Validity of the endotoxin assay in post surgical patients.

In a prospective study 56 patients were studied after major surgery for 10 days to correlate endotoxemia with bacteremia and clinical septicemia. Endotoxemia was detected with a quantitative plasma endotoxin assay using limulus lysate and a chromogenic substrate. Clinical septicemia was defined using a set of predetermined criteria. In 29 patients both bacterial cultures and endotoxin assays were negative. One patient showed clinical signs of septicemia. In 14 patients, low grade endotoxemia (mean 47 +/- 15 pg/ml) was observed in the first three days after surgery. Only three of these patients also showed positive endotoxin tests after this period. Corresponding blood cultures were frequently negative (81%) most likely due to prior and concurrent antibiotic prophylaxis. Two endotoxin positive patients had clinical signs of septicemia. In a total of 16 patients, including two patients from the former group endotoxemia (mean 56 +/- 21 pg/ml) was noted beyond the third post-operative day. Samples for blood cultures collected simultaneously were positive for gram negative bacilli in six patients. Clinical septicemia was present in 12 of the 16 patients. The positive and negative predictive value of the endotoxin test for bacteremia and/or septicemia after the third day of surgery for all patients was found to be 75% and 98% respectively. The results of this small prospective study in surgical patients suggest that this rapid endotoxin test is a useful clinical adjunct for both the detection and exclusion of gram negative bacteremia/septicemia.