Gaining control of waterborne infections

These were some of the key clinical issues discussed by microbiologists at a one-day seminar on waterborne infections organised by Pall Medical in Glasgow last October. The audience heard an outline of the methods used to investigate outbreaks of Pseudomonas aeruginosa in a new purpose-built hospital from consultant microbiologist Dr Michael Weinbren, director of infection control at University Hospital, Coventry and Warwickshire (UHCW). UHCW had an outbreak on the neonatal intensive care unit (ITU) soon after the first (West) wing opened in August 2005. The contamination was tracked down to tap outlets and terminated with the installation of new taps, use of sterile water for bathing babies and alcohol gel on hands after washing. UHCW was constructed in three stages, with the central and East Wings opening in July 2006. Within two weeks of these further wings opening, Pseudomonas was isolated again, from six patients in the general critical care unit (GCC) on the same day. While this level of infection on the unit might not usually be considered an outbreak, in the light of the previous appearance of P. aeruginosa, a preliminary investigation was needed, to establish if this could be the case. Tests were carried out, the results of which suggested that the Pseudomonas infections were linked to the water system. Infection control measures were established. Hands were to be dried and gel used after washing, and patients were only to be washed with water from Pseudomonas negative outlets. However, further sampling revealed six new culture positive outlets and two previously negative outlets testing as positive. Why were the results inconsistent and which outlets were truly negative? No Pseudomonas was found in the taps. The interim control measures were clearly no longer adequate and Dr Weinbren’s team were faced with establishing the organism’s precise location, its route of transmission and whether other areas of the hospital were affected. The team took timed samples at zero, one and then two minutes. The highest concentration of Pseudomonas was found in the first sample and then tailed off, indicating the contamination was at the periphery of the system and not the main pipes. “This finding has important implications for water testing,” Dr Weinbren explained. “Because it clarifies why you would get a negative result for Pseudomonas if the taps on the unit had been run for some time prior to testing. Because Pseudomonas will be harboured in a biofilm, rather than free-floating in planktonic form, we knew that the contamination represented around 1% detached from the system’s total burden.” Hot and cold water at the outlets blend at a thermostatic mixer valve (TMV) and is then carried via a flexible hose. Sampling revealed biofilm within the flexible hose in the positive outlets, supporting a heavy growth of Pseudomonas. Flexible hoses, especially imported non WRAS-approved hoses, have previously been implicated in Legionella outbreaks. Dr Weinbren noted that UHCW was the first UK hospital to be fitted with totally plastic pipework throughout its whole system. Further investigations into possible routes of transmission demonstrated that the risk from hands was probably low. However, use of tap water for washing respiratory equipment and bathing patients seemed likely routes. POU filters are the only real answer to producing large quantities of organism-free water, Dr Weinbren said. Cleaning also proved to be another route of transmission. An investigation into cleaner’s spray bottles surprisingly revealed growth of the outbreak strain of Pseudomonas at a concentration of 108organisms/litre, even after dilution. Most spray bottles were found to be carrying incorrect dilutions. Infection control measures on the GCC were revised by replacing all flexible hoses on the unit with copper piping and placing POU Pall filters on selected sinks. However, Pseudomonas still remained beyond the TMVs. The next step was to find out if the problem was isolated to GCC, because of the type of patient and the number of clinical specimens taken, or whether other areas of the hospital were affected and were transmitting Pseudomonas to the GCC. Early morning sampling (when water has become more static overnight) of 20 outlets supplied from the same tank as the GCC found two outlets, but two floors higher, positive with the same unique outbreak strain as the GCC, with the flexible hoses again being the culprit. Further work revealed that the O-rings and springs of the TMVs serving known positive outlets were seeded with biofilm. Investigation of 1,000 outlets in the rest of the hospital showed up a patchy distribution of Pseudomonas across the first opened West Wing, with clusters in the central block and only a few isolates from the last constructed East Wing. Further plotting of serotypes showed the strains in the West Wing to be unique and unconnected to those found in the Central block and East Wing. The question remained whether Pseudomonas from the water supply was being transmitted to patients elsewhere in the building. Investigations to match Pseudomonas from different areas of the hospital gave no concrete evidence for transmission across the rest of the hospital. Searching for more evidence, the infection control team found antibiotic sensitivity to standard anti-pseudomonal antibiotics among 12 strains of Pseudomonas. However, plates accidentally left out on the bench at 30°C grew Stenotrophomonas maltophilia, a Gram negative pathogen resistant to meropenem, which was shown also to be lurking alongside Pseudomonas in the biofilm found in flexible hoses and TMVs. To find out if Pseudomonas was a problem confined only to UHCW the team anonymously sampled outlets on ITUs in the East and West Midlands. Pseudomonas was found in several water supplies as was S.maltophilia in the one instance it was tested for. “If you go looking for these organisms and carry out the correct tests, you’ll probably find them,” Dr Weinbren commented. While none of these units were previously aware of having a Pseudomonas problem, up to 50% of the outlets in some were found positive. Significantly, the one unit where no Pseudomonas was found had better designed and positioned outlets.

Water is a source of outbreaks

Summarising some of the lessons learnt, Dr Weinbren noted that worldwide, water is increasingly being recognised as the source of Pseudomonas outbreaks on ITU and other high dependency units. “Timing and sampling methods are crucial: testing of only one or two samples could give false reassurance,” he said. “It is increasingly important that those involved in infection control need to understand water systems and plumbing, and the mechanisms by which water gets into patients.” Dr Weinbren remarked that water standards for Pseudomonas in high risk areas could be introduced as the UK currently has none, yet some European countries do. The introduction of prospective studies would also help clarify the risk of transmission of contaminated outlets to patients. For example, is there a higher risk of contamination of outlets on ITUs due to contaminated secretions being discarded in sinks? Dr Weinbren concluded: “Resistant strains responsible for transmission of Pseudomonas on an ITU tend to be investigated further, which may lead to their over-representation in the literature. Yet involvement of sensitive Pseudomonas may be missed because it is generally considered endemic in ITUs. But this background does not mean sensitive strains are not a problem in the water. Water should always be considered as a contributing factor.”