Laboratory diagnosis of catheter-related bacteremia

Abstract

Semiquantitative culture of the external surface of catheter tips has become the currently most used method to detect catheter colonisation and catheter-related bacteremia; however, this method may fail to detect significant colonisation of the internal lumen of catheters, and several quantitative methods have been described for this purpose. Although 15 and 1000 CFU are used to define positive catheters, the sensitivity and specificity are not yet well known, and the definitive cut-off level for a positive result remains to be established. We studied prospectively 91 intravascular catheter tips removed because a catheter-related infection was suspected. The result of external and internal cultures with different cut-off levels, alone or combined, were correlated with clinical data in 22 bacteremic episodes which occurred in these patients, 12 of which were catheter-related. The semiquantitative culture of the external surface alone with a cut-off level of 25 CFU provided the best means of detecting catheter-related bacteremia. This method has 100% sensitivity and 81.1% specificity. We conclude that semiquantitative culture technique of the catheter tip alone is sufficient and this method should be employed for diagnosing catheter-related bacteremia in unselected populations.