Semiquantitative culture in diagnosing venous catheter-related sepsis.

Since Aubaniac (1) described the puncture of the subclavian vein in 1952, and specially after the standardization of parenteral nutrition by Dudrick et al. (11) in 1968, much has been published about complications caused by percutaneous central venous catheterization. Among the various complications provoked by this procedure, a very important one is "primary sepsis" or "catheter-related sepsis", both because of its frequency and because of the morbidity and mortality it causes (18,19). It is, however, difficult to diagnose this complication. The main difficulty lies in differentiating catheters that are really causing sepsis from those that, though showing "positive culture" do not cause bacteremia and are not responsible for the occasional signs of infection that a patient may show (6,7). This difficulty in diagnosing has led to the recommendation that all catheters suspected of causing sepsis be systematically removed. This procedure has the effect of exposing patients in serious condition and with limited venous access to the risks of new punctures. Usually these risks are unnecessary, since 75 to 90% of the catheters removed for this reason are not the real source of infection (3, 17, 19, 21, 22). In 1977, Maki et al. (18) proposed a semiquantitative catheter tip culture that showed considerable correlation with positive hemoculture for the same microorganisms; that is, capable of identifying which "positive catheters" were really causing sepsis. Subsequent research confirmed these results, showing that the semiquantitative catheter tip culture had specificity and sensibility over 80% (10, 15).(ABSTRACT TRUNCATED AT 250 WORDS)