A high proportion of hospitalised patients are immunocompromised due to either therapy, medical condition or surgical procedure and as such are more susceptible than the general population to infectious agents if present in the hospital environment. This paper reviews the presence of Legionella species from 20,149 cooling tower water samples from both hospital and commercial premises over a three year period from April 1992 to March 1995. The highest monthly Legionella isolate rate was 20.4% of samples tested and there was an increased isolation rate in the summer and early autumn months.
Based on these results, maintaining a cooling tower to the Australian Standard AS3666 and performing three monthly tower cleans is not sufficient to guarantee a continued absence or a low level of contamination by Legionella. Recommendations are provided on the usefulness and frequency of regular testing of cooling tower water for plate count (total bacteria count) and Legionella in relation to Australian Standard AS3666, Australian Standards handbook HB-32, Victorian and New South Wales Health department guidelines in the absence of Queensland guidelines.
The paper also examines additional laboratory analyses available to infection control practitioners for monitoring; internal building surfaces and airconditioning systems for air filter efficiency, hydrotherapy pool water quality to appropriate National Health and Medical Research Council (NHMRC) guidelines, potable water quality of both bottled water and bubbler dispensers to NHMRC guidelines, kitchen hygiene of food preparation surfaces to the American Public Health (APHA) standards and food quality to the Australian Food Standards code and the "Guidelines For Ready to Eat Foods" by the Food Unit, Queensland Department of Health. Methods used for the different analyses are performed to meet APHA or Australian Standards. The review shows that there are currently a large number of laboratory services available to infection control practitioners to specify quality requirements of the hospital environment and auxiliary services and monitor these requirements by the initiation of a quality surveillance program.
Cooling Tower Water Analyses
The sources of air borne contamination by Legionella can be via environmental contamination of air intake ducts or from contaminated potable water distributed via aerosols produced in equipment such as hospital bathroom shower heads[2,7]. The contamination of hospital air intake ducts can be due to aerosol drift from hospital or community cooling towers (from 200 metres to up to a three kilometre radius). At some hospitals, Legionella pneumophila has been isolated from cooling towers, and aerosols from these were implicated in transmitting the bacteria to patients[11,23].
A total of 20,149 cooling tower water samples were processed for Legionella analysis (LA) in the period from 1/4/92 to 31/3/95. The geographical area for these samples encompassed Northern New South Wales, Queensland and the Northern Territory. The samples included in the study had either a LA only performed or both a LA and plate count (PC) performed as specified by the client. Samples were obtained from both hospital and commercial cooling towers. As samples were generally received coded, evaluation of results specifically from hospitals could not be performed.
Sample Collection and Transport
Samples were collected into sterile containers, volumes received varied from 50mL to 250mL. All samples were collected and held prior to transport under the control and supervision of the client. Clients were advised to refrigerate samples prior to transport and to transport samples refrigerated if the ambient temperature during the delivery period was not below 25 O C as specified in Australian Standard AS3896.
Samples were refrigerated on receipt at the laboratory prior to processing. Samples received before 7.30pm (Monday to Friday) were processed on the same day.
Methods of Analysis
The method used for LA was the Australian Standard method AS3896. The biocide reduction step used was the centrifugation method rather than the filtration method. Legionellae isolated were identified as Legionella pneumophila serogroup 1, Legionella pneumophila serogroup 2 to 14 or Legionella species (not pneumophila). A negative culture was reported as <10 colony forming units per millilitre (CFU/mL). This is a National Association of Testing Authorities (NATA) accredited method.
Plate Count Analysis
The method used for PC analysis was the spread plate method of the American Public Health Association (APHA) standard method 9215 Water and Waste Water. Results were reported in the range <1,000 to >1,000,000 CFU/mL. This is a NATA accredited method.
Results of Legionella Analysis
Of the 20,149 samples processed for LA in the review period 2225 (11%) had detectable levels of legionellae equal to or greater than 10 CFU/ml. The highest isolation detected during the review period was 270,000 CFU/mL. Using the groupings described by the NSW Health Department, the breakdown of legionellae levels for the positive samples is shown in Table 1.
The NSW Health Department guidelines classify Legionella levels as follows: 10 to 99 CFU/mL as an indicator that maintenance practices may not be satisfactory, 100 to 1,000 CFU/mL as potentially hazardous and levels greater than 1,000 CFU/mL as a serious situation requiring immediate shutdown of the system and decontamination. As shown in Table 1, over half of the positive samples had legionellae detected at potentially hazardous or serious contamination levels. During the review period 97.1% of the positive samples had Legionella pneumophila and 2.9% had Legionella species (not pneumophila), seven samples had detectable levels of both groups isolated. Pathogenic legionellae have been reported in L. pneumophila serogroup 1, L. pneumophila serogroups 2 to 14 and Legionella species (not pneumophila) groups[7,12,17] and therefore the risk assessment at present should solely be based on the concentration reported and not on the species or serogroup of the isolate reported.
The monthly isolation rate ranged from 2.4% (October 1993) to 20.2% (January 1995). As shown in Graph 1. the isolation rate for legionellae peaked in the summer and autumn months.
Relationship of Legionella Isolation to Plate Count Level
There is some confusion to the acceptable levels of PC count results of cooling tower water. The Victorian Health Department guidelines state that levels in excess of 500,000 CFU/mL are unacceptable, whereas the New South Wales Health Department and the Australian Standards handbook HB-32 both state levels in access of 100,000 CFU/mL are unacceptable. The Australian Standard AS3666 makes no reference to PC results and at present the Queensland Health Department does not have guidelines although the department of Workplace Health and Safety is reviewing the situation. Plate count results give a general guide on the effectiveness of biocide used and what level is used as a cut off for acceptability needs to be determined in association with the water treatment or maintenance company responsible for each cooling tower. The action taken for high plate count levels also needs to be determined. Of the 2,225 samples that had a positive LA, 1,275 had a PC also performed. Table 2 shows the break down of the PC results.
As shown in Table 2, 73% of the cooling tower water samples that had a PC performed and had detectable levels of legionellae had a PC result below all acceptable limits. Therefore it can be seen that PC analysis does not bear any relationship to the risk of colonisation by legionellae in a cooling tower. It should be noted that Legionella species and other fastidious bacteria will not grow on the media used for PC analysis and are therefore not detected. An example of this was shown by one sample that had a Legionella pneumophila level of 16,000 CFU/mL and a PC of 2,000 CFU/mL, this has also been reported by other workers. Where the PC is high, and other bacteria are not suppressed on culture media the sensitivity of LA may be reduced and a false negative result may be reported. The reduction in analysis sensitivity can be caused by bacterial overgrowth or the presence of bacteria that synthesised bacteriocins or bacteriocin-like substances that can inhibit Legionella species. Because of this and that there has been reported to be an increase in the risk of Legionnaires' disease where the Legionella/PC ratio is above 10%, there would be an advantage to have a PC performed in unison with each LA.
Testing Frequency Necessary to Reduce the Risk of High Levels of Contamination
During the review period an average of approximately one in 10 samples had detectable levels of legionellae. All cooling towers monitored were maintained as required by AS3666 and, in addition, performed three monthly tower cleans. It could therefore be extrapolated that maintaining a cooling tower to present regulations will not prevent the contamination of cooling tower water with legionellae. Further as over half of the positive samples had the bacteria in levels of either potentially hazardous or serious concentrations, it could be concluded that these regulations do not prevent high level contamination from occurring. Since it appears that the contamination of cooling towers with legionellae is a common occurrence, then steps need to be taken to reduce the risk of a disease outbreak. This could be achieved by regularly monitoring of legionellae levels and reducing the chance of contaminated aerosol dispersion. The later can be achieved by review of the physical characteristics of a cooling tower such as, position in relation to air intake ducts and exposure to general hospital staff and patients and the proficiency of drift eliminators used to reduce aerosols from escaping into the atmosphere. The decision on testing frequency should be based on review of these characteristics. Testing every three months represents a random check usually performed just prior to the next scheduled tower clean. If a tower becomes contaminated shortly after the clean then there will be a period of up to 12 weeks during which the legionellae concentration could reach high levels. This was shown by one cooling tower which was routinely tested every three months and the levels increased from <10 CFU/mL (October 1994) to 40,000 CFU/mL (January 1995) in two consecutive samples.
A survey of one customer (a large property management group based in Brisbane) who monitor all their cooling towers on a monthly basis (as compared to every three months for most other customers) showed that during 1994 they had 92% of their Legionella positive towers at the lowest risk level (10 to less than 100 CFU/mL) and no isolations in the highest risk category (above 1,000 organisms/ml). This compares to a company average of positive samples during the same period, of mostly three monthly tested towers, which had 56% in the lowest risk category and 12% in the highest risk category. It should also be noted that the overall isolation rate of Legionella species from water samples from this customer was similar to the company average (12.5% and 10.6% respectively). That is, the frequency of a positive sample was similar to three monthly tested cooling towers but the concentration of a positive detection was significantly lower (p=<0.001 by Chi squared test). The only way to minimise the risk of a high level of contamination is to do so monthly. At the very least, if three monthly tests are to be performed then the sample should be taken six weeks after the tower clean and not at the end of the three monthly period. This will reduce the chances of long periods of contamination which could go undetected. Realising that contamination of air intake ducts can occur not only by the cooling towers situated within the hospital complex but also by adjacent cooling towers off site (200 metres to a three kilometre radius) further complicates the situation. It would seem that tighter governmental controls in the form of legislation regarding maintenance and routine testing are necessary to significantly reduce the risk of a disease outbreak.
Surface Testing for Filter Efficiency and Duct Cleanliness
There are no Australian Standards for the evaluation of aerial or surface microbiological counts in relation to quality of filter efficiency or the cleanliness of airconditioning ducting. A number of companies and laboratories offer duct testing services. Unless the results of these are referenced to an acceptable standard and results are in a numeric form with standard units of expression (CFU/cubic metre of air or CFU/square centimetre of surfaces) the results can be meaningless. The finding of a wide variety of fungi and bacteria is not uncommon and may not necessarily represent a health risk. One unpublished guideline has suggested a combined fungal and bacterial count in excess of 1,000 CFU/m3 in air and 500 CFU/cm2 on surfaces indicates that an airconditioning system is in need of cleaning. Possibly the best method would be to use surface testing as an in-house reference and by testing surfaces every three to six months a graphical log could be generated of the degree of cleanliness in hospital areas. Sampling can be easily performed by hospital staff using a 5 cm2 template, sterile swab and a specialised transport broth. The test takes 2 days for the bacterial count and 5 days for the fungal count.
Hydrotherapy Pool Water Testing
The National Health and Medical Research Council of Australia (NHMRC) has guidelines for heated spa pools. These guidelines should be attainable routinely for heated hydrotherapy pools. These guidelines state that the PC should be less than 100 CFU/mL and that the Pseudomonas aeruginosa count should be less than 1 CFU/100mL. Water samples should be collected monthly until it is demonstrated that a satisfactory water quality is achieved and then not less than every three months. A water sample of 150 to 200mL is required in a sterile container and the analyses take 3 to 7 days.
Potable Water Quality
There are numerous drinking water outlets situated throughout the hospital area. These will include town water supply (in some situations stored in large tanks in building roof areas or supplied via chilled water bubblers) and in some institutions as purchased bottled water chilled and supplied by free mounted dispensers. All these sources are potentially capable of harbouring high levels of bacteria. The NHMRC guidelines for drinking water quality state that the PC should be less than 500 CFU/mL (or less than 100 CFU/mL for disinfected/chlorinated water) and that the total coliform count (TCC) and the faecal coliform count (FCC) should be less than 1 CFU/100mL. Some water samples from the above sources have been found to have PC in excess of 30,000 CFU/mL, TCC in excess of 300 CFU/100mL and FCC in excess of 50 CFU/100mL. Where unacceptable microbial levels are found disinfection and physical cleaning of the storage or the delivery system is required. All water dispensers need to have a regular maintenance schedule. The testing frequency suggested for bottled water dispensers is every 2 months after initial testing shows acceptable results. For water bubblers testing every 6 months is recommended and for town water tap supply, every 3 months if it has a roof storage tank and every 6 months if supply is direct. Approximately 250mL of water is required in a sterile container to perform the three necessary tests and results would be available in 2 to 5 days.
Legionellosis as a nosocomial infection has been reported associated with hospital potable water supplies[2,5,7]. The factors contributing to the contamination of hospital water distributions include large volume hot water tanks, age of heater and water temperature at faucet. In a given hospital, Best et al. reported that legionellosis occurred whenever L. pneumophila was recovered from >30% of selected sites. Alary et al. reported that large volume hot water tanks (7,005 " 761 litres), mean water temperature at the faucet after 3 minutes of 51.6 " 1.4oC and age of the oldest water heater (28.2 " 1.8) were contributing factors to colonisation by legionellae. Possibly larger water distribution systems are more susceptible to contamination than smaller ones because stagnation is more likely. The relationship between temperature and Legionella contamination relates to the requirement of a temperature of 60oC as measured at the faucet as optimal for growth inhibition. In older heaters the accumulation of sediments or slime that creates an adequate environment for the growth of legionellae and other bacteria species that grow in symbiosis with L. pneumophila is more likely. Hospital hot water systems need to be assessed and, in deemed necessary, a suitable screening program established to monitor for legionellae contamination. Samples need to be collected in 70mL sterile containers and the analysis takes 5 to 10 days, negative reports are issued after the required 10 day incubation period.
Kitchen Hygiene Monitoring of Food Preparative Surfaces
The kitchen facilities of a hospital must be maintained at the highest hygiene level possible. This includes the food preparation surfaces, equipment, utensils and crockery. The American Public Health Association guidelines state that for adequately cleaned surfaces the PC should be less than 2 CFU/cm2 and for utensils the PC should be less than 100 CFU/utensil. There are no guidelines or any real necessity to test for specific pathogens. This monitoring system provides quantitative data to evaluate the surface disinfection procedures presently used and can be incorporated into the evaluation of future disinfectants prior to purchase. Also, by routinely monitoring food preparation surfaces and utensils infection control practitioners can ascertain whether the bacterial build up is excessive. Unacceptably high bacterial levels can lead to food spoilage and could present a health risk to patients.
The collection procedure is the same as described for airconditioning surfaces and can be performed by hospital staff. The method is specifically designed for flat surfaces, but it can be used for testing of unmeasured surface areas such as utensils. In this case the entire utensil surface must be swabbed and care must be taken to prevent contamination by handling during swabbing. Surfaces are usually tested prior to use after disinfection has been performed. This commonly is performed at the start of the morning shift. There is no need to test surfaces while in use as these will routinely have high bacterial loads. The usefulness of this test is to monitor cleaning and disinfection procedures. The frequency of testing usually one a month of a number of areas so that every surface/utensil type is monitored once every 6 to 12 months.
Food Quality Monitoring
Most foods that have not been subjected to a severe heat treatment will contain large numbers of living organisms. In some cases many different types will be present. Some are desirable, while others may be classified as spoilage, indicators of faecal contamination or pathogenic. It is not cost effective to monitor each batch of food for every possible pathogenic microorganism. The Australian Food Standards Code is not complete and only covers a small number of indicator and pathogenic bacteria for the different food groups. The reported Escherichia coli contamination of manufactured sausage would not have been detected by adhering to the AFSC as this group of bacteria are not required to be tested under the manufactured meat microbiological requirements. For this food group only coagulase positive staphylococci and Salmonella are required to be tested. To formulate a more complete list of guidelines for prepared food the assistance was sought of the Queensland Health Departments of the Food Unit and the microbiology laboratory. Table 3 shows the combined suggested guidelines for food analyses using the AFSC and the "Guidelines for Ready to Eat Foods" from the Food Unit.
A sample size of 20 to 100 grams collected into a sterile container is required to perform the analyses. The reporting time varies from 2 to 5 days depending on the analysis type. It is impractical due to monetary restraints to perform all the necessary analyses of each batch of prepared foods. However a testing regime should be established that will perform each test on each food group once over a period of time. This time period will vary from one institution to another depending on available funds.
Choice of Testing Laboratory
Private or public hospital laboratories are NATA registered as medical testing facilities and do not have registration to enable the issue of NATA endorsed documents for the above mentioned analyses unless they have additional registration under biological testing. The choice of a laboratory should be primarily driven on the quality of the testing procedures. By utilising NATA registered biological testing laboratory there is the safeguard that all tests are performed with the relevant quality control procedures to Australian or International standards and that these laboratories are regularly monitored by NATA biological inspections.
Historically the role of the infection control practitioner has been primarily related to medical/surgical procedures and ancillary services which directly contribute to nosocomial infection. Without the creation of a specific section to deal with the microbial quality of the total environment including airconditioning, hydrotherapy pools, drinking water, kitchen preparation areas and food, quality control may well remain with the infection control department. This would seem advantageous as contamination of these facilities could develop into a nosocomial infection on a major scale. The role of the biological testing laboratory has rarely been regarded as a service provider to the infection control team. But it would now seem beneficial to include an environmental/food microbiologist as a regular contributor to infection control meetings. With the introduction of total quality management and quality systems certification to AS/NZS/ISO 9000: 1994 series there is the opportunity for infection control practitioners to widen their scope of activities to encompass proactive monitoring of the less traditional sources of nosocomial infection.
1. Alary, M. and J. R. Joly. 1992. Factors contributing to the contamination of hospital water distribution systems by Legionellae. J. Infect. Dis. 165:565-569.
2. Alary, M. and J. R. Joly. 1993. Occurrence of nosocomial Legionnaires' disease in hospitals with contaminated potable water supply. p39. In J. M. Barbaree, R. F. Breiman and A. P. Dufour (ed.), Legionella: Current status and emerging perspectives. American Society for Microbiology, Washington, D.C.
3. American Public Health Association. 1992. Standard methods for the examination of water and wastewater, 18th ed. p. 9-36 - 9-38. American Public Health Assoc., Washington D.C.
4. Best, M, V. L. Yu, J. Stout, A. Goetz, R. R. Muder and F. Taylor. 1983. Legionellaceae in the hospital water supply. Epidemiological link with disease and evaluation of a method for control of nosocomial Legionnaires' disease and Pittsburgh pneumonia. Lancet 2:307-310.
5. Breiman, R.F. 1993. Modes of transmission in epidemic and nonepidemic Legionella infection: Directions for further study. p 30-35. In J. M. Barbaree, R. F. Breiman and A. P. Dufour (ed.), Legionella: Current status and emerging perspectives. American Society for Microbiology, Washington, D.C.
6. Cameron, R., C. Walker, M. Beers, N. Rose and E. Anear. 1995. Enterohaemorrhagic Escherichia coli outbreak in South Australia associated with the consumption of metwurst. Comm Dis Intell. 19:70-71.
7. Garcia, M., I. Campbell, P. Tang and C. Krishnan. 1993. Incidence of Legionnaires' disease and report of Legionella sainthelensi serogroup 1 infection in Toronto, Canada. p. 37-39. In J. M. Barbaree, R. F. Breiman and A. P. Dufour (ed.), Legionella: Current status and emerging perspectives. American Society for Microbiology, Washington, D.C.
8. Gomez-Lus, R., E. Lomba, P. Gomez-Lus, M. S. Abarca, S. Gomez-Lus, A. Martinez, E. Duran, and M. C. Rubio. 1993. In vitro antagonistic activity of Pseudomonas aeruginosa, Klebsiella pneumoniae, and Aeromonas spp. against Legionella spp., p. 265-267. In J. M. Barbaree, R. F. Breiman and A. P. Dufour (ed.), Legionella: Current status and emerging perspectives. American Society for Microbiology, Washington, D.C.
9. Health Department, New South Wales. 1991. Code of practice for the control of Legionnaires' disease. p. 35. Health Department, New South Wales.
10. Health Department, Victoria. 1989. Guidelines for the control of Legionnaires' disease. Environmental Health Standards. Health Department, Victoria.
11. Kirby, B. D. and A. A. Harris. 1987. Nosocomial legionnaires' disease. Sem Respir Infect. 2:255-61.
12. Marston, B. J., H. Lipman and R. F. Breiman. 1994. A decade of surveillance for Legionnaires' Disease: update on risk factors for morbidity and mortality due to infection with Legionella. Arch Inter Medic; 154 (21):2417-22.
13. Miller, R. D. and K. A. Kenepp. 1991. Aerobiology. Abstr. Pan-Am. Aerobiol. Assoc. Annu. Meet., p. 20. Pan-American Aerobiology Association. Washington D.C.
14. National Food Authority. 1992. Australian food standards code. Australian government publishing service. Canberra.
15. National Health and Medical Research Council. 1989. Australian guidelines for heated spa pools. p. 2. Australian government publishing service. Canberra.
16. National Health and Medical Research Council and Australian Water Resources Council. 1987. Guidelines for drinking water quality in Australia. p.7-8. Australian government publishing service. Canberra.
17. Pearce, M. 1992. Legionellosis in Queensland, A review. Comm Dis Intell. 16:296.
18. Shelley, A. 1992. Correspondence. Microbiological contamination of airconditioning systems.
19. Standards Australia. 1989. AS3666, air-handling and water systems of buildings, microbial control. Standards Australia. Sydney.
20. Standards Australia. 1991. AS3896, waters, examination for legionellae. Standards Australia. Sydney.
21. Standards Australia. 1992. HB-32, control of microbial growth in air-handling and water systems in buildings. p. 40. Standards Australia. Sydney.
22. Stout, J. E., V. L. Yu and M. G. Best. 1985. Ecology of Legionella pneumophila within water distribution systems. Appl Environ Microbiol. 49:221-228.
23. Thacker, B. S., J. V. Bennett, T. F. Tsai et al. 1978. An outbreak in 1965 of severe respiratory illness caused by the legionnaires' disease bacterium. J Infect Dis 138:512-9.
24. Vanderzant, C. and D. F. Splittstoesser. 1992. Compendium for the microbiological examination of foods. (3rd Ed.) p. 59. American Public Health Association. Washington D.C.