Comparison of two culture methods for the enumeration of Legionella pneumophila from potable water samples

• Published in: Journal of water and health Vol. 19; no. 3; pp. 468 - 477
• Main Authors: Boczek, Laura A., Tang, Min, Formal, Casey, Lytle, Darren, Ryu, Hodon
• Format: Journal Article
• Language: English
• Published: London IWA Publishing 01.06.2021
• Subjects: Agglutination tests; bcye agar; Charcoal; Climate change; Drinking water; Enumeration; Laboratories; legiolert; Legionella; Legionella pneumophila; Legionnaires disease; Legionnaires' disease bacterium; Negative wells; Organisms; Plumbing; potable water; simulated home plumbing system; Specificity; Statistical methods; Water analysis; Water pollution; Water sampling; Water shortages; Yeasts; United States > US
• ISSN: 1477-8920; 1996-7829
• DOI: 10.2166/wh.2021.051

Legionella infections have steadily increased in the United States over the last 20 years, and most of these infections have been attributed to contaminated water. The gold standard for confirmation of Legionella presence in water is culturing with Buffered Charcoal Yeast Extract (BCYE) agar. Following many modifications, this method is still time-consuming, expensive, and can take longer than 10 days for full confirmation. The Legiolert is a newer and simpler culture product that is claimed to be able to quantify Legionella pneumophila in 7 days with high sensitivity and specificity and does not need further confirmation for the presence of L. pneumophila. This study compared the culturability of L. pneumophila occurring in a simulated home plumbing system using both Legiolert and BCYE agar methods. Out of 185 water samples, Legiolert and BCYE method detected L. pneumophila in 83 and 85% of the samples, respectively. The two methods were determined to be statistically equivalent for culturability of L. pneumophila, though the detected levels by Legiolert were slightly higher than the BCYE method. The molecular confirmation of positive (n = 254) and negative wells (n = 82) with Legiolert also showed a high specificity of 96.5% (i.e., 3.5% false positives (9/254) and 0% false negatives (0/82)).