Quantitative blood cultures for catheter-associated infections.

The recent studies by Paya et al. (2) concluded that quantitative blood cultures were not useful in diagnosing intravascular-device-related bacteremia associated with arterial, central venous, and pulmonary artery catheters. While obviously more work needs to be done on the value of aspiration through a catheter in the diagnosis of catheterassociated sepsis, a number of studies have found this process useful (1), and I do not think that on the basis of the data presented in the study of Paya et al. one can with confidence conclude that the method is not useful. In seven cases (see Table 1 [2]), there appeared to be a 10-fold difference between the number of organisms isolated from blood drawn through the catheter and the number isolated from peripheral blood. In six cases, however, in both the blood drawn through the catheter and peripheral blood there was more than 100 CFU/ml, and these cases appeared to be the main basis for showing that there was no value in the method for differentiation of the site of sepsis. However, Paya et al. do not appear to have performed appropriate dilutions on these particular blood samples. One cannot therefore claim that there were no differences in the colony counts in these two samples of blood. For instance, if there were 105 CFU/ml in the blood drawn through the catheter and only 103 CFU/ml in the peripheral blood, this would clearly be more than a 10-fold difference. However, the method employed in their paper would not have been able to identify these differences. In addition, the definition of catheter-associated sepsis rested on the definition used in the study of Maki et al. Thus, a catheter is regarded as being positive only if more than 15 CFU could be detected on the catheter tip. However, a number of studies have shown that, particularly with central venous catheters, a cutoff lower than 15 CPU may be more appropriate (1). The use of a lower cutoff value may also have altered their conclusions. Thus, in summary, I agree with the conclusions of the authors in that the method as used by them with lysis centrifugation tubes was only of limited usefulness in the diagnosis of intravascular-device-related bacteremia. However, before the method can be stated to be totally devoid of value, a study needs to be performed in which appropriate dilution is performed such that when there are high numbers of organisms present, 10-fold-or-greater differences in numbers of organisms can be detected by the method in use. This does not appear to have been the case in this study.