Privately Insured Patients in US-Cost Variation in Hospital Spending

How do health care costs vary across the country. Although the team at the Dartmouth Atlas has done this exercise with patients in Medicare, there has been less study of region variation in health care spending among the privately insured with the notable exception of a 2013 Institute of Medicine report. In a study by Cooper et al. presented at the 2016 ASSA meetings, The authors use data from the Health Care Cost Institute (HCCI) to measure regional variation in prices. The data cover about 27.5% of individuals with employer sponsor insurance or 14.2% of the US population.

The study finds that a low correlation between Medicare and private spending per person (0.140). While La Crosse, WI and Rochester, MN are high cost for Medicare, they are actually low cost for patients who are privately insured. Other cities, like San Francisco, in high cost for both. McAllen, TX—one of the most expensive for treating Medicare—is about average cost

Whereas price explains a large portion of national variation in inpatient private spending; for Medicare, utilization is the key drive of differences in healthcare spending.  This finding is similar to what was found in the IOM report.

The authors also find substantial variation in prices, both within and across markets. Variation in hospital prices is not only due to high cost procedures—such as knee replacement—but also relatively low-cost procedures such as lower limb MRIs.

Like most of previous research—such as Leemore Dafny’s finding that hospital mergers increase price—the authors also find that higher hospital market concentration (e.g., if the hospital is a monopoly or in some cases a duopoly) is associated with higher hospital prices.  Specifically, hospital prices are 15.3% higher when they are a monopolist. There is some suggestion that higher hospital quality is associated with higher price.

Although this analysis is not causal, the authors scope of data and statistics can help motivate future studies to better understand regional variation in hospital prices.