Accuracy of diagnostic tests for Legionnaires' disease: a systematic review

Purpose. Rapid and effective diagnosis of Legionnaires' disease (LD) cases is extremely important so that timely and appropriate therapy can be provided, thereby lowering the morbidity and mortality rates and reducing the health and economic costs associated with this disease. Methodology. Diagnosis is established solely by microbiological tests. There are several methods available, each with different performance, sensitivity and specificity characteristics, and further understanding is required. Our objective was to assess the accuracy of urinary antigen detection, direct fluorescent antibody (DFA) staining, serological testing and the polymerase chain reaction (PCR) method versus culture analysis (the reference standard) in patients suspected of being infected with Legionella or patients with laboratory-confirmed LD. We performed a MEDLINE search in November 2014. Two authors independently assessed the trials and extracted data. Pooled analysis was performed through Meta-DiSc version 1.4. Result. The inclusion criteria were met by 11 studies. All the studies evaluated PCR and DFA tests to detect Legionella in clinical specimens, comparing them to culture techniques, and were included in the meta-analysis. The pooled sensitivity and specificity for PCR were 83 % [95% confidence interval (CI): 79-87 %] and 90 % (95% CI: 88-92 %), respectively. DFA was evaluated in one study and the sensitivity and specificity of this test were 67 % (95 % CI: 30-93 %) and 100 % (95% CI: 91100 %), respectively. PCR had high sensitivity and specificity for early diagnosis of LD. Conclusion. Culture analysis is deemed necessary for epidemiological studies, molecular strain typing and antibiotic sensibility evaluations; however, the performance of PCR in recent studies calls for additional, well-designed studies in order to achieve the best standard test, which will enable optimization of the Legionella infection diagnostic.