It seems only fair to tell patients about the conditions that physicians work under and how this might affect the care they receive. Dr. Friedman should be commended for explicitly proposing how this can be done. However, we have several concerns about the physicians' code that Dr. Friedman proposes. We think it is unrealistic to ask physicians to say that they are under no pressure to provide anything other than "optimal" patient care. Physicians are increasingly under intense pressure to consider whether specific medical benefits are worth their costs. Although this pressure should not make physicians withhold necessary care from patients, we think it is appropriate for physicians to withhold marginal benefits that come at great financial costs. It is also unrealistic to require physicians to inform patients whenever marginal benefits are withheld because such rationing is ubiquitous. For example, when physicians see a patient for 12 minutes instead of 15 minutes, or order a "panel 7" instead of a "panel 24," they are potentially withholding a benefit. Because physicians are under intense pressure to see more patients in less time, informing patients of the multitude of tiny benefits that are not pursued may not be the best use of this precious time. A more feasible alternative would be for physicians, at the first visit with the patient, to say "part of my role as a physician is deciding when people really need certain tests or treatments. I have a responsibility not to waste your and everybody else's money on high-cost, low-benefit services, but to decide, using my best medical judgment, when you really need something. Of course, you can always get a second opinion or appeal my decision." In addition, when physicians think that a service is medically necessary but not covered by the patient's insurance, they should disclose this to the patient and help the patient obtain the service.

We agree with Dr. Galanos that a major risk of rationing is that it will be done in a biased manner. Indeed, cost-effectiveness analysis, seemingly the most promising aid in guiding rationing decisions, discriminates against the elderly [1]. We hope that our article will help physicians recognize when they are rationing. We see this as a necessary first step for discussions about when and how bedside rationing can take place fairly, without bias, and in a manner that preserves the integrity of the physician-patient relationship.