Summary

Two hundred and one cases of legionnaires’ disease were reported to the PHLS Communicable Disease Surveillance Centre in 1996. Twenty-four cases (12%) were known to have died. One hundred and one cases were associated with travel, either abroad or in the United Kingdom. Two cases acquired infection in hospital, the smallest number ever reported, and the remaining 98 were presumed to have acquired infection in the community. Fifty-five (27%) of the 201 cases were linked to outbreaks or clusters and the remaining 146 (73%) were reported as single cases. Six outbreaks were associated with industrial premises. Twenty-two of the travel associated cases were part of three travel outbreaks and six clusters. The proportion of cases diagnosed by detection of urinary antigen has continued to increase and in 1996 this method of diagnosis was used for 43% of the cases.

Introduction

Legionnaires' disease is a pneumonic illness associated with environmental sources such as hot or cold water systems or cooling water systems in large buildings, such as hotels, hospitals, or industrial premises. An outbreak of pneumonia among delegates attending a Legionnaires' convention at a hotel in Philadelphia, United States (US) in 1976 gave rise to identification of the bacterium responsible (Legionella pneumophila) and the name of the disease 1 . The disease is transmitted through inhalation of aerosols containing legionella organisms. A National Surveillance Scheme for Legionnaires’ Disease for residents of England and Wales was set up in 1979 by the PHLS Communicable Disease Surveillance Centre (CDSC). Data have been collected each year since 1980 and 3005 cases of legionnaires’ disease in residents of England and Wales had been reported to the scheme by the end of 1996. Overall, travel and community cases have each accounted for 46% of the cases and hospital acquired infections the remaining 8%. The annual totals of reported cases fell between 1989 and 1991, following a peak of 279 cases reported in 1988. Since 1993 annual totals have been increasing again and the 201 cases reported in 1996 was the highest total recorded since 1989.

Methods

Legionnaires’ disease is not a statutorily notifiable disease in England and Wales: cases are therefore reported on a voluntary basis by PHLS, NHS, or private hospital laboratories, consultants in communicable disease control (CCDCs), and other health care workers. CDSC sends a follow up questionnaire to those who report cases asking for clinical, epidemiological, and microbiological information about each case, particularly looking for evidence to show whether the patient's illness was associated with travel, hospital, or occupational exposure during the disease incubation period (2 to 10 days). The PHLS Legionella Reference Unit of the Respiratory and Systemic Infection Laboratory at the PHLS Central Public Health Laboratory sends microbiological confirmation of diagnoses to CDSC. These data are entered onto the national database, which is then searched for other cases that may be linked in time or place or for premises with which cases may have been associated in the past.

CDSC uses the following case definitions for legionnaires’ disease and other forms of legionellosis for surveillance purposes 2 . To meet the case definition for legionnaires’ disease a patient must have clinical or radiological evidence of pneumonia and laboratory evidence of legionella infection. Cases diagnosed by culture of the organism from respiratory specimens (sputum or lung tissue or necropsy specimens), those in whom a fourfold rise in serum antibody levels against L.pneumophila is observed, and, since
1 January 1996, cases diagnosed by detection of L. pneumophila antigen in urine are said to be confirmed. A single high antibody titre with a compatible pneumonic illness defines a presumptive diagnosis of legionnaires’ disease.

Information about other forms of legionellosis – the non-pneumonic, asymptomatic, or suspected cases – are also collected by the scheme and are reported separately inthis review.

Information about cases defined as 'travel associated' – that is, those who stayed in hotel or holiday apartment accommodation in the United Kingdom (UK) or abroad during the ten days before becoming ill – is sent to the European Surveillance Scheme for Travel Associated Legionnaires’ Disease. This scheme, which is funded by the European Commission Directorate General V, has been coordinated by CDSC since 1993 on behalf of the European Working Group for Legionella Infections (EWGLI)3 . Twenty-five European countries currently take part in the scheme and report travel associated cases to CDSC. These cases are entered into a database, which is then searched for other cases linked to the same place of accommodation. CDSC immediately informs the EWGLI collaborator in the presumed country of infection about single cases. All collaborators and the World Health Organisation (WHO) are contacted when clusters are detected. WHO in turn is asked to inform the ministry of health in the country concerned.

A cluster is defined as two or more cases who became ill within six months of each other and who stayed at the same accommodation in the ten days before becoming ill. The term outbreak’ is reserved for clusters for which there is comprehensive information about the results of epidemiological and microbiological investigations, which have been carried out by or in liaison with EWGLI collaborators.

Since June 1996, CDSC has also informed the UK Federation of Tour Operators (FTO) and the British Association of Travel Agents (ABTA) of incidents of travel associated legionnaires’ disease. Anonymised travel histories for all single cases in residents of England and Wales and all clusters that include residents of the UK and/or other European countries are sent to the FTO and ABTA so that hotels where single cases have stayed may be reminded to maintain adequate control measures against legionella in their water and air conditioning systems. When clusters occur, the tour operators may ask for environmental investigations to be carried out at the hotel and may also withdraw their clients from the hotel until they are satisfied that all control measures have been implemented successfully.

Epidemiological

Two hundred and one cases were reported in 1996 (figure 1), with a median age of 55 years. One hundred and fifty-six (78%) of the cases were men, with an age range of 18 to 88 years and 45 cases (22%) were women (aged 26 to 85 years). Twenty-four cases were reported to have died as a result of their infection, 165 recovered, and the outcome of 12 cases was either not known or not reported. The case fatality rate, 12%, was slightly lower than the 13% reported in 1995. A further 12 cases, about whom incomplete information (no date of onset or clinical or epidemiological data) was received, were lost to follow up and are not included.

Microbiological

The methods used to diagnose legionella infection are shown in table 1. One hundred and forty-three infections were diagnosed by a single method, 42 by two methods,

and 16 by three methods. Legionellas were isolated from specimens obtained from only 31 cases (15%). Twenty-eight of these were identified as L.ipneumophila serogroup (sg) 1 and three as L. pneumophila sg 6. L. pneumophila infections were confirmed in 82 cases (41%) by seroconversion (a fourfold or greater rise in antibody titre obtained using L. pneumophila sg 1 yolk sac antigen), 52 (26%) were confirmed by the detection of L. pneumophila sg 1 antigen in urine, and 36 (18%) were diagnosed presumptively by a single high antibody titre. In total, 86 cases (43%) had a positive L. pneumophila sg 1 urinary antigen detection test (with or without an dditional method of diagnosis). Thus 95 cases (47%) were known to have been caused by L. pneumophila sg 1 (culture proven or L. pneumophila sg 1 urinary antigen) and three (1%) to be caused by L. pneumophila sg 6. Although presumed to be caused by L. pneumophila the infecting serogroup of the 103 cases (51%) identified by serology only could not be determined with any confidence as different serogroups of L. pneumophila can share common antigens. Falsely positive L. pnemophila sg 1 serological results due to campylobacter infection have been reported 5 . This possibility was excluded for all 87 patients from whom serum specimens were available by examining them in the presence of ‘campylobacter blocking fluid’ 5 .

Travel associated infection

One hundred and one cases (75 men and 26 women) reported having travelled and stayed away from home overnight in the ten days before becoming ill with legionnaires’ disease. Six men and three women died as a result of their infection. Sixty-eight per cent of the male and 58% of the female travel cases were aged between 50 and 69 years. The commonest months in which cases became ill were June, September, and October. Ninety-six cases were associated with travel abroad:

The commonest destinations were Spain, France, Greece, the Caribbean, Italy, the US, and Turkey. Five cases were linked to travel in the UK (table 2). Three outbreaks and six clusters affecting 22 residents of England and Wales were detected in 1996.

Outbreak I

In March 1996 a party of 34 people from Wales went on a skiing holiday for nine days in Vermont, US. On return, 11 people had respiratory illnesses and three were admitted to hospital. Clinical histories and details of exposure to risk factors were obtained from the entire group and 30 were investigated serologically. One case of legionnaires’ disease was confirmed by a fourfold rise in antibody levels, one other had a compatible illness but was not microbiologically confirmed, one had a non-pneumonic illness, and four had serological evidence of legionella infection but remained well. The party stayed in two adjacent chalets and illness was associated with use of a swimming pool and whirlpool spa in one of the chalets. In the same month another British tourist with travel associated legionnaires’ disease (confirmed by a fourfold rise in antibody titre) was reported to CDSC. This person had also stayed at and used the same two chalets as those of the Welsh group. Public health officials in the US reported that environmental investigations and control measures had been carried out at the two chalets, but that no legionellas had been detected from the whirlpool spa or other water systems and that no other tourists reported respiratory illness.

Outbreak II

On a single day in May 1996, CDSC received three separate reports of travel associated legionella infection in people who had recently stayed at the same hotel apartment block in Minorca, Spain. Details of the cluster were sent immediately to all EWGLI collaborators, WHO, and to the one British tour operator who used these premises exclusively. The tour operator closed the hotel within 12 hours of being informed and moved the guests to other accommodation. Environmental investigations by Spanish health officials revealed that a new hot water system had been installed at the premises in the first three weeks of May and that hot water temperatures had been inadequately maintained while the work was in progress. Legionella spp. were isolated from the water system at the hotel but were not available for further analysis. No clinical isolates were obtained, but urine specimens from two cases were positive for L. pneumophila sg 1 antigen.
Altogether four cases were linked to this outbreak and all became ill between 11 and 24 May. The hotel reopened when all control measures had been satisfactorily completed and no further cases have been reported.

Outbreak III

The first outbreak of legionnaires’ disease ever associated with the Caribbean island of Antigua was detected in October 1996. Two cases occurred in September and one in October 1996. All had stayed at a hotel complex in Mamora Bay. The PHLS and the Caribbean Epidemiology Centre (CAREC) were invited by the Ministry of Health of Antigua and Barbuda to carry out environmental and epidemiological investigations into the source and extent of the outbreak. The hotel’s sickness records showed that the incidence of respiratory illness was higher than expected among staff in August and among guests in September, suggesting that the outbreak had been triggered by a single event. This event was found to be the installation in August of a solar powered hot water system at the hotel. Thirty-one out of 54 hot and cold water samples from the hotel’s water systems were positive for legionella and 19 of these contained L. pneumophila sg 1. Isolates from the hot water tap in a room where one of the cases stayed, the solar hot water storage tank drain, the gymnasium shower, and the spa pool filter housing were all indistinguishable by monoclonal antibody (mAb) subgrouping and restriction fragment length polymorphism (RFLP) typing from a clinical isolate obtained from one of the cases in the outbreak. A serological survey of 11 of the 35 hotel staff who reported a respiratory illness between August and November revealed two people who were retrospectively diagnosed as presumptive cases of legionnaires’ disease with single high titres of antibody against legionella. A questionnaire was sent to 21 guests from the UK, France, Germany, and the US who had reported a respiratory illness while at the hotel in September. All were asked to take part in the serological survey but none was found to have evidence of legionella infection. Several national and international tour operators stopped sending guests to the hotel when the outbreak was identified in November and did not return to the hotel until December when all investigations and control measures were completed.

The clusters

Six clusters were detected in 1996, and included 13 cases from England and Wales. Two of the clusters were associated with hotels in Spain, to one of which a case had been linked in 1990.
In France, three cases from England and Wales and one case from each of Australia, France, and the Netherlands were linked to a large hotel near Paris between May and October. L. pneumophila sg 1 was isolated from one of the cases and the hotel’s hot water system was highly contaminated with L. pneumophila sg 1.
Two cases were associated with a hotel in Italy, to which two cases had been linked in 1995. The hotel was investigated after both incidents. No legionellas were detected in 1995 and L. pneumophila sg 3 was detected in 1996. The cases in 1996 were diagnosed serologically and the infecting serogroup could therefore not be determined with any confidence. Two cases were linked to apartments in Greece, and two to a hotel in the Algarve, Portugal.
FTO and ABTA were informed of all clusters. British guests were withdrawn from the apartments in Greece until satisfactory control measures had been implemented.

Other travel cases

Seventy-nine other cases acquired infection away from home. Nine had travelled to more than one country in the ten days before developing legionnaires’ disease and five had stayed at camp sites, hotels, or other accommodation in England and Wales. Two cases were part of clusters detected by the EWGLI scheme. One had stayed at a hotel in South Africa along with a Danish resident who also became ill and one had stayed at a hotel in Spain where a Scottish case had also stayed before becoming ill. Two cases were linked to cruise ships and two to the Dominican Republic. The remaining single cases were associated with travel to northern and southern Europe, the Middle East, the Far East, India, Canada, North America, Central America, and Bermuda.

Community acquired infection

Ninety-eight cases (81 men and 17 women) were associated with infection from industrial or community settings in 1996, the highest number since 1989. Fifty-seven per cent (55) were aged between 40 and 59 years; 13 men with a mean age of 48.5 years, and two women aged 35 and 56 years died from their infection. There was no seasonal pattern to the community cases; similar numbers of cases occurred throughout the year except for August and December when outbreaks were detected.
Thirty-three cases (34%) were linked to six outbreaks.
Two outbreaks were associated with the same industrial estate. All outbreaks were shown to be due to L.ipneumophila sg 1 either by urinary antigen detection or by culture of a clinical specimen from the patients.

Outbreak I

An outbreak occurred at a plastics factory with two sites in the same town in Wales. Four cases of legionnaires’ disease, one fatal, were reported in employees of the factory over a period of five months. A positive urinary antigen test for three patients confirmed that L. pneumophila sg 1 was the causative agent. All cooling towers at both sites were inspected, environmental investigations were carried out, and control measures were implemented. A cohort study of 200 employees, which included testing blood for L. pneumophila antibody, showed that 17 employees had evidence of past or recent infection, ten with a presumptive or suspected non-pneumonic legionellosis and seven with asymptomatic infection. The four cases of legionnaires’ disease and the employees with serological evidence of infection had all worked for most of their time at one of the two sites belonging to the factory. Epidemiological data implicated the cooling towers at this site as the probable source of infection but this was not confirmed microbiologically.

Outbreaks II and III

In August an outbreak consisting of 14 cases was linked to a large industrial estate in Northamptonshire. Cooling towers from 15 premises on the estate were inspected and L. pneumophilia was isolated from water samples from towers from seven premises. Control measures were taken throughout the estate but in December 1996 and January 1997 a second outbreak of six further cases, including one who died, was linked to the same estate. In February 1997 a sporadic case occurred, which could not be linked to the outbreak. The investigations were widened to include other industrial premises and other possible sources of infection such as car washes. In this second outbreak legionellas were isolated from cooling towers from five premises on the estate. Two of these premises had yielded positive results when investigated during the summer. An isolate of L. pneumophila sg 1 from a respiratory tract specimen from one of the winter cases was indistinguishable by mAb and RFLP analysis from environmental isolates obtained from cooling towers at three of the industrial premises, one of which had been positive in both the first and the second outbreaks.

SEE OUTBREAKS AT CORBY

Outbreak IV

An outbreak associated with an industrial site in the West Midlands occurred in October and November 1996. Five cases of pneumonic and two of non-pneumonic legionellosis were identified, all of whom worked at three separate foundries. Cooling towers at all premises in the vicinity were inspected, sampled, and tested, including one that was not registered with the local authority. One tower, on premises where none of the cases worked, was positive for L. pneumophila sg 1. No isolates were obtained from the cases so clinical and environmental strains could not be compared but L. pneumophila sg 1 antigen was detected in urine from two of the cases.

Outbreak V

In August 1996 a worker at a small plastics factory on an industrial estate in Yorkshire developed legionnaires' disease. A second case occurred in the same workforce six months later. The factory had no wet cooling system, but there were cooling towers on adjacent sites. Water from an uncovered, outdoor tank on the site was used to cool machines in the plastics factory and passed through sight glasses on its way to the machines to confirm its flow. The glass was cracked and created a fine aerosol into the work area. Water from this tank was positive for L. pneumophila sg 1 when tested after the second case was diagnosed. Although no clinical isolates were obtained, L. pneumophila sg 1 antigen was found in urine from one of the patients. Control measures at the factory included replacing the cracked glass, cleaning the tank, treating the water with a bactericide, and covering the tank.

Outbreak VI

A plastics factory in Trent was the source of another outbreak of legionnaires’ disease consisting of two cases. A member of the workforce became ill and died in October 1996 and a retired man who lived near the factory developed legionnaires' disease in November 1996. An unregistered cooling tower at the factory yielded an isolate indistinguishable on mAb subgrouping and RFLP subtyping analysis from an isolate obtained from the fatal case.

Hospital acquired infection

Two single cases of hospital acquired legionnaires' disease were reported in 1996. The first was a psychiatric inpatient who had been admitted 23 days before becoming ill. The second was a patient with impaired immunity and vasculitis due to renal disease who was in hospital for several weeks before developing legionellosis. Both patients recovered from their legionella infections.

Discussion

There was a sharp increase in the number of cases of legionnaires’ disease reported to CDSC in 1996 – 201, compared with 160 in 1995. Cases associated with travel abroad accounted for the second highest number of travel cases reported since 1980, and community acquired cases the largest number since 1989. It is not known why the numbers of cases increased, whether perhaps through improved ascertainment, and/or a decline in maintenance standards at industrial premises.

Twenty-seven per cent of the cases in 1996 were linked to outbreaks or clusters compared with 24% in 1995 4 .

Community acquired cases in England and Wales were predominantly male, tended to be younger, and had a higher case fatality rate than travel associated cases.

Seventy-seven per cent of the 101 cases associated with travel had visited one or more European countries before becoming ill. The remaining cases were linked to travel to other increasingly popular tourist destinations such as the Caribbean, South Africa, and Central America. Antiguan public health officials had no experience of dealing with legionella infection when their outbreak was detected. After the outbreak had been controlled, the island’s environmental health officers, hotel managers and medical staff took part in a series of seminars aimed at promoting good practice for maintaining the island’s hotel water systems and for recognising and reporting cases of legionella infection in tourists. Reporting links between CDSC and CAREC were strengthened as a result of the outbreak. All cases that originate from the Caribbean are now reported to CAREC for immediate follow up at the hotels concerned.

Reporting cases of travel associated legionnaires’ disease to tour operators in the UK has been shown to be both practicable and effective, despite a few local difficulties (complaints, denials and threats from hotel managers) in some countries. The FTO and ABTA, which represent nearly all package holidays sold in the UK, are now able to alert tour operators to the hotels where cases have stayed. The report of a single case does not imply that the hotel named is the source of infection.

Nevertheless, informing the hotel of the report and making sure its managers follow the recommended guidelines means that precautionary measures are taken systematically to maintain safe systems. Vigilance and awareness of what would happen to hotels whose water systems were not maintained adequately was an important message that was made clearly.Three of the cluster alerts from England and Wales resulted in tour operators implementing an immediate withdrawal of guests from hotels or apartments. Two other withdrawals took place in Turkey, following cluster alerts from the EWGLI coordinating centre at CDSC. Investigations at four of these sites identified problems with either the hot and cold water systems or the cooling towers used for air conditioning, and environmental samples from all four sites yielded L. pneumophila. In Turkey, the Ministry of Tourism stated that the hotels would not be allowed to reopen until all control measures were implemented satisfactorily. The combined actions of EWGLI collaborators, local public health officials, the national and international federation of tour operators, and the Ministries of Tourism and Health provide the best opportunity for appropriate and swift action to control and prevent travel associated legionnaires’ disease 6 .

Community acquired infection has often been associated with urban rather than rural areas 7 . All six community outbreaks in 1996 occurred not just in urban environments but specifically in industrial premises. Furthermore, 45 of the 64 sporadic cases lived in industrial towns or cities and 11 of the 32 cases whose occupation was reported also worked in industrial settings, seven of them in the plastics or chemical industries. The reasons for this apparent preponderance of industrial cases in 1996 are unclear but may be linked to the growth of small industrial estates, many of whose premises operate their own cooling towers. In addition newer industries such as plastic manufacture may make greater use of water cooled equipment as well as cooling towers. Although all cooling towers are legally required to be registered with the local authority 8 , unregistered cooling towers were identified in some investigations. Several systems were found to be poorly maintained, and there was evidence of poor practice on the part of some water treatment companies. Although the number of large industrial outbreaks has fallen, the large number of small outbreaks in 1996 suggests that legionnaires’ disease remains a major industrial problem 9,10 . Continuous effective maintenance of water and cooling plant is clearly essential to minimise the risk of infection from legionella 11,12 . Legionnaires' disease is not a notifiable disease in England and Wales but many public health officials would like the disease to become so, as has been the case in Scotland since 1988. When the subject was last debated by the Government in 1989/90 it was concluded that informal reporting links were well established and that notification would not necessarily improve information about cases 13 . It is known, however, that investigations at industrial sites are sometimes carried out by environmental health officials or health and safety staff without first informing the local CCDC. This delays the start of epidemiological investigations, reporting the outbreak to CDSC, and searching for related cases. A review of public health law is expected and will offer an opportunity to reconsider the arguments for making legionnaires’ disease notifiable, taking these and other issues into account. Past studies have identified that legionnaires’ disease is underdiagnosed and underreported 14 . It is now well established that the use of highly specific enzyme immunoassays (EIAs) for the detection of L. pneumophila antigen in urine allows legionellosis to be diagnosed early in the course of infection 15,16 . The PHLS Legionella Reference Unit has used an in-house EIA since 1988 and commercially produced EIAs have become available more recently. Consequently, after data collected on cases referred in 1995 were reviewed, case definitions for England and Wales were changed for 1996 to include cases established using this method as confirmed rather than presumptive. Although this is consistent with changes in definitions made by the Centers for Disease Control and Prevention in Atlanta in 1995 15 , it puts the diagnostic criteria used in England and Wales at variance with those used by WHO and for surveillance in Europe 3,17 . A debate ontinues within EWGLI, which may result in revised criteria and recommendations to WHO to revise their definitions.

It is clear that urinary antigen detection is a valuable diagnostic method, which is leading to earlier diagnosis and treatment for people with suspected legionellosis. Wider use of this test in patients diagnosed with community acquired pneumonia or who present with a pneumonic illness after recent travel should lead to an improved estimate of the true incidence of cases. Making the disease notifiable should improve local reporting of cases to CCDCs and thence to CDSC for epidemiological follow up.

Acknowledgements

The authors are grateful to microbiologists, CCDCs, infection control nurses, and others for their continued support of the surveillance scheme. They also thank Dr John Lee for providing information about environmental investigations.

References

1. CDC. Follow-up on respiratory illness - Philadelphia.
MMWR Morb Mortal Wkly Rep 1977; 26: 9-11.

2. Saunders CJP, Joseph CA, Watson JM.
Investigating a single case of legionnaires’ disease: guidance for consultants in communicable disease control.
Commun Dis Rep CDR Rev 1994; 4: R112-4.

3. Hutchinson EJ, Joseph CA, Bartlett CLR. EWGLI:
aEuropean surveillance scheme for travel associated legionnaires’ disease.
Eurosurveillance 1996; 1: 37-9.

4. Newton LH, Joseph CA, Hutchinson EJ, Harrison TG, Watson JM, Bartlett CLR.
Legionnaires’ diseasesurveillance: England and Wales, 1995.
Commun Dis Rep CDR Rev 1996; 6: R151-5.

5. Boswell TC, Marshall LE, Kudesia G.
False positivelegionella titres in routine clinical serology testing detected by absorption with campylobacter: implications for the serological diagnosis of legionnaires' disease.
J Infect 1996; 32: 23-6.

6. Cartwright RY.
Who is responsible for preventing infection in tourists?
Proceedings of the second conference on international travel medicine;
May 9-121991, Atlanta: 309-10.

7. Bhopal RS, Fallon RJ, Builst EC, Black RJ, Urquhart JD.
Proximity of the home to a cooling tower and risk of non-outbreak legionnaires’ disease.
BMJ 1991; 302: 378-83.

8. Health and Safety Commission.
The notification of cooling towers and evaporative condensers regulations
1992. Statutory Instrument, 1992 No 2225, Health and Safety. London: HMSO, 1992.

9. Bhopal RS, Barr G.
Maintenance of cooling towers following two outbreaks of legionnaires’ disease in a city.
Epidemiol Infect 1990; 104: 29-38.

10. Print J.
Coventry reduces the risks of legionella outbreaks.
Environmental Health 1995; 10: 189-90.

11. Health and Safety Executive.
The control of legionellosis including legionnaires’ disease.
London: HMSO, 1993.

12. CIBSE. Minimising the risk of legionnaires’ disease:
TM13:1991. Chartered Institution of Building Services Engineers; 1992.

13. Employment Committee.
Legionnaires’ disease: further developments.
London: HMSO, 1989.

14. British Thoracic Society.
Community acquired pneumonia in adults in British hospitals in 1982-1983:
a BTS/PHLS survey of aetiology, mortality, prognostic factors and outcome.
QJ Med 1987; 62: 195-220.

15. Plouffe J, File T, Breiman R, Hackman B, Salstrom S, Marston B, et al.
Re-evaluation of the definition of legionnaires’ disease: Use of the urinary antigen assay.
Clin Infect Dis 1995; 20: 1286-91.

16. Birtles RJ, Harrison TG, Samuel D, Taylor AG.
Evaluation of a urinary antigen ELISA for diagnosing Legionella pneumophila serogroup 1 infection.
J Clin Pathol 1990; 43: 685-90.

17. WHO. Epidemiology, prevention and control of legionellosis: Memorandum from a WHO meeting.
Bull World Health Organ 1990; 68: 155-64.

Article by
CA Joseph, TG Harrison, D Ilijic-Car, CLR Bartlett
CDR Review October 1997