Ralph Nader) and the Global Alliance for
TB Drug Development, Goozner argues that DiMasi and colleagues grossly
overstated the true costs of developing useful drugs. He accepts instead the
Global Alliance's estimate that discovering and developing a new
tuberculosis drug would cost only US$115-240 million. Essentially, Goozner
argues that the editors and reviewers of the Journal of Health Economics
overlooked some basic methodological flaws in the paper. His argument is
unpersuasive, though, partly because it defies sound economic reasoning on
opportunity costs. It also ignores important limitations in the Global
Alliance's study of developing a hypothetical drug for a single condition
(tuberculosis) and using relatively small clinical trial sizes, non-market
cost data and highly conjectural failure rates."

Posted by gooznews at June 24, 2004 09:29 AM