October 11, 2015

Why Evidence-Based Medicine Is Insufficient To Drive Lower Health Care Costs

At a recent workshop hosted by the Institute of Medicine, Dr. Elliott Fisher of the Dartmouth Institute of Health Policy and Clinical Practice demonstrated that a sizable percentage of decisions made by health care professionals are discretionary and are within the range of what is popularly called “evidence-based medicine.” His fundamental argument is that we cannot contain health care costs significantly by making health care practices conform to “evidence-based medicine.” Instead, we will need to create systems in which the health care community selects less-costly, but equally-effective, approaches, over more-costly approaches.

The Dartmouth Atlas survey, which I strongly recommend that everyone interested in health care transformation read, proves that there are wide variations in health care cost among American communities with not only no better outcomes, but, in some instances, slightly worse outcomes in the higher-cost communities. A recent posting of an article entitled “The Cost Conundrum” in The New Yorker profiles McAllen County, Texas, one of the two highest-cost areas, the other being Miami. In it, the report Atul Gawande, demonstrates the local health care provider community is oblivious to how out of line its practices and costs are, but the reporter also notes that nothing the practitioners do would ever be prevented by an evidence-based medicine system.

I can relate to his argument based on my own personal experience. Several years ago, I had an angioplasty to address a relatively limited arterial blockage in one of my arteries. My condition was not life threatening, and the blockage was discovered in a routine stress test during a routine physical. Nevertheless, having this entry on my medical record sometimes triggers unintended consequences when I have a health care system encounter.

During Memorial Day weekend, I had a little discomfort on the right side of my chest. When I called my internist and my cardiologist, both offices had other physicians on call and both were not reachable at the office because of the holiday. Both on-call physicians suggested I go to the Emergency Department of the Stamford Hospital to get some precautionary tests. I arrived at 11 am Sunday and got a blood test, an EKG, and a 64-slice CAT scan, as well as having my heart monitored for about two hours. The test results were normal, and indicated that the likelihood that I had heart problem was very low.

Nevertheless, the cardiologist on call suggested that I stay and have 2-3 more blood tests spaced 6-8 hours apart to insure that I did not have an enzyme that indicated heart damage. I accepted her recommendation, and the 3 additional blood tests, as well as 2 additional EKGs, confirmed that the chest discomfort, which, by the way, disappeared by Sunday noon, was not heart-related. I was discharged Monday morning at 9 am.

My cardiologist and internist called me right after the holiday when each returned to their office, and both suggested that it was unfortunate that this event happened on the holiday weekend. Both indicated that they would have been less likely to have admitted me to the hospital because, being familiar with my history and medical condition, they might have made a judgment that my problem was not heart-related. At the same time, they said that when a colleague deals with someone who is not their patient, with my history and my description of symptoms, the tendency is to be more cautious and that hospital admission is more likely. When I related my story to a physician familiar with national health care practices, he told me that if I were in certain parts of the United States, doctors would have ordered me to go in for an angiogram, in addition to the other tests, and to be hospitalized a day or two longer, and that the event would have been far costlier.

I tell this story because it is a good example of how different health care practitioners in different communities, or even the same community, might deal with the same situation very differently. None would be wrong, according to the dictates of evidence-based medicine, but the costs would be radically different, ranging from zero for an encounter with my personal physicians who might have had a “watch-and-wait” approach, to the most aggressive and proactive approach, which would have generated huge health care expenses, probably in the tens of thousands of dollars.

This problem illustrates why I favor a health care payment system recommended by Dr. Fisher, rewarding the creation of “accountable care organizations” that self-manage their care costs downward, while improving or maintaining acceptable levels of health care quality. Dr. Fisher’s approach is described in a number of well-researched and well-argued scholarly articles.