Letters in Applied Microbiology 1999, 29, 81–84 The occurrence of fungi in hospital and community potable waters M. Arvanitidou, K. Kanellou, T.C. Constantinides and V. Katsouyannopoulos Laboratory of Hygiene, Medical School, Aristotelian University of Thessaloniki, Greece 2102/99: received 18 February 1999, revised 5 May 1999 and accepted 5 May 1999 M. ARVANITIDOU, K. KANELLOU, T.C. CONSTANTINIDES AND V. KATSOUYANNOPOULOS. 1999. The prevalence of fungi was investigated in 126 potable water samples (84 hospital and 42 community samples), in parallel with the standard pollution indicator micro-organisms. Filamentous fungi were isolated from 104 of 126 (82·5%) samples and yeasts from 14 (11·1%), whereas their mean counts were 36·6 and 4·4, respectively. Fungi were isolated from 95·2% of community and 76·2% of hospital water samples, with the difference being statistically significant (P ³ 0·05), while yeasts were isolated from 9·5 and 11·9%, respectively. Prevailing genera were Penicillium spp., isolated from 64, Aspergillus spp., from 53, and Candida, from nine of the examined samples. Colony-forming units of yeasts were significantly correlated with those of total and faecal coliforms, whereas the counts of filamentous fungi were significantly correlated with total heterotrophic bacteria counts. These results suggest that tap water is a potential transmission route for fungi both in hospitals and the community in the examined region and may pose a health hazard mainly for the immunocompromised host. INTRODUCTION Fungi are an emerging cause of hospital-acquired infection in the 1990s. Although endogenous infection following colon- ization is the prevailing mode of fungal transmission, exogen- ous acquisition may occur and result in outbreaks (Muittari et al. 1980; Macher and Girman 1990; Beck-Sague et al. 1993). Fungi are natural inhabitants of soil and water and some species behave as opportunistic pathogens in man. Fungi grow in water systems where there are dead spaces and parts of the pipework that have low circulation. Furthermore, the materials forming tanks and washers can contribute to biofilms containing fungi through the slow release of organic material (Niemi et al. 1982). With the increasing number of patients undergoing complex medical therapy, more unusual infectious disease processes have emerged. Fungal pathogens as agents of a variety of infections are part of this trend and are found much more frequently as the cause of both community- and hospital-acquired infections in immuno- compromised patients. In the present study, the total heterotrophic bacteria, total coliform and faecal coliform counts were estimated in parallel Correspondence to: Prof. M. Arvanitidou, Laboratory of Hygiene, Medical School, University of Thessaloniki, 54006 Thessaloniki, Greece (e-mail:vayonaa ¨med.auth.gr). © 1999 The Society for Applied Microbiology with the counts of yeasts and fungi in drinking water. The main objectives of this study were to find the prevalence of yeasts and fungi in potable water samples, to compare the occurrence of these pathogens between hospital and com- munity tap water samples and to investigate the correlation of yeasts and fungi with the standard indicator bacteria. MATERIALS AND METHODS Isolation procedures The presence of fungi was investigated in 126 water samples in parallel with the standard indicator bacteria. Forty-two of the samples were community and 84 hospital tap water samples. Water was sampled after 3–5 min of free flow through a distribution system tap. Before sampling, a solution of sodium hypochloride (100 mg NaOCl l -1 ) was applied to the tap and the water run for an additional 2–3 min after treatment. Gloves and long sleeves were worn when collecting the samples to prevent skin bacteria from contaminating the samples. The membrane filter technique was employed for yeasts and fungi (American Public Health Association 1995). A volume of 100 ml of the samples was filtered, in duplicate, through membrane filters with pores 0·45 mm in diameter.