Lack of correlation between Legionella colonization and microbial population quantification using heterotrophic plate count and adenosine triphosphate bioluminescence measurement.

This investigation compared biological quantification of potable and non-potable (cooling) water samples using pour plate heterotrophic plate count (HPC) methods and adenosine triphosphate (ATP) concentration measurement using bioluminescence. The relationship between these measurements and the presence of Legionella spp. was also examined. HPC for potable and non-potable water were cultured on R2A and PCA, respectively. Results indicated a strong correlation between HPC and ATP measurements in potable water (R = 0.90, p < 0.001). In the make-up water and two cooling towers, the correlations between ATP and HPC were much weaker but statistically significant (make-up water: R = 0.37, p = 0.005; cooling tower 1: R = 0.52, p < 0.001; cooling tower 2: R = 0.54, p < 0.001). For potable and non-potable samples, HPC exhibited higher measurement variability than ATP. However, ATP measurements showed higher microbial concentrations than HPC measurements. Following chlorination of the cooling towers, ATP measurements indicated very low bacterial concentrations (<10 colony-forming units (CFU)/mL) despite high HPC concentrations (>1000 CFU/mL) which consisted primarily of non-tuberculous mycobacteria. HPC concentrations have been suggested to be predictive of Legionella presence, although this has not been proven. Our evaluation showed that HPC or ATP demonstrated a fair predictive capacity for Legionella positivity in potable water (HPC: receiver operating characteristic (ROC) = 0.70; ATP: ROC = 0.78; p = 0.003). However, HPC or ATP correctly classified sites as positive only 64 and 62% of the time, respectively. No correlation between HPC or ATP and Legionella colonization in non-potable water samples was found (HPC: ROC = 0.28; ATP: ROC = 0.44; p = 0.193).