RESUMEN
In Italy no nosocomial infection surveillance database has been established despite the fact that a decrease of nosocomial infection rates was one of the priorities of the Italian National Health Plan 1998--2000. Heart surgery operations are the most frequent high risk procedures in western countries. Active surveillance was performed at the heart surgery wards of two Italian hospitals (Rome and Catania, Southern Italy) in accordance with the methods described for the National Nosocomial Infections Surveillance (NNIS) System of the USA. In both hospitals surgical site infections (SSIs) were the most frequently encountered type of nosocomial infections, accounting for 57.2% in Rome and 50% in Catania, and SSI rates in coronary artery bypass grafts with both chest and donor site incisions, calculated by risk index equal to 1, were above the 90th percentile for the NNIS System. The urinary catheter-associated urinary tract infection (UTI) rate (5.8%) in Catania exceeded the 90th percentile for the NNIS System, while the device-associated UTI (1.6%), bloodstream (4.1%) and pneumonia (8.0%) rates, from the hospital in Rome, did not. All device utilization ratios were lower than the 10th percentile for the NNIS System. Our study demonstrated that the NNIS methodology is applicable to Italian hospitals, although with some limitations mainly regarding the minimal surveillance duration required for significant interhospital comparison, and highlighted the need of a national comparison of surveillance data as benchmark.