More on Regional Spending Variation

In a continuation of the debate that never seems to end, the New England Journal of Medicine has published juxtaposed Commentaries by a researcher questioning some of the Dartmouth Atlas research on geographic and provider spending variation and researchers who have done some of the Atlas’ work.  (NEJM Perspective) (NEJM Perspective2) The underlying importance of this debate is the implication that if there is no improvement in outcome linked to higher-spending regions and providers, substantial savings could be had by reducing payments to those regions and providers.

The author of the Perspective questioning the Dartmouth researchers’ conclusions points out that the research assumes that hospitals are responsible for all the costs associated with patients who were inpatients at the hospital, that the Medicare fee-for-service patients who are the subject of Dartmouth’s research may not be representative of all patients, that the Atlas researchers’ methods may not sufficiently adjust for severity of illness and that the research doesn’t directly examine cause and effect between spending levels and health outcomes.

The Dartmouth authors counter that prospective studies on Medicare members come to the same conclusions as does their retrospective research.  They point out that they do extensive severity and other adjustments to try to ensure that they are comparing the same type patients in different hospitals.  These researchers stick by their conclusion that there are excessive health care costs at the level of the hospital/provider network and that these costs are not correlated with better health.

This is not an unimportant issue, as its outcome should guide public policy.  If, as frankly seems likely, there are providers and systems which spend more to get no better outcomes, then policymakers should find an appropriate remedy, whether it is bundling of payments or reduced reimbursement for hospitals and physicians who persist in inappropriate and expensive care patterns.