There are two primary components to health spending, unit prices and units of service consumed. Growth in per capita spending could be due to either and many analyses have been conducted to understand not only where the growth is coming from, but what, if anything, might be done to slow it. The Dartmouth Atlas project has focused on utilization variation across regions and providers, attempting to understand if the variation could be due to true medical needs or potentially reflects less than optimal medical care. A new report from the group once again finds substantial variation in the care of late stage cancer patients. (Dartmouth Report) As with most of the project’s reports, the population here is Medicare beneficiaries.
The variation is startling. A few examples: about 29% of cancer patients die in the hospital, but the rate ranges from 7% in Mason City, Iowa to 47% in Manhattan. The likelihood of being hospitalized in the last month of life, days in the hospital, admissions to and time in an ICU also showed similar ranges of variation. Interestingly, academic medical centers as a subset of providers also had great variation in care, which is somewhat surprising since these institutions are supposed to be the leading edge of evidence-based medicine. But even across these facilities there was a twofold variation. Facilities with more hospitalizations also tended to use much more aggressive life-extending services, such as intubation, feeding tubes, chemotherapy and CPR. All of these are futile in terminal patients.
The Dartmouth Report authors clearly believe that the variation they uncover reflects unnecessary and inappropriate care; often care that would not be wanted by the patient if they understood all the ramifications. The latter point is what should concern policymakers the most. How a terminal patient spends the last few days and weeks of his or her life is hugely important. Most people don’t want to die comatose in a hospital, attached to all kinds of equipment. Physicians have an obligation to ensure that patients have a fully realistic view of their remaining time and that care is consistent with that.