Modifying DRG-PPS to Include Only Diagnoses Present on Admission

OBJECTIVE: The inability to distinguish complications acquired in hospital from comorbid conditions that are present on admission (POA) has long hampered the use of claims data in quality and safety research. Now pay-for-performance initiatives and legislation requiring Medicare to reduce payment for acquired infections add imperative for POA coding. This study used data from 2 states currently coding POA to assess the financial impact if Medicare pays based on POA conditions only and to examine the challenges in implementing POA coding. METHODS: Medicare payments were calculated based first on all diagnoses and then on POA diagnoses in the Medicare discharge abstracts from California and New York in 2003, using the Diagnosis Related Group (DRG) -based Prospective Payment System (PPS) formula. The potential savings that result from excluding non-POA diagnoses were calculated. Patterns of POA coding were explored. RESULTS: Medicare could have saved 56 million in California, 51 million in New York, and 800 million nationwide in 2003 had it paid hospital claims based only on POA diagnoses. Approximately 15% of the claims had non-POA codes, but only 1. 4% of the claims were reassigned to lower-cost DRGs after excluding non-POA diagnoses. Excluding non-POA diagnoses resulted in reduced payment for operating costs, but increased outlier payments because some of the claims were designated as "unusually high cost" in the lower-cost DRGs. POA coding patterns suggest some problems in current POA coding. CONCLUSIONS: To be consistent with pay-for-performance principles and make claims data more useful for quality assurance, incorporating POA coding into DRG-PPS could produce sizable savings for Medicare.