The detection and interpretation of endotoxaemia.

A considerable body of evidence has accumulated that implicates endotoxin in the pathogenesis of the sepsis syndrome. This has raised interest in the possibility of measuring endotoxin as a surrogate marker of Gram negative infection, particularly since conventional microbiological tests have an inevitable delay. The Limulus amoebocyte lysate (LAL) assay has been used most widely to measure endotoxin in clinical samples. However, there are several important limitations that need to be borne in mind. These include the dangers of contamination, lack of precision and accuracy, and both false positive and false negative results. Endotoxaemia is present in the blood of about 30% of patients with bacteraemia, but endotoxaemia does not predict either Gram negative bacteraemia or Gram negative infection, nor does it predict survival from sepsis. There is some correlation with severity of sepsis, but the level of precision is poor. At the present time, there is no place for routine endotoxin testing in clinical practice. In particular, the positive predictive value of the test is insufficiently high to be of clinical use. It may be that the LAL assay, or one of the newer developments, may be more useful in excluding Gram negative infection, but that remains to be shown.