Risk and detection of pulmonary artery catheter- related infection in septic surgical patients

Specimen cultures were evaluated in 49 catheterized patients who had a known focus of infection (primarily intra-abdominal peritonitis). Bacteria were recovered from 2% of flush solutions, 14% of transducer domes, 18% of diaphragms, and 24% of cardiac output fluids; however, these bacteria were not found in cultures of the pulmonary artery (PA) catheter segments. The rates of positive PA catheter-aspirate cultures were 30.6% on day 1, 20.4% on day 2, and 32.7% on day 3 (not statistically different). PA catheter-aspirate cultures had a sensitivity of 5.7% and a positive predictive value of 30% for catheter-related infection, and 15% sensitivity and 40% positive predictive value for peripheral bacteremia. While 95% (55 of 58) of the catheter-aspirate cultures were false-positives, only 0.5% (3 of 588) were true-positives. Peripheral blood cultures were positive in 10% of the patients, but the catheter segments were sterile or grew different organisms. Arterial line cultures had zero sensitivity and predictive value to detect catheter-related infection, and 15% sensitivity and 40% predictive value to detect peripheral bacteremia. Thus, PA catheter-aspirate cultures, routine peripheral blood cultures, and arterial cultures cannot be recommended to detect PA catheter-related infection. Catheter-related infection confirmed by catheter-segment cultures was 10.2% when the PA catheters were removed after 73 +/- 6.5 (SD) h. Bacteria from catheter-segment cultures corresponded to those from the primary infection site.