Evaluation of MUG-supplemented media for the detection of E. coli in recreational water surveillance.

Four media containing 4-methylumbelliferyl-beta-D-glucuronide were evaluated as a non-confirmatory procedure for E. coli detection in recreational water surveillance. The media included ECD-Agar for membrane filtration and laurylsulphate-tryptose, brilliant-green-bile and lactose as broth media in a three tube most probable number procedure. From six representative water sites, samples were collected weekly over a typical summer season (17.05-27.09.1994) and processed as parallels, using each media at two different incubation temperatures (36 degrees/44 degrees C). Results showed that incubation temperature had no impact on E. coli counts. Each media at a given temperature could be regarded as individual enrichment procedure. None of these enrichment procedures showed a constant and predictable higher sensitivity during the sampling period at all sites compared to the others tested. For parallel results, the rate of agreement, based upon EC-guideline (76/160/EWG) staging of recreational water quality, was 85% for membrane filtration and 75% for the MPN-procedure results. Marked differences could be observed in false-positive specificity showing correlation to the selective characteristics of the media. Subsequently lactose-broth at 44 degrees C performed worst with 30% non verifiable results, while ECD-agar and laurysulphate-tryptose-broth, both at 44 degrees C, had a nearly 100% confirmation rate. Thus, combining high specificity with no lack in sensitivity these two MUG-supplemented media seem to be best suited for E. coli detection in routine recreational water surveillance.