Assessing the environmental health relevance of cooling towers – A systematic review of legionellosis outbreaks

Abstract

Bioaerosols from cooling towers are often suspected to cause community-acquired legionellosis outbreaks. Although Legionella infections can mostly be assigned to the emission sources, uncertainty exists about the release and distribution into the air, the occurrence of the respirable virulent form and the level of the infective concentration. Our study aimed to evaluate studies on legionellosis outbreaks attributed to cooling towers published within the last 11 years by means of a systematic review of the literature. 19 legionellosis outbreaks were identified affecting 12 countries. Recurring events were observed in Spain and Great Britain. In total, 1609 confirmed cases of legionellosis and a case-fatality rate of approximately 6% were reported. Duration of outbreaks was 65 days on average. For diagnosis the urinary antigen test was mainly used. Age, smoking, male sex and underlying diseases (diabetes, immunodeficiency) could be confirmed as risk factors. Smoking and underlying diseases were the most frequent risk factors associated with legionellosis in 11 and 10 of the 19 studies, respectively. The meteorological conditions varied strongly. Several studies reported a temporal association of outbreaks with inadequate maintenance of the cooling systems. A match of clinical and environmental isolates by serotyping and/or molecular subtyping could be confirmed in 84% of outbreaks. Legionella-contaminated cooling towers as environmental trigger, in particular in the neighbourhood of susceptible individuals, can cause severe health problems and even death. To prevent and control Legionella contamination of cooling towers, maintenance actions should focus on low-emission cleaning procedures of cooling towers combined with control measurements of water and air samples. Procedures allowing rapid detection and risk assessment in the case of outbreaks are essential for adequate public health measures. Systematic registration of cooling towers will facilitate the identification of the source of outbreaks and help to shorten their duration.

Introduction

Approximately 57 Legionella species comprising at least 79 serogroups have been described and all of them are considered as potential human pathogens (Robert-Koch-Institut, 2012). Epidemiologically, Legionella pneumophila (Lp) serogroup 1, which is responsible for the majority of human diseases (about 90%), represents the most relevant species and serogroup (Diederen, 2008).

Legionella infections can cause Legionnaires’ disease (LD), a severe pneumonia with a case fatality rate of 10–15% (WHO, 2013). Another possible, milder form of manifestation is Pontiac fever presenting with flu-like symptoms (Diederen, 2008, Robert-Koch-Institut, 2011). Despite mandatory reporting for legionellosis in several countries, the true number of cases is probably highly underestimated. For instance, approximately 600 infections are annually reported in Germany. However, the actual number of community-acquired cases of pneumonia caused by Legionella infections per year is estimated at 15,000–30,000 by CAPNETZ (network of excellence for the Community Acquired Pneumonia) calculations (Robert-Koch-Institut, 2012, vonBaum and Lück, 2011). Thus, 4% of pneumonia cases in Germany, which were not acquired in the hospital, are caused by Legionella infection and up to 80% of these by Legionella pneumophila (vonBaum et al., 2008). The problem of underestimation is also known from other European countries (ECDC, 2010) and other continents (Center for Disease Control and Prevention, 2011).

Legionella species show ubiquitous distribution in the aqueous environment and particularly occur in technical water carrying systems. The main risk of human infection results from the inhalation of aerosolized particles or from the aspiration of water containing Legionella, whereas lateral transmission of Legionella does not occur. Legionella have the ability to grow and replicate within protozoan hosts such as amoebae (Cirillo et al., 1999). Thus, not only the number of free-living bacteria invading the lower respiratory tract but also the presence of respirable vesicles containing Legionella is critical for the occurrence of Legionella infections. Moreover, amoeba-grown Legionella exhibit increased virulence (Cirillo et al., 1999). In this virulent state Legionella are widely insensitive to osmotic changes, biocides and antibiotics and display toxic effects (Fields, 1993, Rowbotham, 1980, Starlinger and Tiefenbrunner, 1996, Steinert, 1999, Swanson and Hammer, 2000). It has been shown that amoebae release respirable vesicles with a diameter of 2.1 to 6.4 μm (Berk et al., 1998) which can contain hundreds of infective Legionella. Principally, a few inhaled amoebae vesicles containing Legionella can already cause an infection (Armstrong and Haas, 2007, O’Brien and Bhopal, 1993). Unfortunately, the infective dose cannot be determined by the routinely conducted cultivation method, because Legionella enclosed in amoebae are not detected and a single bacterium as well as a multitude of bacteria released by amoebae or vesicles, constitutes only one colony forming unit (CFU). Therefore, a correlation between the number of CFU in water or air and the risk of infection cannot be specified (Ishimatsu et al., 2001).

There is an ongoing scientific debate concerning relevant sources and environmental conditions contributing to airborne transmission and aspiration of Legionella resulting in human infection. Artificial aquiferous systems such as water supply facilities, especially heated reservoirs, and cooling towers as well as affected compost represent potential sources of environmental infection (Swanson and Hammer, 2000). Generally, meteorological conditions have influence on the environmental occurrence of Legionella. For example, in the extremely hot summer of 2003 an increase of 17–18 °C in water temperature was measured in the dams and in main routes of drinking water supply of several communities in Bavaria, Germany which may promote a rise in Legionella proliferation (Behling, 2004). Depending on the environmental conditions, considerably higher concentrations of Legionella can even be found in fishponds than in water of swimming pools or cooling towers (Moosavian and Dashti, 2011). On the other hand, Legionella emitted by contaminated cooling towers can be transported over several kilometres within respirable vesicles, which are comparably resistant to physical and chemical influences (Cirillo et al., 1999). Geographical distribution seems to be associated to specific meteorological conditions, for instance low-level inversion preventing vertical intermixture and supporting horizontal transport of aerosol particles (Engelhart et al., 2007).

Previous investigations by our institution have shown, that aerosolized water of urban cooling towers is indeed contaminated by microorganisms including Legionella species in many cases (approximately 13% Legionella-positive samples, 36% Pseudomonas-positive samples) (Fembacher et al., 2007). However, it has to be considered that aerosolized particles containing Legionella are not released continuously and that maintenance actions can be associated with an increased release of Legionella into the ambient air (Critchley and Bentham, 2007, Exner and Pleischl, 2010). Point in time, triggers and extent of Legionella release from biofilms via aerosol leading to a distribution into the environment are widely unknown. Moreover, consistent data on transport distances depending on distinct meteorological conditions (thermal inversion, cold season) and ambient air concentrations in the context of community outbreaks are missing. These problems strike especially cooling towers, which are widely distributed and have been consistently attributed to community-acquired legionellosis outbreaks (Engelhart et al., 2007, Yu, 2008).

Many relevant aspects related to cooling towers as a source of legionellosis outbreaks were covered by a WHO report of legionellosis (WHO, 2007). This report focuses on risk management of outbreaks, disease surveillance, regulatory issues especially control measures and laboratory aspects. Although the WHO overview covers many relevant aspects of legionellosis outbreaks related to cooling towers, it lacks of a systematic and comprehensive review of outbreak studies attributed to cooling towers especially in the last decade. Moreover, molecular subtyping methods have been further developed and routinely applied in the last years, and therefore should be considered in the analysis of cooling towers as a source of legionellosis outbreaks. These techniques enable a valid match between environmental and clinical isolates and verify cooling towers as a source of legionellosis infections. The aim of this study was to investigate the relevance of cooling towers especially with respect to susceptible individuals by extending the knowledge gained from the WHO overview and by confirming its previous findings via analysis of recent studies. Therefore, we conducted a systematic review of legionellosis outbreaks of the last decade to highlight the public health consequences and to reconsider regulations recommended by the WHO.

In the present study a systematic review and analysis of investigations on legionellosis outbreaks attributed to cooling towers published within the last 11 years was performed. Evaluation focussed on characteristics of the outbreaks, including morbidity and mortality, individual risk factors, diagnostic tools, meteorological conditions, maintenance actions conducted shortly before the outbreak as well as sample analysis confirming a relationship between clinical and environmental isolates taken from the cooling towers.