Zbl. Bakt. Hyg., 1. Abt, Orig, A 255, 64-70 (1983)

Epidemic Legionellosis in England and Wales 1979-1982

C. 1. R. BARTLETT and 1. F. BIBBY

PHLS Communicable Disease Surveillance Centre, 61, Colindale Avenue,London NW9 SEQ, England

Abstract

National surveillance of legionnaires' disease in England and Wales identified 588 casesduring the years 1979-1982. The majority of cases appeared to be sporadic but 32 clustersof two or more cases were recognised, 24 in association with hotels, seven with hospitalsand one with a construction site.

Reports (3 published and 3 unpublished) on the investigation and control of six out­breaks are reviewed. All six were found to be common source outbreaks; four hotelassociated, one was nosocomial and one was associated with a construction site. Domesticwater systems were implicated as sources in the four hotels and the hospital. The mosteffective control measure was found to be continuous chlorination of the water supplycombined with raising the hot water temperature to 60 DC or more in the calorifier and atleast 55 DC to a maximum of 60 DC at outlets. Raising hot water temperatures alone wasnot always completely successful as a control measure. Only in the construction site out­break was a cooling water system shown to be the principal source of infection.

Keywords: Legionnaires' disease, surveillance, common source outbreaks, nosocomialoutbreak, water systems, control measures.

Introduction

A laboratory reporting scheme for legionnaires' disease in England and Waleswas established by the Public Health Laboratory Service in 1977, and a morecomprehensive surveillance programme was introduced in 1979. The aim is todetermine the clinical and epidemiological characteristics of legionella infections,and to detect clusters of cases to identify sources and modes of transmission. Briefdetails of the clustering recognised through this scheme are presented and reportson six incidents are reviewed.

Methods

Surveillance of legionnaires' disease in England and Wales is maintained through a na­tional laboratory reporting scheme. About 400 microbiological laboratories voluntarilyreport the identification of this and other infections once a week to the PHLS Cornmunica-

Epidemic Legionellosis in England and Wales 65

ble Disease Surveillance Centre (CDSC). Laboratories reporting legionella infections arerequested to send strains, serum or other specimens to one of five legionella referencelaboratories (Division of Microbiological Reagents and Quality Control, Central PublicHealth Laboratory, Colindale, and Cambridge, Nottingham, Oxford and Preston PublicHealth Laboratories) for confirmation of diagnosis and identification of Legionella speciesand serogroup. The criteria for confirmation are isolation of the organism, or a four foldor greater rise in immunofluorescence antibody titre or demonstration of the organism bydirect immunofluorescence staining in specimens collected from normally sterile sites. Con­valescent antibody titres of 128 or more against the Colindale formolised yolk sac antigen(Taylor et al., 1979), together with a compatible clinical history are accepted as presumptiveevidence of infection. A questionnaire is sent to the physician or microbiologist in chargeof the case by the reference laboratory or CDSC when the diagnosis is confirmed. Detailsof recent hospital admissions, travel and employment are recorded to enable the recognitionof nosocomial clusters and clusters associated with hotels and places of work. The question­naires are returned to CDSC for collation and analysis. Whenever possible, any identifiedclusters are investigated by the Public Health Laboratory Service in collaboration withlocal hospital microbiologists and public health authorities.

Results

Five hundred and eighty eight cases of legionella pneumonia in England andWales were reported during the years 1979-1982. The majority of cases appearedto be sporadic but 32 clusters of two or more cases were identified (Table 1). Innine instances four or more cases were closely clustered in time in association withsingle establishments, mostly hotels. All but two of these hotels are in tourist resortsabroad. Altogether 265 (45%) of 588 reported cases were associated with travel,either in the United Kingdom or abroad, particularly within Europe (Table 2).Accurate data on the number of British tourists visiting countries in Europe arenot available so it is not possible to give country-specific rates. The proportionsof cases that were travel-associated were similar during the years 1979-1982.

Table 1. Clusters of L.D. Associated with Single Establishments 1979-1982

Establishments No. of No. of Clusters with CasesClusters Cases in more than one year

Hotels 24 107 17Hospitals 7 26 2Other 1 3

Total 32 136 19

Indigenous cases, travel-associated cases and clusters showed a distinct seasonalpattern throughout the four years with few reports during the early months of theyear and most during the late summer and autumn. No evidence of person-to­person transmission was produced during the investigations of clusters but fre­quently environmental sources were implicated.

5 Zbl. Bakt. Hyg., I. Abt. Orig. A 255

66 C.L.R.Bartiett and L.F.Bibby

Table 2. Legionella Pneumonia and Travel: Laborarory Reports England and Wales 1979­1982

Country CasesNo Per cent

Spain 115 19Portugal 31 5Italy 30 5France 17 3Greece 13 2Other Europe 15 3Outside Europe 27 5Total abroad 248 42U.K. 17 3No history of travel 323 55

Total 588 100

Six outbreaks and the methods used to control them are briefly described below:Legionella pneumopbila serogroup 1 was the agent responsible in all six incidents.

Outbreak 1

In 1979,58 employees of one firm took part in a golfing tournament in CentralEngland. Four golfers developed pneumonia in the week after the tournament;three were shown to be cases of legionnaires' disease. All four men with pneumoniahad stayed at one hotel whereas none of 26 golfers who had stayed elsewhere wereaffected (Tobin et al., 1981). A woman who had not visited the golf course, buthad stayed at the hotel after the golfers' visit, developed legionnaires' disease oneweek later. 1. pneumopbila serogroup 1 was isolated from the hot and cold watersystems in the hotel; there was no air conditioning and no cooling water systemin the vicinity. As a control measure the water supply to the hotel was chlorinatedcontinuously so as to achieve 1-2 ppm of free residual chlorine at cold water outlets.It was not possible at that time to achieve free chlorine levels at the hot water taps.One year later these measures were reinforced by raising the hot water temperatureto 60°C or more in the calorifier. This produced temperatures of 55-60°C at tapsin all but three bedrooms. The following year a middle aged man developed legion­ella pneumonia five days after staying at the hotel. He had occupied one of thethree bedrooms in which a maximum water temperature of 48°C was recorded.Since then an additional continuous chlorination unit has been installed so as togive at least 1 ppm of free residual chlorine as measured at hot water outlets. Nofurther cases have been identified in association with the hotel.

Outbreak 2

Between September 1979 and August 1980 five cases of legionnaires' diseaseoccurred among British tourists staying at a hotel in a resort in Portugal. The hoteldid not have an air conditioning but a solar heating system was incorporated in thehot water circuit (Cruz de Campos, 1980). The water systems were chlorinated and

Epidemic Legionellosis in England and Wales 67

no further case occurred in association with this hotel until June 1982. It is notknown whether continuous chlorination had been maintained during the inter­vening period.

Outbreak 3

Between December 1979 and July 1980, 11 cases of nosocomial legionnaires'disease were recognised in a district general hospital (Fisher-Hoch et al., 1981).Eight cases had been in-patients shortly before onset, two had been members ofstaff and one had been a visitor. 1. pneumopbila serogroup 1 was isolated fromhot and cold water circuits and from the air conditioning cooling water system inthe hospital. A cluster of four cases occurred during a two week period after thecooling tower was brought into operation for the summer months but there hadbeen three cases before then. Despite cleansing and intermittent chlorination of thecooling water systems, cases continued to occur in the hospital and the outbreakceased only when the water supply to the building was chlorinated. Levels of1-2 ppm of free residual chlorine were maintained in the cold water system andthe hot water was circulated at a temperature of 55°C. A further nosocomial caseoccurred one year later in a patient who had been admitted to the hospital, andhad taken a bath and a shower, on the day that a reserve calorifier had been broughtinto use (Fisher-Hoch, Smith, Colborne, 1982). 1. pneumopbila may commonlybe recovered from calorifiers, particularly from the debris which collects in thebasin. It was found that the maintenance of a constant water temperature of 70°Cthroughout the calorifier successfully eradicated 1. pneumopbila from the cylinder.The hospital's hot water circuit, as well as the cold water system, is now chlorinatedto at least 1 ppm of free chlorine at outlets.

Outbreak 4

!lBetween late July and early September 1980 an outbreak of atypical pneumonia'o~curred among tourists from England and Wales who visited a hotel in Spain;23 met the criteria for diagnosis of legionnaires' disease. Eighteen were investigatedin a case-control study and although no association was demonstrated with bathingor showering it was found that cases, when compared with spouse controls, weremore likely to be the first to bathe in the morning (Bartlett et al., 1983). 1. pneu­mophila serogroup 1 was isolated from pooled hot and cold water samples collectedfrom shower units in the hotel; the building did not have an air conditioning system.The outbreak began within one week of the reconnection of a supplementary watersupply from a well. Continuous chlorination of both hot and cold water systemswas introduced with the maintenance of 2-3 ppm of free residual chlorine. Inaddition the hot water temperature was raised to 50-55°C at bedroom taps. Nonew cases of legionella pneumonia have been identified in association with thehotel since these control measures were introduced more than two years ago.

Outbreak 5

Between June and August 1981four cases of legionnaires' disease occurred amongtourists who stayed at a hotel in Italy. The hotel does not have an air conditioningsystem and there are no cooling towers in the vicinity so it was concluded that thehotel's water system was the probable source of infection. Single super-chlorinationrather than continuous chlorination was adopted as the control measure (Greco,

68 C.L.R.Bartiett and L.F.Bibby

1982). Two further cases occurred in association with the hotel in June 1982 andL. pneumopbila serogroup 1 was later isolated from the calorifier in the hotel.

Outbreak 6

In late September and early October 1981 six cases of pneumonia occurred amongmen working on a power station construction site. A detailed study of sicknessabsences failed to identify any other cases during the preceding three months orsucceeding six months. Three cases were shown to be legionnaires' disease, twohad antibody titres of 64 against the Colindale L. pneumophila serogroup 1 FYSantigen, and appropriate specimens were not available from one patient who died.There were numerous buildings on the site and there was no common hot or coldwater system to which the cases had been exposed. It was found that cases weremore likely than controls to have visited or worked in one part of the site in whichfour small-capacity cooling towers were located (Morton et al., 1983). The towersserved a compressor house and their water systems were interconnected. L. pneu­mophila serogroup 1 was isolated from cooling water samples. The entire systems,including towers, were thoroughly cleansed and a comprehensive treatment pro­gramme was introduced so as to inhibit corrosion, scale formation and organicgrowth. The principal biocide which was used was a biodegradable chlorinatedphenolic thio-ether. Since then no further cases of legionnaires' disease have beenrecognised in association with the site.

Discussion

Clustering of cases of legionella pneumonia in association with single sourceswas found to occur with remarkable frequency in England and Wales during theyears 1979-1982. Hotels were implicated most commonly but just over a quarterof the recognised clusters occurred in hospitals. Hotels often provided a continuingsource of infection over several years. In two-thirds of the 24 hotel-associatedclusters, cases had occurred in earlier years or have been recognised in subsequentyears. There is now considerable evidence to implicate water systems in hotels inEurope as important sources of legionella infections (Bartlett, 1983; Lalla, 1980;Tobin et al., 1981). It is tempting to speculate that the majority of the 265 travel­associated cases of legionnaires' disease reported in England and Wales from 1979­1982 were acquired from such sources.

Cooling water systems have been shown to be sources of L. pneumopbila infec­tions (Band et al., 1981; Dondero et al., 1980) but in England and Wales airbornespread from such a system has only been implicated clearly in one incident (Mortonet al., 1981) and possibly for four of the cases in Outbreak 3 (Fisher-Hoch et al.,1981). Furthermore the majority of clusters were associated with establishmentswhich do not have air conditioning systems so it seems reasonable to draw theconclusion that domestic water systems in hospitals or hotels are the most importantsources of epidemic legionellosis in England and Wales. There is no evidence toindicate that the situation is different in any other European countries.

L. pneumophila has been found more frequently in hot water than cold watercircuits (Dennis and colleagues, 1982), and the investigations of outbreaks suggestthat hot water systems are the principal sources of legionellosis. The hospital

Epidemic Legionellosis in England and Wales 69

calorifiers appeared to be the reservoir of infection for Outbreak 3 (Fisher-Hochet al., 1982) whereas in the hotel implicated in Outbreak 1 there was circumstantialevidence that the peripheral part of the hot water system was the source of at leastone infection. Outbreak 2 suggests that solar heating systems may provide anecological niche for 1. pneumophila but as yet there is no microbiological evidenceto support this hypothesis. The most effective control measure appears to be thatused to deal with Outbreak 4 in which both hot and cold water systems werechlorinated continuously and the hot water temperature raised to 60°C or morein the calorifier and at least 55°C to a maximum of 60°c at outlets. Raising thehot water temperature without additional chlorination did not prevent a furthercase occurring in association with the hotel implicated in Outbreak 1. It is possiblehowever that this case may have acquired the infection from another source becausehe had stayed at other hotels during the incubation period. A single super-chlorina­tion of hot and cold water systems to control Outbreak 5 did not prevent furthercases. The steps taken to control Outbreak 6 (Morton et al., 1983) do appear tohave been effective with continuous chemical treatment of the implicated coolingwater system. A biodegradable chlorinated phenolic thio-ether was selected becauseof its promising performance in controlling 1. pneumophila in cooling watersystems in a pilot field trial (Kurtz et al., 1982).

Although most of the seven nosocomial clusters were recognised at local levelall but one of the hotel-associated clusters were identified through the nationalsurveillance scheme. The central collation of data on cases that were widely scat­tered throughout the country, but had acquired the infection from single sites,provided the means for early recognition of common sources of infection and therapid introduction of preventive measures. This is a good illustration of the valueof national surveillance of communicable disease.

Acknowledgements. We wish to acknowledge the collaboration of all the microbiologistsand physicians who contributed to the national surveillance scheme. We are especiallyindebted to the microbiologists in the legionella reference laboratories, particularlyDr D. N. Hutchinson Dr J. B.Kurtz, Dr A. D. Macrae, Dr J. Nagington, Dr A. G. Taylor andDr J. O'H Tobin.

References

Band, J. D., M. LaVenture, J. P. Davis, G. F.Mallison, P. Skaliy, P. S. Hayes, W. L. Schell,H. Weiss, D.]. Greenberg, and D. W.Fraser: Epidemic Legionnaires' disease: airbornetransmission down a chimney. J. Amer. med. Ass. 245 (1981) 2404-2407

Bartlett, C.L.R., R.A.Swann, ]. Casal, L. Canada Royo, and A. G. Taylor: Recurrent Le­gionnaires' disease from a hotel water system: investigation and control of an outbreakin one building which had been associated with multiple cases over a period of 8 years.Abstract accepted for 2nd International Symposium on Legionella to be held in June1983 in Atlanta, Georgia/U.S.A. (1983)

Cruz de Campos, F.: Personal Communication (1980)Dennis, P.J., J. A. Taylor, R. B.Fitzgeorge, C. L. R. Bartlett, and G. I. Barrow: Legionella

pneumophila in water plumbing systems. Lancet i (1982) 949-951Dondero, T. J., R. C. Rendtorff, G.F.Mallison, R. M. Weeks, J. S.Levy, E. W. Wong, and

W.Schnaffner: An outbreak of Legionnaires' disease associated with a contaminated air­conditioning cooling tower. New Eng!. J. Med. 302 (1980) 365-370

70 C.L.R.Bartlett and L.F.Bibby

Fisher-Hoch, S.P., C.L. R. Bartlett.]. O'H. Tobin, M. B.Gillett, A. M. Nelson,]. E.Pritchard,M.C.Smith, R.A.Swann, ].M. Talbot, and ].A. Thomas: Investigation and control ofan outbreak of Legionnaires' disease in a district general hospital. Lancet i (1981) 932­936

Fisher-Hoch, S.P., C.L.R.Bartlett, C.].Harper, V.H.Lach, S.P.Clark, B.P.C.]ones, ].A.Thomas, and N. R. Crist: Legionnaires' disease at Kingston Hospital. Lancet i (1981)1154 (Corresp.)

Fisher-Hoch, S.P., M.C.Smith, and ].S.Colbourne: Legionella pneumophila in hospitalhot water cylinders. Lancet i (1982) 1073 (Corresp.)

Greco, D.: Personal communication (1982)Kurtz, ].B., C.L.R.Bartlett, V.A.Newton, R.A. White, and N.L.]ones: Legionella pneu­

mophila in cooling water systems: report of a survey of cooling towers in London anda pilot trial of selected biocides. J. Hyg. (Camb.) 88 (1982) 369-380

Lalla, F. de, C.Rossini, and C. Ciannattasion: Legionnaires' disease in an Italian Hotel.Lancet ii (1980) 1187

Morton, S., C.L. R. Bartlett, L. F.Bibby, D. N. Hutchinson, ]. V.Dyer, and P.]. Dennis: Anoutbreak of Legionnaires' disease from a cooling water system in a power station underconstruction. (1983) Manuscript in preparation

Taylor, A. C., T. C. Harrison, M. W.Dighero, and C.M. P.Bradstreet: False positive reac­tions in the indirect fluorescent antibody test for Legionnaires' disease eliminated by useof formolised yolk sac antigen. Ann. Intern. Med. 90 (1979) 686-689

Tobin,]. D'H, C.L.R.Bartlett, S.A. Waitkins, C.l.Barrow, A.D.Maerae, A. C. Taylor,R.].Fallon, and F.R.N.Lynch: Legionnaires' disease: further evidence to implicatewater storage and distribution systems as sources. Brit. med. J. 282 (1981) 573