A number of methods are in common use for calculating 95th percentiles of enterococcal sampling results from recreational waters. However all are liable to give a poor indication of infection risk as calculated from the formula used in the 2003 WHO Guidelines for safe recreational water environments and similar risk based guidelines. Their limitations are particularly evident where the underlying microbial population deviates from the reference distributions used in these guidelines (lognormal, with a log10 standard distribution of ∼0.81).

A Microsoft^® Excel template (the EnteroTester) has been developed for generating workbooks that:

Standardised enterococcal 95th percentiles allow direct comparison between different recreational waters in terms of their infection risk. They overcome the common problem of misclassifying recreational waters to the wrong microbial water quality assessment category.

The template works for data sets containing eight observations or more, and can accommodate up to 80% of the results being below the limit of detection. It produces a narrower upper confidence band than the classic parametric method, thus requiring fewer samples than recommended in the above guidelines.

The template also generates trigger levels for alerting recreational water managers to likely changes in the underlying distribution, and the need to consider whether further sampling or investigation is required. It has been widely trialled in Australia, demonstrating its practical application in managing recreational waters.