Checking that water supplies are free of harmful bacteria is a routine but essential task for healthcare facilities. Water microbiology consultant David Sartory outlines the microbes to look out for, how to carry out tests and take remedial action should the tests show positive
Water is a critical resource in hospitals, but seriously ill and immunocompromised patients are particularly at risk from infection, even from those microbes that are commonly present in water and pose little threat to the healthy. If a tap is used relatively infrequently, the water at the end of the pipe can be prone to stagnation, creating ideal conditions for the formation of a biofilm, allowing bacteria to multiply and making infection even more likely.
The two main waterborne bacterial threats are Pseudomonas aeruginosa and Legionella species. Both are fairly resistant to the chlorine and chloramines commonly used in water treatment and the disinfection of distribution systems, and in the relatively warm environment at the ends of pipes they will grow.
Legionella is usually transmitted in aerosol form, with shower heads and air conditioning units being common culprits. P. aeruginosa is more likely to be spread by direct contact with a tap or sink; if a patient or healthcare worker uses the tap and fails to disinfect their hands properly afterwards, transmission can result.
The major outbreaks and deaths that occurred in the 1970s and 1980s led to a much greater awareness of the problem
P. aeruginosa is widespread, and it is difficult to both prevent and eradicate. Many hospitals have introduced secondary disinfection systems that use either chlorine dioxide or chloramine to try and control it, and while this has reduced the problem, it is still there. Water systems must therefore be designed in such a way that biofilm growth at the ends of the pipes and taps is minimised.
Infection incidents are often isolated, but there have been cases where it has become widespread. A particularly tragic example in December 2011 and early 2012 involved the deaths of four neonatal babies in Northern Ireland. They developed pneumonia after contracting P. aeruginosa infection from tap water used to fill nebuliser units. This high-profile case led to a much greater focus being placed on preventing P. aeruginosa from causing major problems in the future.
While Legionella is less widespread, it is found in hospital pipe waters and occasional cases of legionellosis do still occur. The major outbreaks and deaths that occurred in the 1970s and 1980s led to a much greater awareness of the problem, and most national health and safety authorities, as is the case in the UK, now have very strict guidelines on the management of Legionella. In the UK these guidelines effectively have the weight of legislation, such that if there is an outbreak and the guidelines were not being followed, prosecution would likely result.
Perhaps surprisingly, the only routine water analysis required by UK law in hospitals is that normally required for drinking water, such as for the presence of coliforms. UK Department of Health guidance states that systems supplying water to wards with at-risk patients should be tested for P. aeruginosa, but the required frequency – every six months for P. aeruginosa-negative taps – may be insufficient, and most hospitals will test more frequently, typically every week for at-risk wards.
The sampling regime is important, but it is not always carried out in the most effective way
The sampling regime is important, but it is not always carried out in the most effective way. Ideally, a sample will first be taken from the tap before cleaning to test for the presence of contamination in the tap itself. The tap should then be cleaned, and a second sample taken after flushing. This gives an indication of the quality of the water itself. Two-part sampling is much more relevant to the real-world risk, as people rarely flush water through a tap for a long time before using it – they simply turn it on and use the water. Merely sampling after cleaning does not give a sense of how contamination builds up over time.
Two different tests are available for P. aeruginosa. In the traditional test, which is still widely used, a 100ml sample of water is passed through a membrane filter with a pore size of about 0.45µm. This filtrate is transferred onto an agar medium with cetrimide as a selective agent. The sample is incubated at 37°C for 48h before the plate is read; typical colonies of P. aeruginosa are easy to identify visually as they produce the blue-green pigment pyocyanin. A complication is the growth of atypical colonies (other colonies that are not blue-green but are fluorescent or are red-brown in colour), which have to be selected and subcultured onto confirmation media for identification as P. aeruginosa, which adds additional days onto the testing time.
Figure 1: Step-by-step guide to using Pseudalert
1. Add reagent to water sample
2. Cap vessel and shake to dissolve
The second, newer and faster type of test is Pseudalert/Quanti-Tray, which gives a confirmed quantified result in 24h, with no requirement for further testing. A powder containing a substrate for a diagnostic enzyme for P. aeruginosa is added to the water sample, before incubation at 38°C for 24h. Any P. aeruginosa contamination is immediately obvious under UV light as it glows blue-white.
False results caused by other species are not an issue with the Pseudalert method, as the test reagent contains a suppression system for non-target bacteria. With the rapid detection, simplicity of handling, and highly specific and sensitive results, this method, which has been available for use within the healthcare arena since 2014, has many advantages over the agar method, which must be undertaken by certified laboratories and is a laborious, multi-stage process (see Figure 1).
The method for detecting Legionella is complicated and has not changed for more than a decade; identifying the colonies on a plate requires the skill of an experienced analyst. The incubation period typically takes 5–7 days and, in some cases, up to 10 days, which is a long delay when trying to pinpoint the source of an outbreak.
Routine monitoring is, therefore, essential for those water sources most likely to produce aerosols that might cause infection, and is an important part of those HSE guidelines. Tap water samples are not routinely analysed, although some hospitals do carry out this level of testing. The testing of showers that are used by susceptible patients is much more important.
3. Pour sample into Quanti-Tray
4. Seal using the Quanti-Tray sealer
In addition to P. aeruginosa and Legionella species, there are three other bacteria that can cause hospital acquired infections. Infection with Stenotrophomonas maltophilia can be dangerous because it is difficult to treat. These bacteria are found in water and, while not at the same frequency as P. aeruginosa contamination, there have been cases where the water supply has been implicated as the source of infection within hospitals.
Other potentially problematic bacteria include Ralstonia pickettii, which is being found more frequently, and species of Serratia, which are commonly found in drinking water and have been associated with hospital-acquired infections. For these three bacteria, widespread testing of the water system is only really necessary if there is a specific infection incident.
5. Incubate for 24h at 38°C ± 0.5°C
6. Results – UV fluorescence is positive
While the risks from Pseudomonas and Legionella are well established, there is less awareness of the problems that can be caused by acanthamoebae. These tiny protozoa are widespread in water, and can cause infections in the eyes and nervous tissues. It has also been shown that many pathogens, Pseudomonas and Legionella included, can be ingested by acanthamoebae but not digested. Rather, they will grow and thrive inside the amoebae, adding another multiplication source for the bacteria that are a risk in the water supply of hospitals.
Legionella in particular grows well within acanthamoebae, which adds another layer of risk as the first-line of defence against Legionella in the lungs is the macrophages, whose bacteria-digestion activities closely resemble those of acanthamoebae. The same protection mechanisms that the bacteria have developed to evade acanthamoebae digestion, work just as well in human macrophages.
Detection of acanthamoebae relies on collection of water or biofilm material from a tap. This is then cultured on an agar plate with a lawn of bacteria, commonly heat-inactivated E. coli, and examined for up to 14–21 days’ incubation at 30°C for the presence of amoebae, which will be indicated by areas of clearing on the plate where they have eaten the bacteria. The amoebae can be examined under a microscope, and identified according to their shape and size.
Acanthamoebae are very widespread in nature, particularly in moist environments, and will probably be present in the water distribution system within any hospital
Acanthamoebae are very widespread in nature, particularly in moist environments, and will probably be present in the water distribution system within any hospital. For the majority of people, they pose no health risk. The exceptions are those with compromised immune systems, and contact lens users who rinse their lenses or carrying case with tap water. Yet acanthamoebae are not on the radar as a pathogen in their own right. While routine monitoring is not necessarily required, if there are biofilm issues involving any of the potentially problematic bacteria, then they certainly should be tested for, as they can act as foci for the multiplication of these bacteria.
Once confirmed results are in hand, they are used to inform the remedial actions that should be taken. There are two methods that are primarily used for managing a water system implicated as the source of infection. The most common technique is shock chlorination, where 50ppm (or more) of chlorine is introduced into the system, left for a set time, and then flushed out. This process has its own hazards, as taps must be closed off, and people prevented from using them to avoid harm being caused by the chlorine. Clearly, temporary alternative water sources for patients must be in place before this procedure is carried out.
The alternative process involves raising the temperature of the system to kill off bacteria. While this is relatively straightforward for a hot water system, it is not a practical solution for cold water systems. Modern thermal mixer taps that keep the water at about 43°C, which are increasingly common in hospitals, are also incompatible, as elevating the temperature can distort the needle valve that ensures the correct combination of cold and hot water.
If bacteria are detected and are proving difficult to eradicate, point-of-use filters can be added to shower heads and taps that filter out all the bacteria, preventing the Pseudomonas and Legionella getting through. These can be expensive, need replacing every couple of weeks, and the used filters need to be disposed of as infectious material, so they tend to be used only where there is a significant risk.
The widespread nature of many waterborne pathogens means it is never going to be practical to remove them from the water supply before it enters a hospital. It is only by good practice, efficient cleaning procedures and effective testing protocols that such hospital acquired infections can be prevented and controlled.
Pseudalert and Quanti-Tray are registered trademarks of IDEXX Laboratories.