Conventional wisdom states that the American public is simply unwilling to talk about the need to constrain the costs of health care. Pundits note that the “R word,” rationing, is one of the proverbial political third rails. Nonetheless, a public health official recently has found audiences willing and even eager to talk about cost containment as an overriding policy priority. He elicits this response after he points out to them that by the year 2033, the projected annual premium for a health insurance plan for a family of four will exceed the projected annual median income for such a family (Michael Fine, MD, personal communication).1

It is notable that this public health official is addressing costs in a public policy forum, and is not trying to introduce discussions of cost into individual encounters between patients and their physicians. Danis et al., in this issue, report on a helpful and thoughtfully designed focus-group study of the latter sort of cost conversation.2 They discovered that generally, patients are very leery of physicians bringing up cost issues, especially when the costs are borne by the insurer and not by the individual patient out of pocket.

As Danis et al. report, at least one focus-group participant was fully aware of the apparent disconnect between the policy setting and the medical encounter: “What’s funny is I totally agree with that statement [about the general social value of constraining costs] but I wouldn’t want [my physician] to say it to me. I might say it to him but I wouldn’t want him to say it to me.”

Danis et al. conclude that as the public health official above found, the best setting in which to talk with patients about cost may very well not be in the individual medical encounter. Is this a problem? Does the physician’s refusal to bring up costs challenge the ethical goals of patient autonomy and shared decision-making? The problem may be more apparent than real.

It is instructive to drill deeper into one scenario that Danis et al. presented to their participants: “…they’d had the worst headache of their life for 3 months, for which their doctor recommends an imaging study, either an MRI or CT.” The American College of Radiology featured as the first of its five “Choosing Wisely” admonitions, “Don’t do imaging for uncomplicated headache.”3 As to what counts as a “complicated” headache, a recent guidance from the National Institute for Health and Clinical Excellence in the United Kingdom lists 17 criteria requiring further investigation, and “worst headache of their life for 3 months” matches none of these criteria.4 It appears that this scenario might represent two possibilities:

  • The patient will not benefit from any imaging study, and by recommending such a study, the physician is practicing substandard medicine.

  • The physician has detected evidence of a possible serious, treatable underlying cause of headache and an imaging study may help to rule that cause in or out.

In the first possibility, the physician is simply practicing bad medicine. Involving the patient in so-called shared decision-making around the relative costs of a CT vs. and MRI scan will not change that basic fact and is inappropriate. In the second possibility, the physician presumably knows what sort of underlying cause she is searching for, and can weigh the advantages of MRI vs. CT in detecting that cause while not generating false-positive findings. The patient in this second instance has every right to expect a recommendation from the physician of which imaging test is superior for the task at hand, and not a discussion of comparative costs.

Imagine further that in the second instance, the physician carefully considers the pros and cons of CT vs. MRI and decides that for this patient and clinical presentation, they are equivalent. Assuming that the CT scan is cheaper, the physician can then with confidence recommend the CT scan. There is no reason to say, “By the way, we could talk about having an MRI, but that would be no better in my opinion and would only cost more.” The patient could quite reasonably reply, “If there is no possible extra advantage to an MRI, why would you even bring it up?” Sometimes going through the motions of shared decision-making is more misleading than helpful for the patient.

The point of this close analysis of the scenario is not nitpicking. The assumption of the authors appears to be: We have a serious policy problem in U.S. health care with cost overruns. One source of these overruns is the use of more expensive tests and treatments when cheaper ones would provide almost as much benefit, even if the more expensive technology has a marginal advantage. The ethical practice of medicine requires that patients be at least alerted, if not directly consulted, when cost considerations are included in the physician’s therapeutic recommendations. We now have to ask how to present such discussions of cost in ways that patients would find palatable.

I propose that the above discussion of the scenario suggests a slightly different set of assumptions: We have a serious policy problem in U.S. health care with cost overruns. One source of these overruns is the use of more expensive tests and treatments when cheaper ones would provide almost as much benefit, even if the more expensive technology has a marginal advantage. But another, and perhaps even more important source of these overruns is the use of technology that offers no patient benefit at all, and that is administered by habit or out of a vain search for extra certainty.5–7 Physicians need not feel obliged to introduce discussions of the cost dimensions of such decisions into conversation with patients, since the main question is how to secure maximum patient benefit and avoid harm. Physicians should, however, be prepared to frankly answer any questions patients may have about how costs do or do not factor into such therapeutic decisions.

Danis et al.’s findings suggest that while public discussion about cost containment policies is essential, this discussion ideally engages people as citizens rather than as patients.2 There are various mechanisms for engaging members of the general public in ethical deliberations about health policy matters such as cost containment and equitable access to care.8,9 However, in developing sound ethical policies for cost containment, it will be necessary to distinguish carefully between the rationing problem posed by low-benefit, high-cost interventions and the rather different issues raised by interventions that effectively offer no benefit, whatever the cost.6,7 As the headache-imaging scenario illustrates, it is important to keep these two different problems straight, whether the physician is responding to patients’ questions and concerns in the clinical encounter, or whether citizens are meeting to plan sound public policy.