Objective: To compare the outcome of patients with severe Legionella pneumonia (LP) according to the presence or absence of prognostic factors currently reported in the literature and delays in initiating fluoroquinolones and macrolides.
Design: Retrospective clinical investigation.
Setting: Intensive care unit (ICU) of an university hospital.
Patients: Forty-three consecutive cases with no previous treatment with a macrolide or a fluoroquinolone.
Measurements and main results: The 14 (33%) patients who died of LP were compared with the 29 survivors. Thirty-eight patients (88%) received a fluoroquinolone in combination with a macrolide agent, two patients erythromycin alone and three ofloxacin alone. In univariate analysis, SAPS II more than 46 ( p=0.006) and intubation requirement ( p=0.012) were associated with a higher mortality whereas the administration of fluoroquinolones ( p=0.011) or erythromycin ( p=0.044) within 8 h of arrival on the ICU was associated with better survival. By logistic regression analysis, SAPS II score more than 46 [odds ratio (OR) 8.69; 95% confidence interval (CI) 1.15-66.7; p=0.036], duration of symptoms prior to ICU admission longer than 5 days (OR 7.46; 95% CI 1.17-47.6) were independent risk factors for death. Fluoroquinolone administration within 8 h of ICU arrival (OR 0.16; 95% CI 0.03-0.96; p=0.035) was associated with a reduced mortality.
Conclusions: SAPS II score higher than 46, duration of symptoms prior to ICU admission longer than 5 days and intubation were associated with increased mortality. Initiation of fluoroquinolone therapy within 8 h of ICU admission significantly reduced mortality.