A total of 248 patients (27.8%) died during their hospitalization. Patients with a microbiologically confirmed infection had a statistically greater hospital mortality rate compared to patients without infection (35.5% vs 22.3%, respectively; p < 0.001). Kaplan-Meier plots of the probability of remaining alive during hospitalization for patients with and without microbiologically confirmed infection. Patients with microbiologically confirmed infection had a statistically higher probability of in-hospital mortality compared to patients without evidence of infection (p < 0.001 [log rank test]). Among the baseline characteristics examined, hospital nonsurvivors were statistically older, had greater APACHE II scores, and had a higher Viagra for sale Australia prevalence of underlying malignancy, chronic renal insufficiency, the need for dialysis, and cirrhosis. Hospital nonsurvivors required significantly more central lines, were more likely to receive enteral nutrition, mechanical ventilation, vasopressors, and sucralfate, and to require reintubation. The development of new infiltrates seen on chest radiographs, purulent sputum, temperature > 38.3°C, severe sepsis, primary bloodstream infection with a central vein catheter in place, secondary bloodstream infection, ventilator-associated pneumonia, urinary tract infection, skin or soft-tissue infection, multiple infections, culture positivity for vancomycin-resistant enterococci, and the presence of a microbiologically confirmed infection were statistically more common among hospital nonsurvivors.

Multiple logistic regression analysis demonstrated that underlying malignancy (AOR, 1.90; 95% CI, 1.46 to 2.50; p = 0.015), chronic renal insufficiency (AOR, 2.30; 95% CI, 1.89 to 2.80; p < 0.001), cirrhosis (AOR, 2.89; 95% CI, 2.16 to 3.86; p < 0.001), APACHE II score [AOR, 1.05; 95% CI, 1.04 to 1.07; p < 0.001], mechanical ventilation (AOR, 1.78; 95% CI, 1.39 to 2.27; p = 0.018); temperature > 38.3°C (AOR, 1.50; 95% CI, 1.24 to 1.81; p = 0.035), skin and soft-tissue infection (AOR, 10.71; 95% CI, 3.52 to 32.59; p = 0.033), and therapy with vasopressors (AOR, 3.08; 95% CI, 2.53 to 3.76; p < 0.001) were independently associated with hospital mortality. Variables independently associated with hospital mortality for patients with APACHE II scores of < 20, between 20 and 25, and > 25. Cirrhosis and the requirement for vasopressors were the only variables associated with mortality for all groups of severity of illness examined.