Background: Legionnaires' disease (LD) is associated with high mortality rates and poses a diagnostic and therapeutic challenge. Use of the rapid urinary antigen test (UAT) has been linked to improved outcome. We examined the association between the method of diagnosis (UAT or culture) and various clinical and microbiological characteristics and outcome of LD.
Methods: Consecutive patients with pneumonia and confirmation of Legionella infection by a positive UAT and/or a positive culture admitted between the years 2006-2012 to a university hospital were retrospectively studied. Isolated L. pneumophila strains were subject to serogrouping, immunological subtyping and sequence-based typing. Variables associated with 30-day all-cause mortality were analyzed using logistic regression as well as cox regression.
Results: Seventy-two patients were eligible for mortality analyses (LD study group), of whom 15.5 % have died. Diagnosis based on positive L. pneumophila UAT as compared to positive culture (OR = 0.18, 95 % CI 0.03-0.98, p = 0.05) and administration of appropriate antibiotic therapy within 2 hospitalization days as compared to delayed therapy (OR = 0.16, 95 % CI 0.03-0.90, p = 0.04) were independently associated with reduced mortality. When controlling for intensive care unit (ICU) admissions, the method of diagnosis became non-significant. Survival analyses showed a significantly increased death risk for patients admitted to ICU compared to others (HR 12.90, 95 % CI 2.78-59.86, p = 0.001) and reduced risk for patients receiving appropriate antibiotic therapy within the first two admissions days compared to delayed therapy (HR 0.13, 95 % CI 0.04-0.05, p = 0.001). Legionella cultures were positive in 35 patients (including 29 patients from the LD study group), of whom 65.7 % were intubated and 37.1 % have died. Sequence type (ST) ST1 accounted for 50.0 % of the typed cases and ST1, OLDA/Oxford was the leading phenon (53.8 %). Mortality rate among patients in the LD study group infected with ST1 was 18.2 % compared to 42.9 % for non-ST1 genotypes (OR = 0.30, 95 % CI 0.05-1.91, p = 0.23).
Conclusions: The study confirms the importance of early administration of appropriate antibiotic therapy and at the same time highlights the complex associations of different diagnostic approaches with LD outcome. Infection with ST1 was not associated with increased mortality. Genotype effects on outcome mandate examination in larger cohorts.