Healthcare Outbreaks Associated With a Water Reservoir and Infection Prevention Strategies

Abstract

Hospital water may serve as a reservoir of healthcare-associated pathogens, and contaminated water can lead to outbreaks and severe infections. The clinical features of waterborne outbreaks and infections as well as prevention strategies and control measures are reviewed. The common waterborne pathogens were bacteria, including Legionella and other gram-negative bacteria, and nontuberculous mycobacteria, although fungi and viruses were occasionally described. These pathogens caused a variety of infections, including bacteremia and invasive and disseminated diseases, particularly among immunocompromised hosts and critically ill adults as well as neonates. Waterborne outbreaks occurred in healthcare settings with emergence of new reported reservoirs, including electronic faucets (Pseudomonas aeruginosa and Legionella), decorative water wall fountains (Legionella), and heater-cooler devices used in cardiac surgery (Mycobacterium chimaera). Advanced molecular techniques are useful for achieving a better understanding of reservoirs and transmission pathways of waterborne pathogens. Developing prevention strategies based on water reservoirs provides a practical approach for healthcare personnel.

Hospital water and water-related devices as well as moist environments and aqueous solutions can serve as a reservoir of waterborne pathogens in healthcare settings [1, 2]. The hospital environment may allow contamination by waterborne pathogens, in part because water temperatures are suitable for bacterial growth, and the complex structure of hospital water systems often leads to stagnation, corrosion, and biofilm formation [3]. A variety of water reservoirs have been linked to nosocomial outbreaks including potable water, sinks, faucet aerators, showers, tub immersion, toilets, dialysis water, ice and ice machines, water baths, flower vases, eyewash stations, and dental-unit water stations [4]. Waterborne pathogens have included Legionella, other gram-negative bacilli, nontuberculous mycobacteria (NTM), fungi, protozoa, and viruses [3–5]. Transmission of these pathogens from a water reservoir may occur by direct and indirect contact, ingestion and aspiration of contaminated water, or inhalation of aerosols [1, 2]. Waterborne outbreaks caused by these pathogens and reservoirs have occurred among patients in healthcare settings and have been a serious threat to high-risk patients, especially critically ill patients and immunocompromised hosts, leading to substantial morbidity and mortality [3–5].

The aim of this review article was to (1) review healthcare-associated outbreaks and infections associated with a water reservoir from the published literature, and (2) provide infection prevention strategies and control measures by water reservoirs based on the published scientific evidence and available guidelines.

SEARCH AND SELECTION CRITERIA FOR REVIEWING THE LITERATURE

We searched the published literature via PubMed using the following Medical Subject Headings (MeSH) and keywords: (hospital OR hospitals OR hospital units OR nursing homes OR ambulatory care facilities OR ambulatory care OR dental facilities OR assisted living facilities OR healthcare settings OR patient) AND (waterborne pathogens OR Legionella OR Legionnaires' disease OR bacterial infections OR Mycobacterium infections OR fungal infections OR mycoses OR protozoan infections OR healthcare-acquired infection OR nosocomial OR cross infection OR outbreak) AND (hospital water OR drinking water OR potable water OR ice OR sink OR faucet OR faucet aerator OR shower OR tub or toilet OR water fountain OR water bath OR dialysis water OR decorative fountain OR ice machine OR air conditioning OR heater-cooler unit OR water microbiology OR disease reservoirs).

From January 1967 to July 2015, 2445 publications were identified as a result of our search and were reduced to 1746 by filters of availability in English and abstracts. The 1189 references (January 1997–July 2015) were screened using titles and abstracts, then selected articles were carefully reviewed. For the references published before 1997, we cited a previous review article [4] because of limited numbers of citable references. We excluded articles without abstracts, non-English-language articles, articles that were unavailable in PubMed, reported cases of contaminations or pseudo-outbreaks, and articles that did not include human infections. We identified 179 references by our initial screening process, and then prioritized listed citations to include at least 1 article from each reservoir and organism. Finally, 73 original articles [6–78] were selected and data were retrieved for further analysis. Table 1 provides the features of waterborne healthcare-associated outbreaks and infections by each category, including author, publication year, reservoir, organism, transmission, patient population, infection type, molecular typing, and study type.

HEALTHCARE-ASSOCIATED OUTBREAKS AND INFECTIONS CAUSED BY WATERBORNE PATHOGENS

A review of the literature revealed multiple outbreaks in healthcare facilities due to a variety of pathogens associated with a water reservoir (Table 1). Infections included bloodstream infections, pneumonia, and disseminated diseases (Table 1). Patient populations at risk for waterborne outbreaks and infections included those with hematological and other malignancies or stem cell transplants, immunocompromised patients, location in an intensive care unit (ICU), premature infants in a neonatal intensive care unit (NICU), burn patients, and patients during/after surgery (Table 1).

Causative pathogens included bacteria (Legionella, Pseudomonas, Acinetobacter, Serratia, Stenotrophomonas, Enterobacter, Klebsiella, Alcaligenes, Burkholderia, Chryseobacterium, Elizabethkingia, Halomonas, Ochrobactrum, Sphingomonas, NTM), fungi (Aspergillus, Mucor, Trichosporon, Fusarium, Exophiala), and viruses (norovirus) (Table 1). Details of overall waterborne pathogens in healthcare-associated infections [3, 5, 79–82], and specifically Legionella [83–86], Pseudomonas aeruginosa [87, 88], Aspergillus [89], and NTM [90], have been previously reviewed. Legionella and NTM can reside in hospital water distribution systems, while other gram-negative bacteria and fungi can form biofilms [3, 79]. We briefly updated key points from recent literature on healthcare-associated pathogens in waterborne outbreaks and infections as described below.

Legionella

Legionella can be detected in most water sources at low levels [91]. The level of contamination of Legionella in hospital water systems associated with disease remains unclear. A large healthcare-associated Legionella outbreak occurred at a Pennsylvania hospital despite implementing a Legionella disinfection program with a copper-silver ionization system of hospital water [40]. In this outbreak, a link between clinical isolates of Legionella and hospital environmental samples was confirmed by molecular typing, and Legionella was viable and transmissible despite the presence of copper and silver ion concentrations within the manufacturer's recommended range [40].

Multidrug-Resistant Gram-Negative Bacteria

As antimicrobial resistance in bacteria has become a global concern, multidrug-resistant gram-negative bacteria (Klebsiella, Enterobacter, Pseudomonas, Acinetobacter), including bacteria producing extended-spectrum β-lactamases or carbapenemases (eg, Klebsiella pneumoniae carbapenemase [KPC] and New Delhi metallo-β-lactamase), were described as waterborne pathogens causing healthcare-associated infections [19, 21, 22, 48, 50, 52, 54, 55, 57–59, 70, 71]. Multidrug-resistant organisms were most commonly linked to contaminated sinks as a reservoir (Table 1). Notably, an outbreak of 6 infections due to KPC-2–producing Klebsiella oxytoca in patients with hematological malignancies was linked to contaminated handwashing sinks [57].

Nontuberculous Mycobacteria

Our review found that a variety of NTM, including both rapid-growing and slow-growing species, has led to waterborne healthcare-associated outbreaks and infections as follows: Mycobacterium abscessus with tap water, Mycobacterium avium with potable water, Mycobacterium chelonae with ice and ice machines, Mycobacterium chimaera with heater-cooler units, Mycobacterium fortuitum with hospital water systems and showers, Mycobacterium genavense with tap water, Mycobacterium mucogenicum with bathing and tub immersion, electronic faucets, sinks, showers, and hospital water systems, Mycobacterium neoaurum with hospital water systems, Mycobacterium phocaicum with showers, Mycobacterium porcinum with ice and ice machines and tap water, and Mycobacterium simiae with tap water [7, 13, 20, 27, 28, 31, 34, 35, 37, 41, 45, 51, 60, 62, 66]. NTM can be present in municipal water systems and is relatively resistant to chlorination as well as being difficult to eradicate from contaminated water systems [3].

Fungi

Waterborne fungal outbreaks and infections occurred among patients with hematological malignancies or stem cell transplantation. Aspergillus can be detected from hospital water samples and has been reported to cause invasive aspergillosis [25, 43, 68, 89]. However, there are no criteria for contamination levels of hospital water associated with fungal infections. Furthermore, whether Aspergillus can cause waterborne healthcare-associated infections via aerosols generated from contaminated hospital water remains controversial [92]. Other fungal outbreaks associated with water reservoirs included Mucor with water-damaged plaster, Trichosporon asahii with wash basins, Fusarium with showers and sinks in a hospital water distribution system, and Exophiala jeanselmei with deionized water from the hospital pharmacy [11, 24, 74, 78].

RESERVOIRS AND TRANSMISSION ROUTES OF WATERBORNE PATHOGENS AND INFECTION PREVENTION STRATEGIES AGAINST WATERBORNE OUTBREAKS IN HEALTHCARE SETTINGS

All water with the exception of sterile water and filtered water is contaminated with microbes (eg, potable water, tap water, showers, and ice). In healthy persons, contact or ingestion of such water rarely leads to infection. However, contact or ingestion of such water may cause infection in immunocompromised persons or when applied to nonintact skin. Water-related reservoirs in healthcare settings during 1997–2015 were as follows: bathing and tub immersion, decorative water fountains, deionized water, dialysis water, electronic faucets, faucets, heater-cooler units, hospital wastewater systems, hospital water systems, ice and ice machines, ice baths, potable water, showers, sinks, tap water, toilets, wash basins, water baths, water birth, water-damaged plaster, and water-saving devices (Table 1). Transmission routes were primarily direct contact with contaminated water and water-related devices/activities or inhalation via aerosols generated from the contaminated water (Table 1). Waterborne healthcare-associated outbreaks and infections continue to occur and were mostly associated with well-recognized water reservoirs as previously described [4]. Moreover, recent studies document electronic faucets (P. aeruginosa, Legionella, M. mucogenicum) [13, 14, 93–95], decorative water wall fountains (Legionella) [10, 91], and heater-cooler devices used for cardiac surgery (M. chimaera) [20, 96] as water reservoirs. Infection prevention and control measures by each reservoir category are summarized in Table 2. The Centers for Disease Control and Prevention (CDC) has published recommendations of infection prevention and management for hospital water [1, 2].

Potable Water, Tap Water, and Hospital Water Systems

Various healthcare-associated infections were linked to contaminated potable/tap water and hospital water systems, especially among immunocompromised and severely ill patients [21–26, 28, 29, 37–39, 62–65, 68–70]. The common pathogens included gram-negative bacilli (eg, Pseudomonas, Stenotrophomonas) [21, 30, 38, 39, 69, 70], Legionella [23, 40], and NTM [27–29, 31, 37, 60, 62, 66]. Waterborne organisms may exist in potable water at acceptable levels of coliform bacteria (<1 coliform bacterium/100 mL) [4]. Standards for potable water should adhere to public health guidelines, and hot water temperature at the outlet should be maintained at the highest temperature allowable [1, 2].

Sinks

Some studies demonstrated a transmission link between sinks colonized with a pathogen and patients using molecular typing methods [47–52, 54–59]. The most frequent pathogens associated with sinks were gram-negative bacilli (eg, Pseudomonas, Acinetobacter, Serratia), as gram-negative bacilli can survive in wet environments, including sinks, for a long time [4]. Sinks should be decontaminated if epidemic spread of gram-negative bacteria via sinks is suspected [4]. Further, multidrug-resistant gram-negative bacilli have been linked to contaminated sinks and healthcare-associated infections as stated above [48, 50, 52, 54, 55, 57–59]. Although hand hygiene is an essential precaution against healthcare-associated infections in healthcare personnel, transmission from contaminated sinks to their hands during hand washing can occur [3]. Separate sinks should be used for handwashing and disposal of contaminated fluids [1, 4].

Showers

Healthcare-associated infections have been linked to use of contaminated showers or inhalation of aerosols by immunocompromised patients [41–43, 45]. Potential pathogens related to showers were Legionella, Pseudomonas, NTM, and Aspergillus [41–45]. The use of showers in immunocompromised patients (eg, neutropenic patients) should be avoided [4].

Bathing and Tub Immersion

Tub immersion used in hospitals for physical hydrotherapy and for cleaning of burn wounds can cause cross-transmission and transmission from environmental reservoirs [4]. Water contamination of central venous catheters during bathing was linked to bloodstream infection [7]; therefore, catheter sites should be covered with transparent occlusive dressing. Pathogens associated with bathing and tub immersion were Pseudomonas, Legionella, Alcaligenes, and NTM [6–9]. Tubs should be drained and cleaned after use of each patient, and strict adherence to proper disinfection of tubs between patients is essential [1, 4].

Electronic Faucets

Electronic faucets have been introduced in healthcare facilities, mainly to save water consumption and costs and avoid healthcare personnel's hands from being contaminated by touching the handle [93]. However, several studies have revealed that water samples from electronic faucets were heavily or frequently contaminated by P. aeruginosa and/or Legionella, compared with handle-operated faucets [93–95], and a P. aeruginosa outbreak due to electronic faucets among immunocompromised patients in a NICU has been reported [14]. Electronic faucets are more likely to be colonized because of low amounts of water flow, more favorable conditions for growth of waterborne pathogens (temperature of about 35°C in the column; materials made of rubber, plastic, and polyvinylchloride), and less flushing. Although the degree of risk posed by electronic faucets has not been quantified and there is no guideline for the water quality of electronic faucets, some authors have recommended monitoring water samples periodically from electronic faucets as well as removing electronic faucets from high-risk patient areas (eg, stem cell transplant units), given the difficulty of decontamination even with hyperchlorination [14, 93–95].

Decorative Water Wall Fountains

Two immunocompromised patients were exposed to a decorative water fountain in radiation oncology, which was heavily contaminated despite standard maintenance and disinfection (ozone generator, filter, and weekly cleaning), and developed Legionella pneumonia [10]. Another study described laboratory-confirmed Legionnaires' disease (pneumonia) in 8 patients, 6 of whom had exposure to a decorative fountain near the main hospital entrance; high counts of Legionella were detected from foam materials on the fountain despite routine maintenance [91]. A fountain's water has a closed recirculating system that can stagnate, and the spraying function may generate aerosols [91]. The results from both studies suggested that decorative water fountains can pose a risk for transmission of waterborne pathogens and should not be installed in healthcare facilities serving immunocompromised patients, even with standard cleaning, disinfection, and maintenance; at a minimum, water safety should be strictly monitored [10, 91].

Heater-Cooler Units

Heater-cooler devices are frequently used in cardiopulmonary bypass during cardiac surgeries to warm and cool patients’ blood [97]. A M. chimaera outbreak associated with heater-cooler units during cardiac surgeries as a water source has been recently documented [20, 96]. Patient infections included prosthetic valve endocarditis, bloodstream infection, and vascular graft infection. The transmission was potentially caused by aerosols generated from stirred water in contaminated heater-cooler unit water tanks. In a report of surgical site infections due to Mycobacterium wolinskyi, a cold-air blaster and a self-contained water system in a heater-cooler unit was identified as a potential reservoir [98]. The CDC's recommendations for healthcare facilities are as follows: (1) ensure that heater-cooler units are safe and properly maintained according to the manufacturer's instructions; (2) enhance vigilance for NTM infections in patients after cardiac surgeries using heater-cooler devices; and (3) if NTM infections are suspected, review microbiology data (NTM-positive cultures) and medical records of surgical procedures within 4 years after cardiac surgeries [97]. The US Food and Drug Administration recommended the following maintenance of heater-cooler units: (1) use sterile water or filtered water (≤0.22 µm) to rinse or fill water tanks or when making ice for cooling patients during a surgical procedure (avoid using tap water); (2) monitor cleaning, disinfection, and maintenance for heater-cooler units regularly with written documentations of the quality control program; and (3) avoid using heater-cooler units with cloudiness or discoloration in the fluid lines or circuits (http://www.fda.gov/MedicalDevices/Safety/AlertsandNotices/ucm466963.htm).

WATER SAMPLING AND MOLECULAR STRAIN TYPING IN OUTBREAK INVESTIGATION OF WATERBORNE HEALTHCARE-ASSOCIATED INFECTIONS

Infectious disease physicians and infection preventionists should recognize the emergence of waterborne healthcare-associated pathogens and the unusual change of these rates, and begin an initial investigation if waterborne healthcare-associated cases are suspected [3]. Microbiologic sampling can be considered to enhance epidemiological investigations and to develop infection prevention and control measures. Microbiologic water sampling is not routinely recommended, but can be performed in the following situations: support of outbreak investigations, research purposes, evaluation of a potentially hazardous environmental situation, and quality assurance (eg, biological monitoring of dialysis water or assessment of infection control measures) [1, 2]. Established water sampling methods (eg, volume, media, and incubation temperature) should be utilized to ensure recovery of waterborne pathogens; water samples should be filtered before culturing when low counts are expected and large volumes are needed (>100 mL) [1].

Molecular typing has been increasingly applied to outbreak investigations in healthcare settings, and the advantages as well as disadvantages of molecular methods have been described by other authors [99–101]. In our review, molecular typing methods used for waterborne healthcare-associated pathogens when applicable were mostly pulsed-field gel electrophoresis, followed by random amplification of polymorphic DNA (Table 1). Recent studies using whole-genome sequencing demonstrated more powerful evidence of water reservoirs in transmission dynamics of waterborne pathogens, including an outbreak of P. aeruginosa (1 pneumonia/17 colonizations) from contaminated water via sinks among infants in a NICU; transmission of P. aeruginosa from contaminated water outlets and a thermostatic mixer valve among burn patients receiving hydrotherapy even in plumbing of a new hospital; and 2 healthcare-associated cases of pneumonia caused by Legionella pneumophila serogroup 1 from a contaminated hospital water distribution system [30, 56, 102]. Thus, whole-genome sequencing is now promising for genomic comparative analysis in investigations of waterborne healthcare-associated outbreaks and can provide more accurate and informative strain typing to assess the relatedness of pathogens isolated from clinical and environmental water sources.

CONCLUSIONS

We reviewed waterborne healthcare-associated outbreaks and infections and summarized current infection prevention and control. With emergence of reservoirs and pathogens that have been unrecognized so far, waterborne healthcare-associated outbreaks and infections continue to occur and affect patients' health and safety, underscoring the significant role of water as a reservoir for healthcare-associated infections. Water contamination can cause pseudo-outbreaks [3, 80] as well as outbreaks/infections, both of which require substantial efforts and resources for investigation and control. Waterborne healthcare-associated infections are preventable for some pathogens and reservoirs, but eliminating contamination of waterborne pathogens as natural inhabitants of water systems may be difficult in healthcare settings. It is essential for healthcare personnel to understand reservoirs and transmission pathways of waterborne pathogens for developing prevention strategies and control measures of healthcare-associated infections. Multiple approaches of engineering and hygiene measures as well as surveillance and clinical management for hospital water can reduce the risk for contracting waterborne healthcare-associated infections [3]. Advancement of pathogen identification and molecular typing methods has enhanced outbreak investigations as well as provided better understanding of reservoirs and transmission routes of waterborne pathogens. The safe level of microbial water contamination, which would preclude healthcare-associated infections for any waterborne pathogen in susceptible patients, remains to be determined. Further scientific and practical evidence on hospital water reservoirs, pathogens, and healthcare-associated infections is needed to address unresolved issues on infection prevention and control.

Notes

Acknowledgments. We are very grateful to Karen Crowell, a clinical information specialist at Health Sciences Library, University of North Carolina, for her assistance with literature search.

Potential conflicts of interest. H. K. received financial support from the JSPS Postdoctoral Fellowship for Research Abroad. All other authors report no potential conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.

References