Assessment of severity of ICU-acquired pneumonia and association with etiology.

Di Pasquale, Marta; Ferrer, Miquel; Esperatti, Mariano; Crisafulli, Ernesto; Giunta, Valeria; Li Bassi, Gianluigi; Rinaudo, Mariano; Blasi, Francesco; Niederman, Michael; Torres, Antoni

OBJECTIVES We evaluated the association between severity of illness and microbial etiology of ICU-acquired pneumonia to define if severity should be used to guide empiric antibiotic choices. DESIGN Prospective observational study. SETTING ICUs of a university hospital. PATIENTS Three hundredy forty-three consecutive patients with ICU-acquired pneumonia clustered, according to the presence of multidrug resistant pathogens. INTERVENTIONS None. MEASUREMENTS AND MAIN RESULTS Two hundred eight patients had ventilator-associated pneumonia and 135 had nonventilator ICU-acquired pneumonia. We determined etiology in 217 patients (63%). The most frequent pathogens were Pseudomonas aeruginosa, Enterobacteriaceae, and methicillin-sensitive and methicillin-resistant Staphylococcus aureus. Fifty-eight patients (17%) had a multidrug-resistant causative agent. Except for a longer ICU stay and a higher rate of microbial persistence at the end of the treatment in the multidrug-resistant group, no differences were found in clinical and inflammatory characteristics, severity criteria, and mortality or survival between patients with and without multidrug-resistant pathogens, even after adjusting for potential confounders. Patients with higher severity scores (Acute Physiology and Chronic Health Evaluation II and Sepsis-related Organ Failure Assessment) and septic shock at onset of pneumonia had significantly lower 28- and 90-day survival and higher systemic inflammatory response. The results were similar when only patients with microbial diagnosis were considered, as well as when stratified into ventilator-associated pneumonia and nonventilator ICU-acquired pneumonia. CONCLUSIONS In patients with ICU-acquired pneumonia, severity of illness seems not to affect etiology. Risk factors for multidrug resistant, but not severity of illness, should be taken into account in selecting empiric antimicrobial treatment.