Objective: To assess whether combination antibiotic therapy improves outcome of severe community-acquired pneumonia in the subset of patients with shock.
Design: Secondary analysis of a prospective observational, cohort study.
Setting: Thirty-three intensive care units (ICUs) in Spain.
Patients: Patients were 529 adults with community-acquired pneumonia requiring ICU admission.
Interventions: None.
Measurement and main results: Two hundred and seventy (51%) patients required vasoactive drugs and were categorized as having shock. The effects of combination antibiotic therapy and monotherapy on survival were compared using univariate analysis and a Cox regression model. The adjusted 28-day in-ICU mortality was similar (p = .99) for combination antibiotic therapy and monotherapy in the absence of shock. However, in patients with shock, combination antibiotic therapy was associated with significantly higher adjusted 28-day in-ICU survival (hazard ratio, 1.69; 95% confidence interval, 1.09-2.60; p = .01) in a Cox hazard regression model. Even when monotherapy was appropriate, it achieved a lower 28-day in-ICU survival than an adequate antibiotic combination (hazard ratio, 1.64; 95% confidence interval, 1.01-2.64).
Conclusions: Combination antibiotic therapy does not seem to increase ICU survival in all patients with severe community-acquired pneumonia. However, in the subset of patients with shock, combination antibiotic therapy improves survival rates.