Bacteriological monitoring of drinking water has not changed essentially during the last decennium. Escherichia coli and fecal coliforms are still traditionally used as index organisms, although methods are slightly different in various countries.

Investigations carried out by the authors have demonstrated that conventional methods can still be criticized in two respects. The spontaneous dying off of the Salmonella in water often occurs at a considerably lower rate than that of E. coli, which can lead to falsely negative results. In addition, investigations on human enteric flora composition, using newer methodology, have demonstrated that coli-aerogenes bacteria other than E. coli occur in relatively high concentrations. Two approaches will allow the elimination of these deficiencies of conventional monitoring of drinking water. They are using the entire group of Enterobacteriaceae as index organisms or considerably increasing the sensitivity of testing for E. coli.

For the latter purpose, an instrumental, continuous method has been developed. Water samples are first passed through a hollow fiber ultrafiltration device. Thereupon enrichment is carried out by culturing in lactose broth at 41°C. After stripping off the carbon dioxide (CO₂) formed by fermentation, the liquid is mixed with a glutamate solution buffered at pH 3.9, the CO₂ formed set is freed by acidification, and then absorbed in a phenolphtalein buffer which is ultimately examined colorometrically. The sensitivity of this method is 1 colony forming unit (cfu) of E. coli/100 ml or better. It has the additional advantage of allowing much more representative sampling of a water supply than customarily carried out.