PCR for the detection of invasive aspergillosis in bronchoalveolar lavage systematic review and meta-analysis

D. Yahav*, T. Avni, I. Levi, L. Leibovici, M. Paul

Author address: 

Petah-Tikva, IL)


Objectives: To assess the sensitivity and specificity of PCR in bronchoalveolar lavage (BAL) for the diagnosis of invasive aspergillosis (IA), and to examine the standards of PCR performance and their effects on the diagnostic accuracy of PCR. Methods: Systematic review. We included prospective, retrospective or casecontrol studies using PCR on BAL or bronchial secretions for the diagnosis of IA. We included only patients at risk for IA as defined by the EORTC host criteria (both cases and controls). Any PCR test was acceptable, including standard PCR, real-time, nested, multiplex or other PCR, using any primer gene and targeting all or specific aspergillus spp. We used only PCR tests taken within 14 days of infection onset. In the primary analysis we considered only proven and probable IA, defined by EORTC criteria. In a secondary analysis we also considered possible IA as disease. Risk of bias assessment was conducted using the QUADAS-2 tool. Coupled sensitivity and specificities were summarized using the bivariate model, from which hierarchical summary receiver operating curves (HSROC) were drawn. Covariates effects were investigated through meta-regression of diagnostic odds-ratios (DORs)

abstract No: 


Full conference title: 

22nd European Congress of Clinical Microbiology and Infectious Diseases
    • ECCMID 22nd (2012)