Sacristan and colleagues suggest that we ask not why some pharmaceutical companies supported the Task Force but why others did not. In fact, I raised both questions in my editorial. One might expect reimbursement agencies to be enthusiastic about the development of guidelines that are effective for the purposes claimed—distinguishing cost-effective drugs from the others. But they held back: Emptores cavebant.

Sacristan and coworkers would have us infer that this indicates bias on the part of purchasers. My inference was rather that this should raise suspicions about the guideline process itself. As emphasized in the editorial, the "lack of consensus about methods" in economic evaluation is as old as the activity itself. Disagreements over "classic" questions are not rooted in faulty analysis or incomplete data. They arise from differing conceptions of the question itself, and of its context, which is why so little convergence has occurred over time.

Consistency of method may be essential to science, but here we address evaluation. Only humans have values, and their values are not and never have been consistent.

Regarding sources of bias, we would do well to remember that pharmaceuticals are archetypical of "knowledge-based" products, with high fixed costs and low marginal or variable costs. Most of the money is spent before the first pill is rolled. The initial research and development outlay (and we hope much more) can only be recouped by selling individual items at a price well above their unit cost. You must move the product or lose your huge initial investment.

The result is a heavy emphasis on marketing—through any possible channel, including economic evaluation—as a condition of survival. Drug industry behavior is a natural and inevitable consequence of this cost structure; the "conventional prejudgments" referred to by Sacristan and colleagues are simply an attempt to keep a firm grasp on the obvious.

Ms. Jung's suggestion of a neutral agency for evaluation is attractive, but financing such an agency is not so simple. Consumers' Union is supported entirely by subscribers, fiercely defending its complete independence, financial or otherwise, from sellers. And rightly so; those who pay the piper usually manage, one way or another, to call the tune. How long could an industry-funded institute of medical research preserve its independence?

In the end, drug buyers and reimbursers will have to do their own evaluations and make their own purchasing decisions. Offers of participation and scientific cooperation from sellers always spring from the same underlying motive, to move the product. What else can they do?

Robert G. Evans, PhD

University of British Columbia

Vancouver, British Columbia, Canada