Two recent events underscore dilemmas at the heart of the future of Medicare, and health care reform more generally. The first one was the Hospital Trust Fund’s finding that Medicare will run out of money in 2017. The other was Medicare’s decision not to pay for virtual colonoscopy, a noninvasive CT imaging test to screen for colon cancer.

The story behind this decision and the reaction it has provoked illustrates just how difficult it is to control health care costs.

Medicare’s decision was based on a comparison of the risks and benefits of virtual colonoscopy and standard optical colonoscopy, an invasive procedure. After an exhaustive review of all published studies, and extensive consultation with experts and organizations, Medicare concluded that there was not enough evidence to determine whether virtual colonoscopy was as good as regular colonoscopy, particularly for the older patients that Medicare covers.

While some proponents argued that virtual colonoscopy would encourage more people to be screened for colorectal cancer, Medicare noted that there was no clear evidence that more screening had occurred in places where the procedure was available.

Medicare also addressed the question of whether the higher cost of virtual colonoscopy factored into its decision; it said that the costs were considered but emphasized that its decision was based on uncertainty about the clinical benefits.

When Medicare first proposed to deny payment for virtual colonoscopy, prominent industry groups, professional societies, and cancer patient advocacy organizations that strongly favored reimbursement protested vigorously. Members of Congress signed letters urging Medicare to change its position. After Medicare made its decision, opponents continued to argue that the scientific evidence was good enough to conclude that virtual colonoscopy is a reasonable alternative, and that Medicare beneficiaries should be able to decide with their doctors which approach was best for them.

We agree that respect for patient choice is at the heart of the practice of medicine and should be fully reflected in public policy. But this commitment should be understood in the broader content of what also matters morally for current and future Medicare beneficiaries – preserving guaranteed access to high quality medical care.

Adding to the challenge, what is really best for patients often is not clear from the analysis of published scientific studies. The conclusion that evidence is or is not sufficient, or that a particular technology has crossed the line from “promising” to proven, is determined by groups of individuals applying personal judgments to scientific results. The evidence does not speak for itself – people speak on its behalf.

In its recent review of virtual colonoscopy, the Blue Cross Blue Shield Association commented on the inevitable impact of subjective considerations in these policy decisions: “Given that much of the evidence supporting colorectal cancer screening is indirect, it is not so surprising that consensus groups reviewing the same evidence might come to different conclusions…”

The Association pointed out that even though they examined the same evidence about virtual and standard colonoscopies, different review groups emphasized different risks, like radiation exposure during virtual colonoscopy or anesthesia reactions during standard colonoscopy. Those choices about relative weighting of scientific information were judgments, not facts, and this is true of all such clinical policy decisions, no matter how “evidence-based.”

Bottom line: Medicare’s decision about virtual colonoscopy is a window into two difficult challenges at the core of any serious attempt to preserve Medicare and secure the goal of universal access to comprehensive health care.

First, we will be faced with tough decisions about which health services to buy, based on a lot of uncertainty and differences of opinion. More research on what works best for whom in medicine is critical, but uncertainty and differences in judgment will never be completely eliminated.

Second, these decisions will require ensuring that the needs and interests of all of us are respected. In the colonoscopy case, we need a mechanism to meet the special needs of those few patients for whom regular colonoscopy is medically contraindicated.

Let’s assume that the weight of evidence over time supports the conclusion that virtual and standard colonoscopies are equally effective. If virtual colonoscopy continues to be more expensive, then the decision not to reimburse for it becomes an ethical no-brainer. But in the meantime, it appears to be anything but.

And the colonoscopy example is hardly unique. New and more expensive alternatives to existing treatments and diagnostics tests are constantly emerging. There is strong demand from patients and clinicians for proton beam therapy for prostate cancer, MRI scans for uncomplicated back and knee pain, and coronary stents for mild symptoms of heart disease. In each of these cases, there is no clear scientific evidence that the new intervention produces better health outcomes than older ones.

How should American health care handle these difficult situations? First, we need to develop inexpensive and efficient ways to do the studies necessary to determine whether newer interventions are better.

One option would be for Medicare and private insurance companies to pay for new technologies only for those patients who agree to participate in comparative studies. The decision about unrestricted coverage for the new intervention would then be delayed until there is evidence that it produces better outcomes.

Meanwhile, we might allow patients who want new treatments or diagnostic tests to pay some or all of the extra cost or to purchase supplemental insurance to cover them.

We also need to do a better job of engaging the public and the medical community in discussion about the difference between what is newest in medicine and what is best. Considerable cost saving may be found if less expensive (often older) interventions turn out to be just as good as or better than more expensive (often newer) interventions. In the end, however, we may face much more ethically challenging choices, in which a new intervention offers a very limited additional benefit, but only at substantially higher cost.

Medicare’s virtual colonoscopy decision is but one of countless choices that will need to be made for both Medicare and health reform to survive. The key to success will be to anticipate the complexity, the necessity of trade-offs, and the fact that “evidence” does not make decisions – people do.

Ruth Faden, PhD., M.P.H., directs the Johns Hopkins Berman Institute of Bioethics. rfaden@jhsph.edu, 443-287-0413. Sean Tunis, M.D., directs the Center for Medical Technology Policy, a nonprofit organization in Baltimore and is the former Chief Medical Officer for Medicare. Sean.tunis@netzero.net, 410-963-8873.