Hospital water and life-threatening illnesses

Kevin G Kerr
Department of Microbiology, Harrogate District Hospital, Harrogate, Yorkshire, UK

Anna M Snelling
Bradford Infection Group, University of Bradford, Bradford, Yorkshire, UK

There is a growing appreciation of the true scale of the clinical and economic costs of nosocomial infections. As a result, these infections, and the measures taken by healthcare providers to prevent them, have come under intense scrutiny not only from healthcare regulators but also from patients, the media and the wider public.

Infections caused by methicillin-resistant Staphylococcus aureus (MRSA) and Clostridium difficile have enjoyed particular notoriety and many of the interventions designed to control healthcare-associated infections have focused on these pathogens. Such measures include screening patients for MRSA carriage, antimicrobial stewardship programmes and hand-hygiene initiatives.

However, by targeting infection-control practices and resources against a subgroup of healthcare-associated infections, there is a risk that other pathogens are overlooked. Particularly important in this respect are Gram-negative bacteria such as Pseudomonas aeruginosa, a long-established cause of healthcare-associated infections and Stenotrophomonas maltophilia which is an emerging nosocomial pathogen.

Pseudomonas aeruginosa infections
Pseudomonas aeruginosa is a very versatile pathogen and a common cause of a wide range of infections in hospitalised patients including bacteraemia, pneumonia, urosepsis and wound infection.1 Indeed, the US National Healthcare Safety Network reported that P. aeruginosa is the sixth most frequently occurring nosocomial pathogen, the second commonest cause of ventilator-associated pneumonia and the seventh commonest cause of catheter-related bloodstream infection.

There are particular patient groups that are at increased risk of infection and these include immunocompromised patients, especially those with chemotherapy-induced bone marrow suppression, those requiring intensive care and patients with severe burn injuries.1

Morbidity and mortality associated with 
P. aeruginosa infection is high, in part reflecting the underlying vulnerabilities of patients affected, but also because of the increasing problem of antibiotic resistance manifested by this species.1 The emergence of multi-drug resistant strains presents a significant challenge in this respect.

Furthermore, infections caused by multiple-antibiotic-resistant strains of P. aeruginosa are associated with increased attributable hospital costs and length of patient stay.2 Of particular concern is the advent of pan-resistant strains, which are resistant to all currently available antibiotics and thus may render some infections untreatable.1

Stenotrophomonas maltophilia infections
Stenotrophomonas maltophilia is associated with an ever-widening spectrum of nosocomial infections which, as with P. aeruginosa, are most frequently observed in patients who are debilitated or immunocompromised. Individuals who are rendered neutropenic following chemotherapy for haematological malignancy are particularly at risk.3  

Likewise, effective management of S. maltophilia infection is hampered because the bacterium is resistant to many currently available antibiotics. In particular, S. maltophilia is inherently resistant to the carbapenem class of antimicrobials, which are regarded as the ‘drugs of last resort’ for treating infections caused by multi-resistant 
bacteria.3

Given the morbidity and mortality arising from infections caused by both of these species – coupled with the acknowledged difficulties in antimicrobial therapy – it is clear that robust infection control strategies should be considered a priority by healthcare providers. Crucial to the design and implementation of infection prevention measures for any nosocomial pathogen is an understanding of the potential sources of the bacterium as well as routes of transmission and modes of acquisition in the hospital setting.

P. aeruginosa, S. maltophilia in hospital water
Both of these species are ubiquitous in the hospital environment and can be readily isolated from an extremely wide range of reservoirs. Sites which are most likely to harbour P. aeruginosa and S. maltophilia are water-related and include taps, sinks, showers, flower vases and hydrotherapy pools as well as potable water.1,3 In addition, these bacteria can contaminate devices or sites where moisture or humidity is high such as mop heads and other cleaning items and respiratory therapy equipment.1,3

Bacterial persistence in these environmental niches is also promoted by their resistance to many commonly used disinfectants. Furthermore, both species have the ability to form complex layered communities known as biofilms on inanimate surfaces, including the inside of water pipes and hoses. As well as impeding their physical removal during cleaning activities, bacterial cells embedded in biofilm can manifest increased tolerance to disinfectants.  

Although there is no disagreement that P. aeruginosa and S. maltophilia are common inhabitants of the hospital environment, there has, however, been no consensus on the implications of their presence for patient care. Certainly, there is no shortage of reports to implicate hospital water supplies as the source of P. aeruginosa infection. However, because many of these were published before the advent of molecular typing (‘genetic fingerprinting’) which permits detailed analysis of the relationship between clinical and environmental strains, the conclusions that can be drawn from these earlier publications are limited.4

More recently, however, outbreak investigations that have employed molecular typing techniques have provided evidence that hospital water supplies and outlets are an important source of infecting strains. For example, a protracted outbreak of infection with P. aeruginosa on a neurosurgical intensive care unit (ICU) which lasted 16 months, during which the outbreak strain was isolated from tap water on the unit, was only terminated following replacement of all sinks on the unit.5

Similarly, during investigation of an outbreak of S. maltophilia infection involving five preterm infants on a neonatal ICU, the outbreak strain was isolated from tap water from three outlets on the unit, but not from any other environmental source.6 Outbreak investigations should not be confined to sampling of mains water supplies and distribution points. This was amply illustrated in a report of an outbreak of P. aeruginosa infection involving six ICUs of a large teaching hospital. Despite extensive sampling of the environment, the outbreak strain could be isolated only from bottles of still water; both in-use and unopened. The outbreak ended following removal of bottled water from the units.7

Prospective studies
Despite evidence from investigations such as these, it has been argued that isolation of P. aeruginosa and S. maltophilia from the hospital water supply during outbreak situations merely reflects contamination of taps and sinks by the patient rather than the converse. Theoretically, of course, there is no reason to suppose that both routes of contamination might take place and results from prospective studies conducted in non-outbreak situations have helped to gain a better understanding of this issue.

Prospective studies involve the screening of patients on admission to a unit (to exclude patients who are already infected or colonised) and regular screening thereafter to identify as accurately as is possible when patient acquisition of the bacterium occurred.

Contemporaneous water and/or outlet specimens are obtained to determine whether patients may have acquired the infection from a water source or vice versa.  One such prospective study investigated a 16-bedded ICU over a six-month period. Weekly tap water specimens were obtained, as were throat, rectal, sputum and urine cultures from patients on admission and weekly thereafter.

Results demonstrated that P. aeruginosa was indeed acquired by patients following exposure to contaminated tap water, but also that colonised patients could also be the source of contamination for taps.8

Another study conducted over a 12-month period in five ICUs of a single hospital found that 56/132 (42%) of P. aeruginosa isolates obtained from patients colonised or infected with the bacterium were identical to those obtained from taps on the units.9  

Although most investigations have focused on taps, tap water and sinks as a source of P. aeruginosa and other environmental Gram-negatives, there is increasing evidence to suggest that shower heads and shower water might represent an important source of these bacteria.10

Interventions to prevent infection
Accumulating evidence of the importance of hospital water as a source of infections caused by P. aeruginosa and S. maltophilia in both endemic and outbreak situations has provided the impetus to develop interventions to counter this problem. In some circumstances, the solution has been relatively straightforward; for example, the outbreak of S. maltophilia infection described earlier was brought under control after the introduction of sterile water for washing infants in place of tap water.6

In other instances, however, control of the outbreak has been more difficult to achieve. Replacement of taps on a paediatric surgical unit was required to halt an outbreak of P. aeruginosa infection after a number of other measures including chlorination of the water supply and disinfection of sinks was unsuccessful.11

Another outbreak of P. aeruginosa infection associated with contaminated hand hygiene sinks was not controlled by various attempts at disinfection and was terminated only after the sinks were replaced.12

Interventions to control a water-associated infection-control problem often include measures directed at the hospital water distribution system, such as increasing water temperature or the use of copper/silver ionisation, but these may, at best, be only partially successful.13 It can be argued that this is because such interventions are not effective at eradicating biofilm-associated bacteria from water outlets or they can be difficult and costly to implement throughout large distribution systems.

Point-of-use filtration
An alternative technique for controlling water-borne infection and one which represents a potential solution to the problem of biofilms is point-of-use filtration. An outbreak of P. aeruginosa blood-stream infection on an haematology unit was resolved after the fitting of 0.2µm point-of-use filter cartridges on all taps and showers.14 Point-of-use filters were also introduced as a control measure in a surgical ICU in which water-associated P. aeruginosa infections had become endemic.

Previous attempts at infection control had included use of sterile water for purposes such as mouth care and regular cleaning of taps, but, as these had proved unsuccessful, point-of-use filters were installed on all water outlets.  

Prior to filtration, 113/117 (97%) water specimens grew P. aeruginosa with molecular typing showing that all the water isolates and 25/27 (92.6%) of patient isolates were identical. After the introduction of point-of-use filters, 0/52 water samples yielded P. aeruginosa and the number of patients colonised or infected with the bacterium was significantly reduced.

The decline in infections and associated reduction in the use of antipseudomonal antibiotics more than offset the cost of water filtration.15 Although not intended as a specific control measure for P. aeruginosa and 
S. maltophilia, a guideline from the French Ministry of Health in 2002 advised that hospitals install point-of-use filters in high-risk areas.

In recent years, treatment of hospital water supplies with chlorine dioxide has been evaluated with varying degrees of success, principally as a control measure for Legionella pneumophila, but there is, as yet, little published data on the effectiveness of this approach against 
P. aeruginosa and S. maltophilia.

Conclusions
Pseudomonas aeruginosa and S. maltophilia are important causes of nosocomial infection and are associated with significant morbidity and mortality especially for immunocompromised patients and other vulnerable patients on intensive care units. Management of these infections is becoming increasingly difficult because of antibiotic resistance and the emergence of pan-resistant strains and, consequently, this has focused attention on infection prevention and 
control strategies.

Control measures should take into account an ever-growing body of evidence which implicates hospital water supplies as a potential source of both P. aeruginosa and S. maltophilia and should also include guidelines for the provision of microbiologically safe potable water for immunocompromised patients.

A range of such interventions have been employed, but, in many instances, they have not been universally successful, although accumulating experience from point-of-use filtration suggests that this technology may play a valuable role in preventing water-associated P. aeruginosa and S. maltophilia nosocomial infections within defined areas of the hospital where at-risk patients are found.

References

1. 

Kerr KG et al. J Hosp Infect 2009;73:338-44.
2. 
Mauldin P et al. Antimicrob Ag Chemother 2010;54:109-15.
3. 
Denton et al. Clin Microbiol Rev 1998;
11:57-80.
4. 
Anaissie EJ et al. Arch Intern Med 2002;162:1483-92.
5. 
Bert F et al. J Hosp Infect 1998;39:53-62.
6. 
Verweij PE et al. Epidemiol Infect 1998;120:251-6.
7. 
Eckmanns T et al. Clin Microbiol Infect 2008;14:454-8
8. 
Rogues AM et al. J Hosp Infect 2007;67:72-8.
9. 

Blanc DS et al. Intensive Care Med 2004;30:1964-8.
10. 
Feazel LM et al. PNAS 2009 106;1693-9.
11. 
Ferroni A et al. J Hosp Infect 1998;39:301-307.
12. 
Hota S et al. Infect Control Hosp Epidem 2009;30:25-33
13. 
Petignant C et al. Infect Contro Hosp Epidem 2006;27:953-7.
14. 
Vianelli N et al. Haematologica 2006;91:983-5.
15. 
Trautmann M. Am J Infect Control 2008;36:421-9.