Comparison of endotoxin (LPS) measurements and plate counts for bacteriological assessment of drinking water.

Plate counts were compared with endotoxin (LPS) concentrations in 229 samples of unchlorinated drinking water. Plate counts were obtained on two different media (Kings agar B and R2A), and different fractions of LPS (total, free, and bound) were determined using a chromogenic LAL test kit. The closest correlation was found when total LPS was compared with R2A plate counts (r = 0.60), but based on regression analysis this relation was shown to be insufficient for practical use. Possible explanations for this discrepancy between plate counts and LPS are discussed, and it is concluded that the use of LPS as a surrogate indicator in drinking water bacteriology should not be rejected solely on the basis of a poor correlation with plate counts.