Diagnostic performance of Pneumonia multiplex PCR in critically ill immunocompromised patients.

backgroundAdmissions of immunocompromised patients to intensive care units (ICUs) are on the increase. The main reason for admission is acute respiratory failure, predominantly of infectious origin. In such circumstances, early and appropriate antibiotic therapy guarantees a better prognosis. Rapid diagnostic techniques such as multiplex polymerase chain reaction (PCR) have shown their value in both diagnosis and treatment in immunocompetent patients. To date, little data are available on immunocompromised patients.

methodsIn this retrospective, single-center study, we analyzed data from critically ill immunocompromised patients admitted for acute respiratory failure requiring invasive ventilation, in whom a respiratory specimen was taken and processed simultaneously by BioFire FilmArray Pneumonia Panel multiplex PCR (BFPPm PCR) and conventional culture (CC). Samples had to be taken from deep respiratory tracts less than 48 h after mechanical ventilation. The primary endpoint was the evaluation of the diagnostic performance of BFPP mPCR compared with CC. The secondary endpoint was the therapeutic impact of the results of BFPP mPCR.

resultsOne hundred and fourteen patients were included, with immunosuppression mainly of a hematological (35.1%) and oncological (35.1%) nature. The mPCR positivity rate was 36.8%, with the majority identifying enterobacteria (51%) and a median turnaround time of between 2h30 and 4 h. Comparison of rapid techniques with CC showed sensitivity of 89%, specificity of 83%, predictive positive value of 52% and negative predictive value of 98%. Concordance between the two techniques was complete in 84.2% of cases. mPCR enabled antibiotic therapy to be modified in 17.5% of cases, mainly de-escalation.

conclusionThe use of mPCR in the diagnosis of pneumonia in immunocompromised patients shortens the time required to obtain results, and is particularly effective in eliminating the presence of multi-resistant germs. Bacteria detected in culture and not included in the mPCR spectrum were mostly bacteria of low pathogenicity or sensitive to the antibiotics usually prescribed. The mPCR technique could reduce exposure to broad-spectrum antibiotics in this population.