Comparative performance of a new method, Legiolert, vs. standard methods for the quantification of Legionella pneumophila in potable and nonpotable water samples

Poster-Vendor
Poster Presentation

Prepared by , D. Broder, T. Knight, A. Pednault, B. Swalla

Contact Information: veronica-newport@idexx.com; 207-556-4000


ABSTRACT

Legionella pneumophila is a Gram-negative bacterium that is commonly found in both potable and nonpotable water systems and can cause a severe pneumonia-type illness termed Legionnaires' Disease. A key step in mitigating L. pneumophila risk is to perform routine monitoring to quantify the presence of L. pneumophila, for which there are several methods in routine use internationally. In this study we compared the performance of standard plate-based culture methods defined by the CDC, Standard Methods, or ISO to Legiolert, a quantitative MPN-based culture method. Independent field trial data was generated in both North America and Germany. To compare a Legiolert 100mL potable water protocol to method ISO-11731-1/2 for examination of 100mL of potable water, 1604 German potable water samples were analyzed by 4 commercial laboratories yielding 290 yielded data pairs. Mean relative difference analysis as per ISO-17994 methodology revealed better sensitivity for Legiolert (+35. 3%) and the specificity was 96. 4% (1102/1143 positive wells confirmed by subculture). To compare a Legiolert 10mL potable protocol to method SM9260J for examination of 10-500 mL of potable water, 491 U. S. potable water samples were analyzed by 1 commercial laboratory yielding 74 data pairs. A two-tailed Wilcoxon Signed Rank test (used for non-parametric data) revealed higher sensitivity (Prob>|S| = < 0. 0001, cutoff value = 0. 05). A two-tailed t-test was also run and revealed no statistical difference. The specificity was 100. 0% (199/199 positive wells confirmed by subculture). To compare a Legiolert 0. 1 mL nonpotable protocol to both the CDC method "Procedures for Recovery of Legionella from the Environment" and method SM9260J for examination of 0. 1 mL of nonpotable water, a total of 1063 U. S. nonpotable water samples, predominantly from cooling towers, were analyzed by 2 commercial laboratories yielding 141 data pairs. A two-tailed Wilcoxon Signed Rank revealed no statistical difference(s) (Prob>|S| = 0. 2488 vs. the CDC method (n=92) and Prob>|S| = 0. 7278 vs. SM9260J (n=49)). A two-tailed t-test was also run and revealed no statistical difference vs. either standard method. The specificity was 95. 5% (378/396 positive wells confirmed by subculture). In summary, each of the test methods were either equivalent to or less sensitive than Legiolert for L. pneumophila quantification in these water matrices. Legiolert has a counting range of 1-2272 (MPN) compared to petri plates which are