IOM Report on Geographic Variation in Spending

It seems that geographic variation in health spending provides a never ending series of reports and briefs.  Because of concerns about the effect of such variation in regard to Medicare spending, Congress asked the Institute of Medicine to investigate and report on the topic, with an eye toward informing policy decisions that might include some kind of geographic reimbursement modifier.  The Interim Report has been issued, which focuses solely on Medicare fee-for-service variation.   (IOM Report)   The final report will cover variation in commercial, Medicaid and Medicare Advantage spending as well, and will make recommendations regarding what, if any, adjustments should be made to payment mechanisms.  The report notes that one potential issue with any adjustments is that health care decision-making typically occurs at the individual practitioner or provider organization level, not by a geographic region.  So the mechanism by which a geographic adjustment would affect that decision-making needs to be clarified.

The authors note that geographic variation exists at all levels of geographic analysis, down to the individual practitioner, and across all service categories.  And even the same practitioner may be high cost for some conditions but not for others.  This fact alone suggests that geography has far less to do with variation in spending than does individual provider practice pattern.  Among other implications, this means that the Hospital Referral Regions, or even the Hospital Service Areas, used by Dartmouth for example, have little value in understanding variation.  Similarly, if value is being measured by health process or clinical outcomes, there is not a consistent level of performance on these measures within any unit of geographic measurement.  And while some variation is explained by socio-demographic and health status differences across units of geography, a great deal of variation is not explained.  And resource use, measured by utilization, does not appear correlated with either high or low quality, undermining the whole rationale of geographic based spending analysis, that somehow lowering spending would get at least equal quality.  At a service-type level, the most variation occurs with post-acute services, especially home health and skilled nursing facility care, and the next greatest in acute inpatient care.  Other categories have little variation.  Overall, the Interim Report suggests that the IOM is unlikely to recommend that a geographic adjustment be made to Medicare reimbursements.