You are here:
Comparison of two methods for the enumeration of Legionella pneumophila from potable water samples
Citation:
Boczek, L., K. Carlson, H. Woo, D. Lytle, M. Rodgers, AND H. Ryu. Comparison of two methods for the enumeration of Legionella pneumophila from potable water samples. Presented at WQTC 2018, Toronto, Ontario, CANADA, November 11 - 15, 2018.
Impact/Purpose:
Legionella pneumophila is the causative agent of the human respiratory disease legionellosis and is responsible for a significant percentage of waterborne disease outbreaks in the US. These outbreaks are frequently associated with the premise plumbing systems of large buildings. The availability of better detection methods for these bacteria is needed to aid facility managers in implementing building water management plans. Culturing techniques for the isolation of Legionella, such as using Buffered Charcoal Yeast Extract (BCYE) Agar), have been developed and improved upon since the late 1970’s. However, these methods can be labor intensive, prone to overgrowth by non-target bacteria that obscure the Legionella colonies, and require additional analysis to determine whether the Legionella recovered are L. pneumophila. A newer method introduced by IDEXX® called Legiolert has recently been introduced that promises to resist the growth of non-target bacteria while specifically recovering and quantifying L. pneumophila from water samples. In this study, we investigated the Legiolert method by comparing its performance using environmental water with the standard culture method.
Description:
One-liter water samples were collected from a simulated home plumbing system in sterile polypropylene bottles containing sodium thiosulfate. A portion of each sample was analyzed using Legiolert according to the manufacturer instructions (10 ml of sample added to 90 ml of diluent) and the remainder of the sample was concentrated using membrane filtration, and 0.1 ml was plated onto BCYE agar plates containing antibiotics. In this way, both methods analyzed the equivalent of 10 mL sample volume. Isolates from BCYE plates were confirmed as being Legionella by streaking onto BCYE plates without L-cysteine, and then further identified as L. pneumophila using latex agglutination tests specific for L. pneumophila. Legiolert samples were further screened using L. pneumophila specific quantitative PCR on randomly selected wells from the Legiolert trays. The Pearson correlation and regression analyses were used for the parallel comparison of L. pneumophila levels enumerated by BCYE agar count and Legiolert methods. In addition, a McNemar’s binomial analysis was used to compare these methods when used to determine only the presence or absence of L. pneumophila in the samples. A total of 204 samples were analyzed and of those, 163 (80%) were positive for presence of L. pneumophila using both methods. The BCYE culture method identified 10 (4.9%) samples as L. pneumophila positive however these were negative using Legiolert. Conversely, 20 (9.8%) of the samples were positive only by Legiolert. The McNemar’s statistics showed no statistical difference (P=0.1), indicating both methods are equally sensitive for determining the prevalence of L. pneumophila. Of the 20 samples positive by Legiolert only, 15 samples resulted in failing to identify any Legionella using the BCYE method due to the overgrowth of non-target microorganisms. The log transformed quantities determined by both methods showed a high cross-correlation (Pearson’s r of 0.9149), and the slope of the regression equation was near one. The Legiolert method had a relatively high specificity (i.e., 3.5% false positives from 254 positive wells and 0% false negatives from 82 negative wells) which is comparable to other published studies. In addition, there was no evidence of interference by non-target microorganisms when using the Legiolert method. Altogether, the new Legiolert method performed as well or better than the standard agar-based method in qualitative and quantitative detection of L. pneumophila in premise plumbing water samples.