Improving the Diagnosis of Bloodstream Infections: PCR Coupled with Mass Spectrometry

Jordana-Lluch, Elena; Giménez, Montserrat; Quesada, M. Dolores; Ausina, Vicente; Martró, Elisa

In a number of cases in both studies, PCR/ESI-MS detected microorganisms in whole blood specimens with a paired negative blood culture. The clinical relevance of the additionally detected microorganisms was investigated through clinical records review in order to discriminate between probable contaminants and true pathogens. The proportions of detected microorganisms with clinical significance not isolated by conventional methods were 7.5% ( The system gathers all the information obtained from each primer pair 172 blood culture negative cases) [21] and 7.2% (31 out of 431 blood culture negative cases) [22] . These findings are highly relevant, as conventional methods were not able to diagnose the etiology of infection in the culture-negative patients. The sensitivity of the system was calculated using different approaches in each study. Jordana-Lluch et al. disregarded those specimens with a polymicrobial identification by either or both methods, as the events with one correct detection but with a disagreement in the second one were difficult to catalogue as "true positive" or "false positive. " In those terms, the sensitivity of the PCR/ESI-MS was 50.0%. Laffler et al. performed a theoretical approximation of the sensitivity based on the historical blood culture positivity rate in their center. They extrapolated the experimentally obtained PCR/ESI-MS positivity rate in order to obtain the number of negative blood cultures that, if processed by the PCR/ESI-MS, would have additionally tested positive. The estimated sensitivity of PCR/ESI-MS using this theoretical approach was 85.9%. This extrapolation may have led to a biased estimation of the sensitivity.