Assessment of severity of intensive care unit-acquired pneumonia and association with etiology : a prospective study

Rationale: The current 2005 guidelines on hospital-acquired pneumonia do not take into account the severity of disease in the empiric antibiotic therapy algorithm, as it was in the former 1996 guidelines. Moreover, studies have shown that the 2005 guidelines predict potentially multi-drug resistant (MDR) microorganisms less accurately than the 1996 guidelines. We therefore evaluated the correlation between the severity of illness and the microbial etiology of Intensive care unit (ICU)-acquired pneumonia (ICUAP). Methods: Prospective, observational study in 6 medical and surgical ICUs with a total of 45 beds, from January-2007 to December-2011. We assessed the characteristics, microbiology, systemic inflammatory response and outcomes of 343 consecutive patients with ICUAP clustered according to the presence or not of MDR pathogens. Results: 208 patients had ventilator associated pneumonia (VAP) and 135 patients had non-ventilator ICUAP, of whom 90 needed subsequent intubation. 47% of the study population had septic shock at the onset of pneumonia and 38% developed it afterward. In 217 (63%) patients we determined etiology. The most frequently isolated pathogens were P. aeruginosa, Enterobacteriaceae, methicillin-sensitive (MSSA) and methicillin-resistant (MRSA) S. aureus. 71 patients had early-onset pneumonia and no risk factor for MDR pathogens, while 272 had late-onset pneumonia and/or at least one risk factor. No differences were found in terms of severity scores (APACHE-II, SAPS-II, SOFA), clinical and inflammatory characteristics between these groups. 58 patients had a MDR causative agent. No differences were found in any clinical and inflammatory characteristic and severity criterion between patients with or without MDR, except for a trend to higher frequency of VAP (72% vs. 58%, p=0.063) in the MDR group. Patients with higher APACHE-II and SOFA scores and septic shock at the onset of pneumonia had a significantly lower survival rate both at 28 and 90-day, in presence of a high rate of initial adequate treatment. Conclusions: in patients affected by ICUAP severity of illness affects mortality but does not seem to affect etiology, particularly the presence of MDR pathogens.