Eventually, it will dawn on many Americans that nothing less than rigorous, science-based, and mandatory guidelines will be necessary to control the ever-rising costs of health care. By “mandatory” I mean that they will be used to set coverage benefits, determining what will and will not be provided, and using cost as one criterion. They will have to start with the Medicare program and, if the past is any guide, they will be adopted by the private sector.

There are no other methods on the horizon, much less in the proposed reform legislation, at all likely to bring the rate of annual cost escalation down from the present 6% level to the 3% needed for long-term affordable and sustainable health care. Only the power, and the will, to say no to patients, doctors, hospitals, and industry can do that job. And only rigorous treatment standards, and if necessary price controls, can do that.

The recent outburst of anger by many at the recommendation by the U.S. Preventative Services Task Force that routine mammography not begin until age 50 shows how difficult it will be to move toward that goal. A number of disturbing features of the outburst make that clear. The most notable feature was the absence of any serious objection to the scientific quality of the research behind the recommendation. Instead, it took the form of a charge (with no evidence to support it) that its tacit aim is to ration care, and a forerunner of likely efforts to turn recommendations into rules.

Hardly surprising was that the assault was spearheaded by those groups most effected professionally and monetarily by the recommendation, the National Cancer Institute, the American Cancer Society, and the American College of Radiology. 

A common criticism as well was that the recommendation was a threat to the doctor-patient relationship, presented as the final and absolute arbiter of treatment decisions. That criticism echoed a major reason (along with industry opposition) why the $1.1 billion comparative effectiveness program initiated as part of the President’s stimulus package was immediately neutered by prohibiting its research findings to be used to formulate treatment guidelines or even recommendations. The screening guidelines were taken as an ominous warning of worse to come, that of the slippery slope hazards of a recommendation, and even one which in fact took no power away from doctor-patient decisions.

Perhaps the most common reaction, however, was one bound to plague every future finding of comparative effectiveness research – much less the even more desirable cost-effectiveness research (if we ever get it). How can we tolerate any finding, the indignant response seemed to say, that shows that even one statistical death will result from a research-dominated calculus based on population studies?

It was once said that one picture is worth a thousand words. It might now be said that one statistical death in a government recommendation that encompassed thousands is one too many. The task force calculated that there would be one death in every 1,900 women undergoing annual mammography over a 10-year period; that is 19,000 procedures. Of course many more lives could be saved if every woman over age 20 was screened annually. Even the critics of the government recommendation did not go that far, tacitly recognizing that it makes sense to set some limits – but not conceding the rationality of doing so.

Shortly before the breast screening debate got under way The New England Journal of Medicine published an account of a little-known program in Washington State called the Health Technology Assessment Program, designed to be used by various state health agencies. The purpose of the program, initiated by the legislature with only one negative vote in 2006, is to assess the clinical effectiveness, safety, and cost effectiveness of those technologies thought likely to be overused or underused or to raise questions of safety or cost effectiveness.

If it is determined that a technology should not be covered, even in cases of medical necessity, the state will not allow it in their programs. The assessment agency can, that is, ration care failing to meet its standards, but also mandate beneficial treatment that can raise costs. 

That program points us in the right direction for cost control: an agency that assesses technologies and makes binding decisions about them. If the Mayo Clinic, Geisinger, and Intermountain are taken to be exemplary models for the organization and delivery of health care, the Washington technology assessment program should be emulated to control costs.

The Senate health reform bill’s proposed Medicare commission to monitor the program and keep its spending in check, combined with the work of a strengthened comparative effectiveness effort (to encompass cost effectiveness as well), could come close to doing what the Washington state program does. That would be serious reform of a kind now absent, and the ingredients for doing so can be found in the Senate and House bills, but in a diluted form. They only need some fortification. A lot of it.

Daniel Callahan is editor of the Health Care Cost Monitor and author of the new book, Taming the Beloved Beast: How Medical Technology Costs are Destroying Our Health Care System.