• Report

Refining cost to charge ratios for calculating APC and DRG relative payment weights. Interim report

Citation

Dalton, K., Freeman, S., & Bragg, A. (2008). Refining cost to charge ratios for calculating APC and DRG relative payment weights. Interim report. (Prepared for the Centers for Medicare & Medicaid Services. RTI Project Number 0209853.008). Research Triangle Park, NC: RTI International.

Abstract

In August 2006, RTI International was awarded a contract from the Centers for Medicare
& Medicaid Services (CMS) to investigate charge compression and other possible sources of
aggregation bias in setting the cost-based relative resource weights under the inpatient
prospective payment system (IPPS) using charges from claims data and cost-to-charge ratios
(CCRs) from cost report data. Aggregation bias can occur if services with systematically
different markup rates are grouped together to compute a single CCR. When that average ratio is
used to convert charges to cost, the cost estimates for component services with higher markup
rates will be overstated, while the estimates for component services with lower markup rates will
be understated. In the 2006 contract, RTI used inpatient claims and cost report data to identify
sources of aggregation bias and develop ways to improve the accuracy and precision of the CCR
component of the IPPS cost-based weight methodology. “Charge compression” is a particular
form of aggregation bias where weights for high-cost cases are systematically understated and
weights for low-cost cases are systematically overstated. The best known example of this comes
from a hospital industry practice of applying lower markups to expensive medical devices and
implantable items, but higher markups for routine medical supplies. RTI found strong evidence
of charge compression as well as other forms of aggregation bias in the relative weights for
diagnosis-related groups (DRGs) under the inpatient prospective payment system (IPPS), and
considered both accounting and statistical methods to reduce this bias.