[Epidemiology of antibiotic resistance: implications for empirical treatment in intensive care]

The adequacy of initial antibiotic therapy is an important prognostic factor in severe infections. Concerning nosocomial infections, the selection of appropriate empirical therapy should take into account the incidence of offending pathogens within a specific unit. The changing trends in the hospital's microbial resistance patterns should be known to the physicians. The bacteria involved and the susceptibility testing vary widely among institutions and among countries. Many risk factors for acquisition of resistant pathogens have been identified. The duration of stay in hospital, previous colonization, and antibiotic treatment are the most frequently cited risk factors. When P. aeruginosa is suspected, an ureido-penicillin/aminoglycoside combination is usually effective. However, in some units with high levels of resistance, the beta-lactam should be ceftazidime or imipenem. When enterobacteria are suspected, a third generation cephalosporin, alone or in combination with an amino-glycoside or a broad spectrum penicillin associated with a beta-lactamase inhibitor is appropriate. Early nosocomial staphylococcal infections are treated with nafcillin or oxacillin, alone or in combination with an aminoglycoside. In units with a high rate of MRSA, the initial antibiotic therapy should include a glycopeptide.