Predicting Legionella contamination in cooling towers and evaporative condensers from microbiological and physicochemical parameters.

Background: Inhalation of Legionella-containing aerosols generated by cooling towers (CT) and evaporative condensers (EC) where water risk management is not performed correctly has been linked to a high percentage of community outbreaks of Legionnaires' disease (LD). Likewise, microbiological and physicochemical characteristics of the water in these facilities have been associated with this bacterium. The main aim of this study was to assess the risk of Legionella colonization in CT and EC based on the data for microbiological and physicochemical water quality provided by the Environmental Health Department and Laboratory of the City Council of L'Hospitalet de Llobregat (Barcelona, Spain).

Methods: Legionella was analysed in 789 samples collected from 127 CT and EC in 46 companies in Catalonia from 2002 to 2019. A two-step logistic regression analysis was carried out to assess the risk of colonization by Legionella in the studied facilities according to the microbiological (aerobic heterotrophic bacteria) and physicochemical (pH, alkalinity, hardness, turbidity, conductivity, total iron and Langelier Index) water parameters. The optimal cut-off points for the water parameters predictive of Legionella contamination were defined as the values on the receiver operating characteristic (ROC) curve where sensitivity and specificity were jointly maximized.

Results: Legionella was isolated in 8.49% of the 789 analysed samples, 22.39% of which were heavily contaminated (with counts higher than 1.0 × 10 CFU/l). L. pneumophila was isolated in 82.09% of the samples, with 41.82% belonging to serogroup 1. Logistic regression analysis revealed that aerobic heterotrophic bacteria concentrations ≥6.90 × 10 CFU/ml [Odds ratios (OR) (95% CI) = 3.56 (1.39-9.43), p = 0.01], a pH ≥ 8.70 [OR (95% CI) = 3.60 (1.34-10.09), p = 0.01], and water hardness ≥5.72 × 10 mg/l [OR (95% CI) = 6.30 (2.34-18.56), p < 0.001] were each independently associated with a higher risk of CT and EC colonization by Legionella.

Conclusions: The present study shows the importance of risk assessment for improving the control measures aimed at preventing or reducing Legionella populations in CT and EC, thus minimizing potential dangers for public health.