Incentives to Reduce Use of Low-value Services

There is a widespread belief that a lot of health spending in the US is on “inappropriate”, “low-value” or “wasteful” services.  There have been wild estimates that as much as a third of all spending falls into this category.  Because of this believe, interventions have been created that might reduce the use of these services, including provider education, patient education and incentives to both to reduce the use of low-value services.  One of the incentive mechanisms is “value-based” insurance design, which generally offers lower cost-sharing for high-value services and higher cost-sharing for low-value ones.  Research recently published at the National Bureau of Economic Research examines the experience of one large Oregon public employer that implemented a value-based health plan design.   (NBER Paper)   This particular program raised cost-sharing by as much as 150% for sleep studies, upper GI endoscopies, advanced imaging services and overused surgeries, like spinal surgery for pain; all deemed to be low-value in many cases.

The new design was implemented in 2010 and the researchers used claims data from 2008 to 2013 to identify changes in use of the targeted services, comparing utilization by members covered by the new design compared to that in three public employee groups not subject to the increased cost-sharing.  The results suggest the raised cost-sharing was modestly effective; overall use of the targeted services fell by about 12%.  Sleep study use was reduced by about 20%, low-value surgeries by the same amount, endoscopies by 12%, and  advanced imaging by 7.7%.  There was no change in the trend of use in the comparison groups not subject to the value-based design.  The modest reductions suggest that patients may be relatively price-insensitive to the out-of-pocket cost of these services.  The authors did not indicate whether some of these members had access to health savings or health spending accounts, which might blunt the impact of higher cost-sharing amounts.

The authors did note, however, that they cannot distinguish whether all the reduced utilization was in fact low-value, since there are cases when all of these services can be appropriate, even critical to improving health.  That is a big problem with the enhanced cost-sharing approach; it can be a blunt instrument, just like use of high-deductible plan designs.  In addition, the services covered in this particular analysis don’t have a high utilization to begin with, so the modest reduction likely led to relatively small spending decreases.  While value-based insurance designs show promise, they may not have as significant an impact on spending as promoted by some advocates.