Hospital-Insurer Bargaining and the Relationship Between Charges and Prices

Tuesday, June 14, 2016: 1:55 PM
F50 (Huntsman Hall)

Author(s): Ellerie Weber; Chapin White; Eric Floyd

Discussant: Kevin Pflum

According to recent reports by the Health Care Cost Institute, rising spending for the privately insured results from rising payments (“prices”) by payers to hospitals and facilities, rather than from increased utilization. Also documented is that charges (“list prices”) vary greatly across markets, and charges are growing rapidly. However, it remains unclear whether increasing charges and prices are related, because researchers are unable to obtain joint data on the two; until now, contract terms have been deemedtrade secrets by hospitals and plans, and never before used by researchers. We have acquired new data from the Colorado All Payer Claims Database (APCD) that – uniquely among large claims data – contains charges from over 100 hospitals as well as the negotiated final prices paid to those hospitals by approximately 20 commercial and public health plans. We combine the Colorado APCD with the New Hampshire APCD data toanswer this heretofore outstanding question about the relationship between prices and charges. We define the ratio of allowed amounts to billed charges as the “reimbursement rate” (RR); a RR of one means that the hospital was paid the total amount for which they billed. Preliminary findings from New Hampshire suggests that final prices (allowed amounts) are on average 67% of billed charges, with significant variation across major procedure groups and across the site-of-care (inpatient versus outpatient) setting. Moreover, the distribution of this RR is heavily left-skewed, with a high proportion of ones. Our main investigation estimates how allowed amounts vary with changes in charge prices, which provides evidence as to whether charges are economically meaningful. We then explore whether certain characteristics, such as product, hospital and health plan characteristics moderate the relationship between prices and charges. Finally, we use the charge and price data to predict the broad contract terms between health plans and hospitals, for example, whether contracts are written as per-diem, discounted charges, or case-based (DRGs). Our results inform price transparency and other health policies that aim to reduce healthcare costs.