Impact of Mortality-Based Performance Measures on Hospital Pricing: the Case of Colon Cancer Surgeries

We estimate price regressions for surgical procedures used to treat colon cancer, a leading cause of cancer mortality. Using a claims database for self-insured employers, we focus on transaction prices, rather than more commonly available billing data that do not reflect actual payments made. Although the responsiveness of prices to hospital performance depends on the impact of quality on the slope of the quantity-demand of the payers, which are not known a priory, it is often assumed that higher performing hospitals are able to command higher prices. To test this hypothesis we construct performance rankings, based on hospital excess-mortality and incorporate them into our price models. We are interested in the type information available to large payers who negotiate prices on behalf of their members. To get a cancer-specific index we emulate the widely-reported risk-adjustment methodology used in the federal Hospital Compare reporting system for ranking cardiac performance. The effects were consistently negative in all models (adverse quality reduces price), though not significant. However, we observe a rational pricing structure whereby higher treatment complexity is reflected in higher price differentials, controlling for patient characteristics and market structure.