RESUMO
Faster respiratory pathogen detection and antibiotic resistance identification are important in critical care due to the severity of illness, significant prior antibiotic exposure and infection control implications. Our objective was to compare the performance of the commercial Unyvero P55 Pneumonia Cartridge (Curetis AG) with routine bacterial culture methods and in-house bacterial multiplex real-time PCR assays. Seventy-four bronchoalveolar lavage specimens from patients admitted to a Scottish intensive care unit (ICU) over a 33-month period were tested prospectively by routine culture and viral PCR and retrospectively by Unyvero P55 and in-house bacterial PCR. Sensitivity/specificity was 56.9%/58.5% and 63.2%/54.8% for the Unyvero P55 and in-house bacterial PCR panels respectively; sensitivity for in-panel targets was 63.5 and 83.7% respectively. Additional organisms were detected by Unyvero P55 and in-house bacterial PCR panels in 16.2% specimens. Antibiotics were changed on the basis of routine test results in 48.3% cases; of these, true-positive or true-negative results would have been obtained earlier by Unyvero P55 or in-house bacterial PCR panel in 15 (53.6%) and 17 (60.7%) cases respectively. However, a false-negative molecular test result may have been acted upon in six (21.4%) cases with either assay. Sensitivity/specificity of Unyvero P55 antibiotic resistance detection was 18.8%/94.9% respectively. Molecular testing identified a number of respiratory pathogens in this patient cohort that were not grown in culture, but resistance detection was not a reliable tool for faster antibiotic modification. In their current set-up, molecular tests may only have benefit as additional tests in the ICU pneumonia setting.