Objective: To confirm the sensitivity of the polymerase chain reaction (PCR) technique (versus blood cultures) and to gain a better understanding of the incidence of true- and false-positive results when using this technique.
Design: Observational study.
Setting: Fourteen-bed, level 3 intensive care unit.
Patients: Hundred twenty-six critically ill adult patients. Hundred ninety-seven blood culture and PCR samples taken as clinically indicated for suspected sepsis, according to routine ICU protocol.
Measurements and results: The PCR product (16SrDNA: 341F-1195R) was sequenced and compared with a database of known species (Genebank) to identify the bacterial nucleic acid. The PCR or blood culture result was classified as a true-positive if there was other microbiological or clinical supporting evidence.