The data: This analysis uses the 5 percent HCFA Standard Analytical Files (SAF) for 1996 through 1998 to compare the use of coronary artery stents among hospitals. This data source was selected because it includes data from all short-term acute care facilities in the United States, allowing representation of all 100 Top Hospitals™ winners and non-winners.

The comparison groups: The analysis compared non-winners with 100 Top Hospitals™ winners, the latter of which were classified according to the number of years in which they received this designation:

• All winners
• Winners of two or more years
• Winners of three or more years
• Winners of four or more years

The outcome measure for this analysis was the implantation of a coronary artery stent in a PTCA patient. PTCA and stent placement were identified using International Classification of Diseases, 9^th Revision, Clinical Modification (ICD-9-CM) procedure codes.

The data: This analysis uses calendar year 1998 International Classification of Clinical Services (ICCS) data to analyze the use of ReoPro, Integrilin, and Aggrastat among patients undergoing PTCAs. We used this source because it contains detailed inpatient drug data.

The comparison groups: 100 Top Hospitals™ were divided into two groups — those hospitals that had won this award in three or more years, and those that had won the award in one or two years. This division was made to see whether multi-year winners performed differently from those who had won the award once or twice. These two 100 Top Hospitals™ groups were compared against a reference group of hospitals that had fallen into the lower half in the rankings of the 1999 100 Top Hospitals™ study.

Patients within each hospital group were identified as having undergone a PTCA with or without stent based on ICD-9-CM procedure codes. Patients within these groups were further classified based on their principal ICD-9-CM diagnosis codes.

The percentage of patients receiving ReoPro was identified within each of these subgroups using ICCS data.