I remember an incident a view years back that really made me despair about our collective ability to get a handle on health care costs. I was consulting to a large, multi-site radiology practice that was taking a look at its MRI practice. While there were a small set of MRI scans that made up the bulk of the work (e.g. knee, hip, spine, etc.), each radiologist had small differences in how they ordered those scans, i.e. which view of the joint they wanted in what order. These differences tended to reflect differences in training.The site chiefs believed that it would be more efficient if the practice could agree on standardized approaches (known as protocols) to ordering these scans. This would be especially useful in cases where a study ordered by one radiologist (e.g. during an ER visit) was read by another in a different location. It would also be easier to develop standard templates for the radiologists to use in documenting their findings.All the chiefs could see the benefits of this idea. We put together a committeewith several subcommitteesto develop the protocols. Several of the chiefs volunteered some of their younger colleagues, because they thought this would be a good leadership development opportunity. We worked hard over several weeks and developed a proposed set of protocols for review.When the final set came before the chiefs for review, several became controversial. The draft protocols for knee, hip, and shoulder contained many more views of the joint than any of the chiefsall very experiencedwould generally order. Implementation of these protocols would have led to a significant increase in costs and a degradation in appointment access, since these studies would now have taken longer.The protocols were remanded, so to speak, back to our committee. What became clear in the ensuing dialogue was that the younger physicians who had recently completed training were demanding more views. Our first reaction was to assume that this reflected a lack of confidence, leading to a need for more views. It turned out, though, that it really reflected changes in the underlying technology that made the additional views more likelybut still not very likelyto reveal positive findings of one form or another. Rather than deferring to the chiefs, the younger physicians stuck to their guns, convinced that settling for a smaller number of views was poor medicine.We were at an impasse. I remember a later discussion among the chiefs. One of the chiefs, a very respected and experienced radiologist, ruefully admitted that he felt uncomfortable second-guessing his younger colleague about the knee protocol. You should see her at work, he said. she can find things I never would have found. While we did implement some of the protocols, we ultimately abandoned our effort to develop a consensus around the rest.One of the iron rules of health care is that if the technology is available, it will be used. And, yes, that means we will sometimes find things that would not have been found before and that patients will benefit as a result. But it may well be the case that one might have to scan many, many patients with this new and more expensive technology in order to find a single case of the finding in question. The physician wrestles with a tension between his or her dutyenshrined in both law and professional ethics to the patient in front of them and their responsibilityoften less clearly definedto be good stewards of the resources of the health care system. Patients, for our part, tend to complain about waste, fraud and abuse in the health care system when it is tied to the actions of others. When it comes to our own care, however, we tend to want the physician to do everything possible.I dont have a tidy set of ideas for reining in health care costs, mostly because I can easily imagine why most of them wont work very well. I am fairly certain, though, that if we dont find a way to use our ever expanding array of health care technology more judiciously, the health care system is going to collapse under its own weight.