Abstract
INTRODUCTION: Water distribution systems in hotels have been related to outbreaks caused by Legionella spp. Certain measures, including disinfection by chlorination, maintaining increased temperatures are usually undertaken to prevent Legionella outbreaks. However, these preventive strategies are not always effective, since there are several factors (e.g., synergistic interactions with other microbes, physico-chemical factors, biofilm formation, availability of nutrients) that promote survival and proliferation of the pathogen in water pipes., Accordingly, there is a need of a holistic approach in development of preventive models for Legionella outbreaks associated with water distribution systems. METHODS: Water samples were collected from hotel water systems and were tested for the presence of Legionella, E. coli, total coliforms, total mesophilic count and Pseudomonas. In each sample, temperature and chlorine were also tested. Other epidemiological factors were additionally recorded including number of rooms, stars, proximity of sampling point to the boiler, etc. Data were processed by generalized linear analysis, and modeling based on logistic regression analysis to identify independent predictive factors associated with the presence of Legionella in hotel water systems. RESULTS: According to the generalized linear model, temperature affected (p<0.05) the presence of Legionella regardless of the species or the water supply (hot or cold). Additionally, opportunistic (P. aeruginosa) or non-opportunistic (E. coli, coliforms) pathogens were significantly associated (p<0.05) with the presence of all Legionella species. Temperature also exhibited a positive effect to all pathogens tested except for Pseudomonas according to the linear model. Multivariate analysis showed that Pseudomonas, total coliforms, HPC and temperature had a statistically significant effect on the presence of Legionella. Based on a binomial model, cold water had a positive effect on Legionella. Type of sampling and proximity of the sample to the boiler seemed to pose different effect on Legionella depending on the cfu/L. The number of hotel stars and rooms did not appear to have any effect in all tested models. DISCUSSION: Collectively, these results indicate the need for development of individualized water safety plans tailored by the presence of other microbiological agents, and unique physico-chemical factors, which could facilitate the survival of Legionella.in hotel water systems.