Quelle surveillance des infections nosocomiales en réanimation ?

During the last 20 years, the landscape of healthcare-associated surveillance in the intensive care unit has profoundly changed. Principles of surveillance have been established by the Centers for Diseases Control (CDC), with rates expressed as incidence density, use of results for feedback, and for designing and evaluate infection control programs. Most national surveillance networks use similar methods, but there is no available study evaluating cost-effectiveness of this surveillance strategy. Infection rates, however, have decreased in most surveillance systems, including French ones. Surveillance of healthcare-associated infections (HAI) in the ICU setting is mainly useful for local purpose. Using infection rates to compare performance between units and in some countries to publicly disclose results is appealing. There are, however, uncertainties on detection and diagnosis of HAI, adjustment on comorbidities and surveillance accuracy. These limitations impede any comparisons and benchmarking between units. Process indicators are increasingly recognized as useful, with impact on HAI rates, especially if used as bundle measures in a multifaceted control program. Surveillance must stay the cornerstone of infection control program. Its extent and complexity, however, are debated, its interest being counterbalanced with process indicators, in constrained healthcare system budget and personnel.