Determinants of Legionella spp. presence in hospital water surveillance: a five-year, three-centre study from Greece

Hospital water systems are persistent reservoirs of Legionella spp. and a recognised threat to immunocompromised inpatients.

The aim of this study was to characterise the burden and factors associated with Legionella spp. contamination across three Greek hospitals over five years.

We retrospectively analysed 292 water samples from three acute care Greek hospitals (March 2020-April 2025), cultured by a single accredited laboratory using the methods defined in International Organization for Standardization 11731:2017. A sample was classified as positive when viable Legionella spp. were recovered at or above the laboratory limit of detection; records reported as zero counts or below the limit of detection were treated as non-detections.

Overall, 20.9% (61/292) of samples were positive, with marked between-hospital differences: 21.2%, 35.0% and 4.4% (χ² = 27.1, P < 10^-6). Hospital 2 carried the heaviest burden (mean: 2188 colony-forming units [cfu]/L; 9.7% >10,000 cfu/L) and produced nine of the ten highest readings from hot-water taps. After adjustment, between-hospital differences remained the factor most strongly associated with detection (hospital 2 vs 3: odds ratio [OR] = 15.0, 95% confidence interval [CI]: 4.8-46.4; hospital 1 vs 3: OR = 8.6, 95% CI: 2.0-38.0; both P < 0.01). Hot-water sampling was independently associated with detection (OR = 2.76, 95% CI: 1.36-5.59, P = 0.005).

Differences between these three hospitals, together with hot-water systems, were the factors most strongly associated with Legionella spp. Persistent contamination in two of three networks supports a shift from outlet-level reactive disinfection towards whole-system, risk-based water-safety planning aligned with World Health Organization and European Society of Clinical Microbiology and Infectious Diseases Study Group for Legionella Infections guidance. As a three-hospital retrospective study, these findings are preliminary and require confirmation in larger datasets.