Interlaboratory comparison of culture- and PCR-based methods for Legionella pneumophila detection in drinking water samples

The European Directive 2020/2184 concerning the quality of water for human consumption now includes Legionella among the pathogens for assessment in domestic water systems. It states that “for risk-based verification and to complement spread-plate culture methods, rapid culture methods, non-culture-based methods, and molecular-based methods may be used.” In this study, 33 laboratories across Italy analyzed a number of unique water samples ranging from 10 to 30 for the presence of Legionella pneumophila. All laboratories used the standard spread-plate culture method, 31 laboratories also performed the Legiolert rapid liquid culture method (IDEXX Laboratories), and 27 out of 33 performed the DI-Check Legionella pneumophila real-time PCR method (Diatheva). In all, 23 laboratories executed all three methods. Data generated from 817 samples were collected and statistically analyzed. The Legiolert method allowed analysis with a smaller sample volume (100 mL and 10 mL), compared to the standard culture method with which it was shown to be comparable with K agreement values of 0.785 and 0.840 in the two mentioned volumes, respectively. The standard real-time PCR method was more sensitive (93%) than the spread-plate culture method. Sensitivity values of 95.2% and 98.8% were also obtained by comparing two new real-time PCR procedures with the spread-plate culture method, tested to shorten the analysis, and the standard culture method. Finally, data obtained from the analysis of drinking water samples with the spread-plate culture method using both Buffered charcoal yeast extract and non-selective glycine vancomycin polymyxin cycloheximide media showed the greater capacity of the latter in the recovery of Legionella (P < 0.0001).