Original Article
Potable Water as a Cause of Sporadic Cases of Community-Acquired Legionnaires' Disease
N Engl J Med 1992; 326:151-155January 16, 1992
- Abstract
- Abstract
Background.
The environmental sources of sporadic, community-acquired legionnaires' disease are largely unknown, and culturing of water sources after identification of a case is currently not recommended. We conducted a prospective study of sporadic cases of community-acquired legionnaires' disease to determine whether the environmental reservoirs could be identified.
Methods.
We cultured samples of potable water obtained from sources to which each of 20 patients with culture-confirmed, community-acquired legionnaires' disease had been exposed during the two weeks before the onset of symptoms. Monoclonal-antibody subtyping and restriction-endonuclease analysis were performed on the legionella isolates recovered from both the patients and the associated environmental cultures.
Results.
For 8 of the 20 patients, isolates of Legionella pneumophila with identical subtypes were identified in cultures from both the patient and the potable water to which the patient had been exposed. The environmental reservoirs linked to the infections were the water supplies of two private residences, two nursing homes, two hospital outpatient clinics, and an industrial plant.
Conclusions.
Potable-water supplies that harbor L pneumophila are an important source of community-acquired legionnaires' disease. Future studies should include attempts to identify the environmental sources of this infection. (N Engl J Med 1992;326:151–5.)
Article Activity
- Article
LEGIONELLA PNEUMOPHILA is a common cause of nosocomial and community-acquired pneumonia. In the majority of studies, nosocomial legionellosis has been linked with exposure to contaminated hospital water-distribution systems.1 2 3 In contrast, the reservoir for sporadic, community-acquired legionnaires' disease has not been well studied. Given the uncertainties about the epidemiology of community-acquired legionnaires' disease, the Centers for Disease Control have recommended that samples of water from sources associated with these cases should not be cultured.4 To clarify whether an environmental investigation after a diagnosis of community-acquired legionnaires' disease would yield useful information, we prospectively studied sporadic cases of community-acquired legionnaires' disease. We sought to determine whether the reservoir of L. pneumophila could be identified, and if so, for what proportion of the cases it could be identified.
Methods
Case Finding and Definition
All patients with culture-confirmed legionnaires' disease seen at the University of Pittsburgh Health Center, Mercy Hospital of Pittsburgh, or the Pittsburgh Veterans Affairs Medical Center or referred to the Special Pathogens Section of the medical center were included in this study.
Pneumonia was defined as the presence of acute respiratory symptoms and new pulmonary infiltrate. Respiratory symptoms had to include at least one of three major symptoms (cough, sputum production, and fever) or two of five minor symptoms and signs (pleuritic chest pain, dyspnea, altered mental status, pulmonary consolidation, or a white-cell count greater than 12.0×109 per liter). The pneumonia was considered to have been community-acquired if symptoms preceded hospital admission and were the basis for it and if the patient had not been admitted to the hospital within the previous two weeks.
Environmental Cultures
Each patient or a family member gave consent for samples to be obtained at the patient's residence to be cultured for L. pneumophila. Since the incubation period for legionnaires' disease ranges from 2 to 10 days, the patient or family member was questioned about the patient's activities during the 2 weeks before the onset of symptoms. We obtained water samples from each patient's home. If the patient lived anywhere other than at home during the two-week period (e.g., at a hotel), water samples from that site were also studied. If the patient had received any outpatient treatment at a local hospital (such as hemodialysis, radiation therapy, or a blood transfusion) within the period, samples of the water supply at that hospital were also cultured. Finally, in the case of all patients who were employed, samples from the workplace were also included in the culture protocol. Water samples were obtained from hot-water tanks, and swab samples were obtained from sinks, showers, and tub outlets.
Specimen Processing and Isolate Identification
The specimens were processed to isolate L. pneumophila, as described elsewhere.5 This included the use of selective differential mediums and pretreatment of specimens with acid. The cultures were incubated for five to seven days at 37°C in a humidified atmosphere, and colonies morphologically consistent with legionella were subcultured on blood-agar and buffered-charcoal yeast-extract agar plates. They were identified definitively by direct fluorescent-antibody testing.6 Isolates of L. pneumophila serogroup 12 were identified definitively by the Centers for Disease Control, Atlanta (courtesy of Donald Brenner and W. Lanier Thacker).
Subtyping
Isolates of L. pneumophila serogroup 1 from the patient and the patient's environment were typed by monoclonal-antibody techniques as described elsewhere.7 Each isolate was identifiable by code number only, and typing was performed without knowledge of its source.
Restriction-Endonuclease Analysis
The pairs of legionella isolates from each patient and that patient's environment that were found to be identical on the basis of serogrouping or monoclonal-antibody subtyping were studied further by restriction-endonuclease analysis. For comparison, this analysis was also performed on one environmental isolate associated with each of the four cases in which there was no such match and on one isolate of L. pneumophila serogroup 1 from each of 12 households in Pittsburgh where no disease had occurred. The latter isolates had the following subtypes: Bellingham (two), Oxford (two), and OLDA (eight). Specimens from these 12 households were cultured in a separate survey of Pittsburgh homes (unpublished data). Whole-cell DNA restriction-endonuclease analysis was performed on coded isolates as previously described.8 Purified whole-cell DNA was analyzed individually with three restriction endonucleases: EcoRI, EcoRV, and HindIII.
Results
Twenty consecutive culture-confirmed cases of community-acquired legionnaires' disease were identified, involving 5 women and 15 men. The mean age of the patients was 60 years (range, 28 to 75). The underlying diseases included myelodysplastic disease (in 3 patients), chronic obstructive pulmonary disease, diabetes mellitus, renal disease, arteriosclerotic heart disease (2 patients each), cancer, vasculitis requiring steroid use, heart disease requiring transplantation, and seizure disorder (1 patient each). Seven patients had no underlying disease. These numbers exceed 20 because 2 patients had two underlying diseases. Twelve patients were cigarette smokers, of whom 5 had no underlying disease. Six of the 20 patients died of legionnaires' disease.
L. pneumophila serogroup 1 was isolated from 18 of the 20 patients, L. pneumophila serogroup 3 from 1, and L. pneumophila serogroup 12 from 1. Several monoclonal subtypes of L. pneumophila serogroup 1 were identified among the patients (Table 1Table 1
Results of an Investigation of Possible Linkage between Environmental Reservoirs of L. pneumophila and 20 Cases of Sporadic Community-Acquired Legionnaires' Disease.*).A source of exposure was identified for 40 percent of the community-acquired cases of legionnaires' disease (8 of 20) (Table 1). For three patients, the subtypes of the infecting strains (OLDA, Bellingham, and Benidorm) matched those of the isolates recovered from the residential water systems; two of these three residences had large-volume water systems serving a number of housing units. In two cases, the subtype of the isolates from the patients (OLDA) matched that of the strain recovered from the water supplies in the nursing homes where the two patients had lived. Two other patients had received outpatient radiation therapy at community hospitals where L. pneumophila was recovered from the water supply. One of these patients was infected with L. pneumophila serogroup 12, and this serogroup was found in the hospital water supply. The second patient (Patient 7, Table 1) was infected with L. pneumophila serogroup 1, subtype Bellingham, which was also found in the water supply of the community hospital where he had received radiation therapy. Both the isolate from the patient and that from the water supply had a unique pattern of sensitivity to antibiotics that further confirmed their association.9 Finally, the subtype of the isolate of L. pneumophila serogroup 1 (Philadelphia) that was recovered from one patient matched the subtype of the isolates found in the patient's workplace, an industrial plant. The details of this case have been presented elsewhere.10
Restriction-endonuclease analysis was used to study the matching pairs of isolates from the patient and the environment — seven pairs involving L. pneumophila serogroup 1 and an identifiable monoclonal-antibody subtype and one pair involving L. pneumophila serogroup 12. The isolates identified in each pair had identical patterns on restriction-endonuclease analysis (Fig. 1Figure 1
Restriction-Endonuclease Analysis of Pairs of Matching Isolates from the Patient and the Environment.). Two pairs with the subtype OLDA (Patients 4 and 6, Table 1) shared the same pattern, whereas all the other pairs had unique patterns.The unique typing characteristics of these matched pairs of isolates were investigated by monoclonal-antibody subtyping and restriction-endonuclease analysis of 16 isolates of L. pneumophila serogroup 1 that were obtained from 16 different and unrelated sources of water in the Pittsburgh area. When used in combination, the two methods were capable of dividing the 16 isolates into eight different subgroups (data not shown). Five of these isolates had typing characteristics identical to those of the pair of isolates corresponding to Patient 1 (Table 1). Three isolates had the same typing pattern as that of the pairs of isolates corresponding to Patients 4 and 6. All the other environmental isolates were distinct from the isolates obtained from our patients.
Discussion
Unlike other common microbial agents that cause pneumonia, legionella has a source that is extrinsic to the host, and thus theoretically the disease can be prevented by control measures directed at the environmental source. The policy of performing environmental cultures of hospital water-distribution systems after the discovery of nosocomial cases of legionnaires' disease has led to strategies successful in preventing nosocomial legionnaires' disease.1 2 3 , 11 Because nine large studies of community-acquired pneumonia have ranked L. pneumophila among the three most common microbial causes,12 13 14 15 16 17 18 19 20 it appears logical that further investigative efforts should be directed at identifying the environmental reservoir for cases of community-acquired legionnaires' disease.
In our study, we cultured samples of potable water from the sources to which the patients had the greatest recent exposure. Cooling towers were not included in our surveillance protocol, although in most cases they were not located near the patient's workplace or home. This was a potential weakness of our study, since we might have been able to link additional reservoirs of legionella to infection had we sampled the cooling towers. On the other hand, reports associating legionnaires' disease with cooling towers have decreased dramatically in number in the past few years, and it has been suggested that such towers are not a major disseminator of legionnaires' disease.21
Isolates of L. pneumophila serogroup 1 from the patient and the environment were typed with standardized monoclonal-antibody methods.7 In addition, when isolates from a patient and that patient's environment were found to be identical by either monoclonal subtyping or serogrouping, the epidemiologic association was strengthened by restriction-endonuclease analysis of DNA. Correlation of both the monoclonal-antibody subtype and the restriction-endonuclease pattern for the isolates from the patient's respiratory tract and from the environment was considered to indicate an association.
For 40 percent of the patients with community-acquired legionnaires' disease (8 of 20), an epidemiologic association was demonstrated that implicated potable water as the source of infection. The implicated sources of infection included private residences (in three cases), nursing homes (in two), an industrial plant (in one), and community hospitals (in two) (Table 1). The fact that there was heterogeneity of L. pneumophila types (i.e., two serogroups, four monoclonal-antibody types, and seven restriction-endonuclease patterns) and that the type found in the patient precisely matched the type found in the environmental sample make it extremely unlikely that these results could have occurred by chance. In addition, the diversity of restriction-endonuclease patterns among environmental isolates with the same monoclonal-antibody subtype further supports the epidemiologic association between exposure to specific sources of potable water and subsequent disease.
Contrary to the report by Redd and Cohen,4 L. pneumophila is not found frequently in single-dwelling homes. A separate study conducted while this one was in progress found that the frequency of recovery of L. pneumophila from Pittsburgh homes in which legionella pneumonia had not occurred was only 6 percent ( 14 of 218) (unpublished data). In that study, L. pneumophila serogroup 1 was isolated from 12 homes, serogroup 3 from 1 home, and serogroup 4 from 1 home. The pattern of recovery of monoclonal-antibody subtypes from homes not associated with a case of legionnaires' disease was similar to that found in homes associated with such a case.
In three of our patients, infection was linked to exposure to L. pneumophila in the home. Two of these patients lived in large, multidwelling residences with public water supplies. We have previously reported two cases of community-acquired legionnaires' disease that were linked to contaminated household water systems22; however, the homes of these two patients were supplied with water from private wells.
In other, smaller surveys conducted in Quebec and Pittsburgh, only 20 to 30 percent of the homes sampled yielded L. pneumophila. 23 , 24 One study conducted in the Chicago area reported that 32 percent of the household water supplies were positive for legionella; however, the majority of the residences sampled were large apartment buildings.25 Large water-distribution systems are more susceptible to colonization with legionella, since large storage tanks provide an environment optimal for the growth of L. pneumophila (with lower temperatures, commensal microbial flora, scale, and sediment).26 27 28 29 30 31 In our study, the water supplies of two multidwelling residences, two nursing homes, two community hospitals, and an industrial plant were linked to the seven cases of community-acquired legionnaires' disease. Each of these water-distribution systems was supplied by large water-storage tanks.
Determining the source of exposure for community-acquired legionnaires' disease caused by L. pneumophila serogroup 1 can be complicated if this organism is recovered from more than one of the implicated sources. But the epidemiologic reservoir can be identified by molecular typing methods, as was done in the case of Patient 6 (Table 1), who was infected with L. pneumophila serogroup 1, subtype OLDA. During the incubation period, this patient was exposed to the water systems of three different institutions in the course of residence in a nursing home, a one-week stay as an inpatient in a local hospital, and daily hemodialysis treatments at a second hospital. The water systems of all three institutions were found to harbor L. pneumophila serogroup 1. Only the isolates recovered from the nursing home matched the OLDA subtype (Table 1); restriction-endonuclease analysis of the isolates confirmed the association.
We have shown that if environmental sources are studied to investigate cases of sporadic community-acquired legionnaires' disease, the source of exposure can often be identified. Future investigators, as well as the authors of previous reports of community-acquired legionnaires' disease putatively attributed to cooling towers, should now consider the possibility that alternative reservoirs, including the water supply in a patient's residence, may be the source of the exposure. This is an opportune time to reassess the recommendation that environmental cultures for legionella should not be pursued after the diagnosis of community-acquired legionnaires' disease.4
Source Information
From the University of Pittsburgh and the Veterans Affairs Medical Center, Pittsburgh (J.E.S., V.L.Y., P.M.); the Université Laval, Quebec, Canada (J.J.); and Stanford University Medical Center, Stanford, Calif. (N.T., L.S.T.). Address reprint requests to Dr. Yu at the Infectious Disease Section, Veterans Affairs Medical Center, University Dr. C, Pittsburgh, PA 15240.
- References
References
1
Best
CrossRef | Web of Science | Medline2
Ezzeddine
CrossRef | Web of Science | Medline3
Snyder
CrossRef | Web of Science | Medline4
Redd
CrossRef | Web of Science | Medline5
Vickers
Medline6
Cherry
Web of Science | Medline7
Joly
Web of Science | Medline8
Tompkins
Web of Science | Medline9
Vickers
Web of Science10
Muraca
CrossRef | Medline11
Muraca
CrossRef | Web of Science | Medline12
Macfarlane
CrossRef | Web of Science | Medline13
Klimek
CrossRef | Web of Science | Medline14
Muder
CrossRef | Web of Science | Medline15
Aubertin
CrossRef | Web of Science | Medline16
Fang
Web of Science | Medline17
Ruf
CrossRef | Web of Science | Medline18
Falcó
CrossRef | Web of Science | Medline19
Friis-Moller
CrossRef | Web of Science | Medline20
Bates
CrossRef | Web of Science21
Muder
CrossRef | Web of Science | Medline22
Stout
CrossRef | Web of Science | MedlineErratum, JAMA 1987;257:2595.
23
Lee
CrossRef | Medline24
Joly
Web of Science | Medline25
Arnow
CrossRef | Web of Science | Medline26
Vickers
Medline27
Fisher-Hoch
CrossRef | Web of Science | Medline28
Fields
Web of Science | Medline29
Stout
Web of Science | Medline30
Witherall
Web of Science31
Wadowsky
Web of Science | Medline
- Citing Articles (29)
Citing Articles
1
S. F. DUFRESNE, M. C. LOCAS, A. DUCHESNE, C. RESTIERI, J. ISMAÏL, B. LEFEBVRE, A. C. LABBÉ, R. DION, M. PLANTE, M. LAVERDIÈRE. (2011) Sporadic Legionnaires' disease: the role of domestic electric hot-water tanks. Epidemiology and Infection1-10
CrossRef2
Christopher Nalepa, Terry Williams. 2010. Biocides. , 318-410.
CrossRef3
Victoria Ng, Patrick Tang, Frances Jamieson, Steven J. Drews, Shirley Brown, Donald E. Low, Caroline C. Johnson, David N. Fisman. (2008) Going with the Flow: Legionellosis Risk in Toronto, Canada Is Strongly Associated with Local Watershed Hydrology. EcoHealth 5:4, 482-490
CrossRef4
R.H. Doust, M.A. Mobarez, H. Bagheri, N. Khoramabadi. (2008) Interaction of Legionellae and Free-Living Amoebae Within Hospital Water Supplies. Research Journal of Parasitology 3:3, 104-113
CrossRef5
C-W. Chang, Y-H. Hwang, W-Y. Cheng, C-P. Chang. (2007) Effects of chlorination and heat disinfection on long-term starved Legionella pneumophila in warm water. Journal of Applied Microbiology 102:6, 1636-1644
CrossRef6
Brendan Flannery, Lisa B. Gelling, Duc J. Vugia, June M. Weintraub, James J. Salerno, Michael J. Conroy, Valerie A. Stevens, Charles E. Rose, Matthew R. Moore, Barry S. Fields, Richard E. Besser. (2006) Reducing Legionella Colonization of Water Systems with Monochloramine. Emerging Infectious Diseases 12:4, 588-596
CrossRef7
Megan Young, Helen Smith, Bruce Gray, Ben Huang, Jackie Barten, Chris Towner, Sandy Plowman, Bahavak Afshar, Norman Fry, Barry Blair, John Savill, Bradley McCall. (2005) The public health implications of a sporadic case of culture-proven Legionnaires' disease. Australian and New Zealand Journal of Public Health 29:6, 513-517
CrossRef8
Linda P. B. Verhoef, Ed P. F. Yzerman, Jacob P. Bruin, Jeroen W. Den Boer. (2004) Domestic Exposure to Legionellae for Dutch Legionnaires' Disease Patients. Archives of Environmental Health: An International Journal 59:11, 597-603
CrossRef9
Teresa Brooks, Raegan A. Osicki, V. Susan Springthorpe, Syed Abdus Sattar, Lionel Filion, David Abrial, Serge Riffard. (2004) DETECTION AND IDENTIFICATION OF Legionella SPECIES FROM GROUNDWATERS. Journal of Toxicology and Environmental Health, Part A 67:20-22, 1845-1859
CrossRef10
James D. Heffelfinger, Jacob L. Kool, Scott Fridkin, Victoria J. Fraser, Jeffrey Hageman, Joseph Carpenter, Cynthia G. Whitney, . (2003) Risk of Hospital‐Acquired Legionnaires’ Disease in Cities Using Monochloramine Versus Other Water Disinfectants • . Infection Control and Hospital Epidemiology 24:8, 569-574
CrossRef11
Ronald M. Atlas. (1999) Legionella: from environmental habitats to disease pathology, detection and control. Environmental Microbiology 1:4, 283-293
CrossRef12
J.R. Köhler, M. Maiwald, P.C. Lück, J.H. Helbig, V. Hingst, H.-G. Sonntag. (1999) Detecting legionellosis by unselected culture of respiratory tract secretions and developing links to hospital water strains. Journal of Hospital Infection 41:4, 301-311
CrossRef13
A J Smith, J Hood, J Bagg, F T Burke. (1999) cross-infection control: Water, water everywhere but not a drop to drink?. British Dental Journal 186:1, 12-14
CrossRef14
Jacob L Kool, Joseph C Carpenter, Barry S Fields. (1999) Effect of monochloramine disinfection of municipal drinking water on risk of nosocomial Legionnaires'disease. The Lancet 353:9149, 272-277
CrossRef15
A. Rampling, C.J. Butt, A.A. West, M. Tully, K.T. Palmer. (1997) Community-acquired Legionnaires' disease following minimal exposure to a contaminated source. Journal of Infection 35:3, 300-302
CrossRef16
Stout, Janet E., Yu, Victor L., . (1997) Legionellosis. New England Journal of Medicine 337:10, 682-687
Full Text17
JETTE MARIE BANGSBORG. (1997) Antigenic and genetic characterization of Leaionella Proteins: Contribution to taxonomy, diagnosis and pathogenesis. APMIS 105:S70, 5-53
CrossRef18
T. J. Marrie, W. Johnson, S. Tyler, G. Bezanson, D. Haldane, S. Burbridge, J. Joly. (1995) Potable water and nosocomial Legionnaires' disease – check water from all rooms in which patient has stayed. Epidemiology and Infection 114:02, 267
CrossRef19
M. Maiwald, M. Schill, C. Stockinger, J. H. Helbig, P. C. Lück, W. Witzleb, H.-G. Sonntag. (1995) Detection ofLegionella DNA in human and guinea pig urine samples by the polymerase chain reaction. European Journal of Clinical Microbiology & Infectious Diseases 14:1, 25-33
CrossRef20
J. Hood, J. Hay, T. Gillespie, D.V. Seal. (1994) A fatal case of community-acquired Legionnaires' disease acquired from potable water. Journal of Infection 29:2, 231-233
CrossRef21
C.F von Reyn, J.N Marlow, R.D Arbeit, T.W Barber, J.O Falkinham. (1994) Persistent colonisation of potable water as a source of Mycobacterium avium infection in AIDS. The Lancet 343:8906, 1137-1141
CrossRef22
M. Maiwald, K. Kissel, S. Srimuang, M. Doeberitz, H.-G. Sonntag. (1994) Comparison of polymerase chain reaction and conventional culture for the detection of legionellas in hospital water samples. Journal of Applied Microbiology 76:3, 216-225
CrossRef23
S.B. Surman, L.H.G. Morton, C.W. Keevil. (1994) The dependence of Legionella pneumophila on other aquatic bacteria for survival on R2A medium. International Biodeterioration & Biodegradation 33:3, 223-236
CrossRef24
L.H.G. Morton, S.B. Surman. (1994) Biofilms in biodeterioration — a review. International Biodeterioration & Biodegradation 34:3-4, 203-221
CrossRef25
Ling-Ling Hung, Denise C. Copperthite, Chin S. Yang, Frank A. Lewis, Francis A. Zampiello. (1993) Environmental Legionella Assessment in Office Buildings Of Continental United States. Indoor Air 3:4, 349-353
CrossRef26
R. Nechwatal, A. Prull, H. Lutz, W. Ehret, O. J. Klatte, H-J Zeissler. (1993) Nosocomial outbreak of legionellosis in a rehabilitation center. Demonstration of potable water as a source. Infection 21:4, 235-240
CrossRef27
R.S. Bhopal, R. Wagstaff. (1993) Prospects for the elimination of Legionnaires' disease. Journal of Infection 26:3, 239-243
CrossRef28
Desforges, Jane F., , Tompkins, Lucy S., . (1992) The Use of Molecular Methods in Infectious Diseases. New England Journal of Medicine 327:18, 1290-1297
Full Text29
(1992) Sporadic Cases of Legionnaires' Disease. New England Journal of Medicine 326:25, 1699-1701
Full Text
- Letters
