DISEASE CAUSED BY LEGIONELLA PNEUMOPHILA AND CHLAMYDIA PNEUMONIAE
An unrecognized bacterial genus until 1977, Legionella was isolated by the Centers for Disease Control during the investigation of an explosive outbreak of pneumonia among American Legion conventioneers in Philadelphia. More than 25 species of Legionella have subsequently been isolated, and 14 serotypes of Legionella pneumophila have been identified. About 90% of human infections are caused by L. pneumophila; other species recovered from patients include L. micdadei, L. bozemanii, L. dumoffii, L. gormanii, and L. feeleii. Legionella is usually associated with community-acquired and nosocomial pneumonia; however, members of this genus have been recovered from a variety of extrapulmonary infections.
L. pneumophila, the etiologic agent of Legionnaires’ disease, can be isolated from water found in nature (e.g., lakes, creeks), heat-exchange units (e.g., cooling towers), and domestic supplies (e.g., taps, showers), and most epidemics of Legionnaires’ disease have been traced to contaminated water from one of these sources. On the basis of the epidemiologic observations, investigators have assumed that Legionnaires’ disease develops after a person inhales aerosols containing the organism; however, contemporary studies have suggested that aspiration may play a prominent role in the pathogenesis of the infection. Person-to-person transmission has not been documented.
A facultative, intracellular pathogen, L. pneumophila is capable of persisting and multiplying within pulmonary alveolar macrophages. Thus, an intact cellular immune system is essential for preventing and controlling infection by the organism. Not surprisingly, many patients who experience Legionnaires’ disease are immunosuppressed. Optimal antimicrobial treatment includes drugs that penetrate macrophages (macrolides, rifampin, fluoroquinolones), and relapses have occurred following short courses of therapy.
The risk for Legionnaires’ disease is increased in men, persons over the age of 50 years (mean age, 56 years), and smokers. Risk is also increased among patients who receive corticosteroids or cytotoxic agents, who are alcoholic, or who have diabetes mellitus, cancer, chronic obstructive pulmonary disease, or renal failure requiring dialysis or transplantation. Case-control studies have also found a significant association between Legionnaires’ disease and the occupation of construction worker, the presence of excavation sites near the home, and a history of travel within the preceding 2 weeks.
The onset of Legionnaires’ disease is typically abrupt. The initial symptoms are usually nonspecific and include fever, anorexia, malaise, myalgias, and headache. Recurrent chills or rigors are experienced by most patients. The fever may increase in a stepwise fashion, and temperatures above 40°C are seen in more than half of cases. Pulmonary symptoms appear 2 to 3 days after the onset of illness, and they may progress within a few hours to include dyspnea and cyanosis. The cough is initially dry but becomes productive of nonpurulent secretions and, subsequently, frankly purulent sputum; hemoptysis occurs in up to one third of cases. Chest pain, which is usually pleuritic, is common. Diarrhea, which consists of the passage of three to four loose or watery stools daily, occurs in about one half of all cases, and this symptom may precede or follow the respiratory complaints.
The physical examination typically reveals an acutely ill patient with tachypnea and an unremitting fever above 38.4°C; a relative bradycardia is found in two thirds of patients. Auscultation of the chest reveals moist crackles and rhonchi; evidence of consolidation or a pleural effusion may appear later in the course. An abnormal neurologic evaluation is recorded in up to one third of patients; changes in mental status include emotional lability, confusion, delirium, hallucinations, lethargy, and stupor. The neurologic examination may also reveal fine or coarse tremors, hyperactive reflexes, and signs of cerebellar dysfunction, such as dysarthria or ataxia.
A total WBC count of more than 10,000/mm3 is found in three fourths of patients, and immature granulocytes are common; however, the leukocyte count rarely exceeds 20,000/mm3. Thrombocytopenia and disseminated intravascular coagulation occur infrequently. The urinalysis may demonstrate proteinuria or hematuria. Evidence of acute renal failure may be present; the kidney dysfunction is usually a consequence of shock, disseminated intravascular coagulation, immune complex glomerulonephritis, interstitial nephritis, hemoglobinuria, or myoglobinuria. A number of nonspecific biochemical abnormalities have been noted in patients with Legionnaires’ disease, including modest elevations in bilirubin and hepatic enzymes, hypoalbuminemia, hypophosphatemia, and hyponatremia. Cerebrospinal fluid is almost always normal. The initial chest radiograph characteristically reveals a patchy, unilobar infiltrate; any area of the lungs may be involved, although lower lobe disease is most common. The pulmonary infiltrate usually progresses to complete consolidation, and occasionally cavitation may occur; patients who are receiving corticosteroids are at greater risk for development of a lung abscess. The pulmonary process can spread to involve adjacent lobes or the opposite lung. Pleural effusions are typically small.
A number of laboratory methods are available for diagnosing Legionnaires’ disease. A serologic assay, the indirect fluorescent antibody (IFA) test, represents the most commonly employed diagnostic procedure. A fourfold increase in IFA titer to at least 1:128 between acute and convalescent phases of the disease is considered confirmatory. Unfortunately, an interval of 6 to 8 weeks is often required to detect a diagnostic rise in IFA titers. A single acute-phase IFA titer of 1:256 or above in a patient with an illness compatible with Legionnaires’ disease has been believed to represent strong presumptive evidence of the infection; however, recent studies have cast doubt on the predictive value of a single antibody titer in the patient with sporadic legionellosis. The urinary antigen assay has become commercially available; this technique is 80% to 90% sensitive in patients infected with L. pneumophila serogroup 1, which is responsible for about 80% of the sporadic cases of Legionnaires’ disease. Direct fluorescent antibody (DFA) staining is a rapid means of diagnosing the disease. This technique can be used to examine expectorated sputum, transtracheal aspirates, pleural fluid, and lung or extrapulmonary tissues for the presence of L. pneumophila. In laboratories with extensive experience with the procedure, DFA staining of sputum is about 70% sensitive and 95% specific. The existence of multiple serogroups of L. pneumophila contributes to the difficulty of DFA staining. L. pneumophila can be recovered from primary cultures of sputum, pleural fluid, blood, lung biopsy specimens, and other clinical material. Special media, such as buffered charcoal yeast extract agar and biphasic blood culture broth, are required for isolation. It usually requires 3 to 7 days to detect the growth of L. pneumophila. Occasionally, Legionella organisms can be seen on the Gram’s stain of clinical material; they appear as faintly staining, small, gram-negative bacilli. Special stains (e.g., DFA, Dieterle silver impregnation) are usually required to detect the organism in pathologic specimens. Finally, DNA amplification through the use of the polymerase chain reaction (PCR) has been shown to be very sensitive and highly specific, and the technique has been used to detect Legionella in bronchoalveolar lavage specimens and throat swabs; PCR for the diagnosis of Legionnaires’ disease will likely become widely available.
The overlap in the clinical manifestations of sporadic legionellosis and other pneumonias is usually sufficient to obscure the correct diagnosis early in the patient’s course. The presence of a progressive pneumonic disease associated with multisystem abnormalities; a paucity of sputum; the absence of a predominant organism on Gram’s stain or culture; and the persistence of prostration, unremitting fever, and recurrent rigors despite therapy with penicillins, cephalosporins, or aminoglycosides should raise the possibility of Legionnaires’ disease. Infection with Mycoplasma pneumoniae and Chlamydia pneumoniae can resemble disease caused by L. pneumophila; however, patients with M. pneumoniae and C. pneumoniae infection tend to be healthy younger adults, and they usually experience an illness that is more insidious in onset and less intense in severity. Psittacosis, tularemia, and Q fever can produce clinical syndromes that parallel those associated with Legionnaires’ disease; these alternative diagnoses can usually be excluded by a detailed epidemiologic history. Because patients with Legionnaires’ disease are often immunocompromised hosts, they are at risk to experience either concurrent or sequential infection with other pathogens, including common pyogenic bacteria, such as Streptococcus pneumoniae and Mycobacterium tuberculosis.
Under laboratory conditions, L. pneumophila demonstrates susceptibility to a large number of antimicrobial agents, including macrolides/azalides, tetracyclines, and fluoroquinolones. Clinical experience during the past two decades has supported the use of erythromycin as the drug of choice for the therapy of Legionnaires’ disease; similarly, observations made during the initial outbreak of Legionnaires’ disease in Philadelphia 1976 and subsequent reports indicate that tetracycline or doxycycline is also effective. In generally small, noncomparative studies, azithromycin, clarithromycin, ciprofloxacin, oflaxacin, and perfloxacin have all been shown to be curative. At present, some experts recommend that a fluoroquinolone be utilized as initial therapy. The use of ciprofloxacin or ofloxacin eliminates some of the nettlesome problems associated with high-dose erythromycin therapy, including phlebitis at the infusion site, gastrointestinal intolerance, a number of potential drug-drug interactions, and the need to infuse the macrolide with large volumes of fluid.
In general, high doses of antimicrobials should be administered to patients with suspected or confirmed Legionnaires’ disease. For patients with a milder clinical illness and without significant underlying diseases, some experts recommend clarithromycin or azithromycin as initial therapy; for immunosuppressed or critically ill patients, a fluoroquinolone might be selected. If erythromycin or doxycycline is employed and if the patient appears seriously ill, rifampin should also be administered. Of note, all antimicrobial regimens have been associated with treatment failures, and if the patient fails to demonstrate some improvement within 48 to 72 hours, an alternate therapy should be considered.
The response to appropriate antimicrobial therapy is characteristically prompt, and most patients show substantial improvement within 24 to 48 hours. Occasionally, the chest radiograph will demonstrate a progression of the infiltrate during the first few days of treatment; however, some clearing of the pneumonia usually occurs within the first 2 weeks of therapy. Complete resolution of pulmonary infiltrates may take 4 to 8 weeks. With clinical improvement, patients may be given oral antimicrobials. A 3-week course of antimicrobial therapy should be completed to reduce the likelihood of relapse; patients with lung abscesses require more prolonged therapy.
The prognosis among patients with Legionnaires’ disease is correlated with the severity of underlying disease and the use of appropriate antimicrobials. The overall case fatality rate for sporadic legionellosis ranges from 6% to 25%; however, among patients who are immunosuppressed and who do not receive appropriate antimicrobial therapy, mortality rates of up to 80% have been noted. Persistent malaise, fatigue, and problems with memory are common complaints among patients who experience Legionnaires’ disease.
Pontiac fever represents the second most common syndrome associated with L. pneumophila. Although the etiology remained obscure until Legionella was isolated, the disease was first observed in 1968 among employees in a county health department building in Pontiac, Michigan. Since that time, a limited number of epidemics have been reported. Aerosolized contaminated water has been implicated as the source of these outbreaks. Pontiac fever is a nonpneumonic form of legionellosis that affects previously healthy adults. The illness is remarkable for a short incubation period (24 to 48 hours) and a high attack rate (up to 95% of exposed persons). Clinically, Pontiac fever is characterized by fever of up to 40°C, chills, myalgias, headache, fatigue, a dry cough, and vague neurologic symptoms, such as dizziness. A leukocytosis is frequently noted among hospitalized patients. The acute illness usually resolves within 48 to 96 hours without specific therapy. Residual complaints are common, and these symptoms include lassitude, forgetfulness, and an inability to concentrate. Neither secondary cases nor deaths have been reported. Other Legionella species, including L. feeleii, have been identified as causes of outbreaks of Pontiac fever.
Resulting from hematogenous dissemination, extrapulmonary infections with L. pneumophila usually occur as a complication of Legionnaires’ disease. Pericarditis and hemodialysis fistula infections have been described in association with Legionnaires’ disease. A few other localized infections, such as pyelonephritis and myocarditis, have been identified at postmortem examination. On occasion, infections with Legionella species can occur in the absence of obvious lung involvement; L. pneumophila has been identified as the cause of bacterial endocarditis in a patient with prosthetic valves, and Legionella has been implicated as a cause of granulomatous hepatitis, fever of unknown origin, and postoperative wound infections.
C. pneumoniae (TWAR agent) was first isolated from the conjunctivae of a Taiwanese child in 1965 and subsequently associated with acute respiratory tract infections in humans in 1983. Following the recognition that the microbe can cause sinusitis, pharyngitis, bronchitis, and pneumonia, epidemiologic studies demonstrated that infections with C. pneumoniae are prevalent; for example, approximately 50% of all adults and 75% of aged individuals have serologic evidence of prior infection, and the microbe has been implicated in about 5% of episodes of sinusitis and 10% of cases of community-acquired pneumonia.
Like other chlamydial species, C. pneumoniae is an obligate intracellular pathogen. In contrast to C. psittaci, which can also cause pneumonia in adults, C. pneumoniae does not have a zoonotic reservoir, and all disease is believed to result from human-to-human transmission via respiratory tract secretions. The incubation period seems to last several weeks. Although outbreaks of disease among close contacts, such as family members, have been reported, transmission appears inefficient, and most cases are likely the consequence of acquisition from asymptomatic carriers. Finally, C. pneumoniae can remain viable on environmental surfaces for hours, raising the possibility that fomites play a role in transmission.
Infection with C. pneumoniae leads to both cellular and serologic immune responses; the latter includes IgM and IgG antibodies, and those immunoglobulins serve as important markers for the diagnosis of acute infection and for epidemiologic studies. Of note, the seroprevalence studies have demonstrated that epidemics of infection occur in 4- to 6-year cycles and that attack rates are highest in children 5 to 14 years of age. The seroepidemiologic investigations have also suggested that immunity is not long-lived and that many people are infected repeatedly throughout life.
The symptoms and signs of bronchitis and pneumonia caused by C. pneumoniae do not appear to be unique. As in other nonpyogenic respiratory tract infections, the onset of disease is usually insidious. Often, the illness is biphasic; initially, the patient experiences pharyngitis that may be associated with hoarseness, and after the upper respiratory tract symptoms abate, the patient notes fever and a nonproductive cough. The most notable physical examination finding may be a relative bradycardia; focal rales on chest auscultation are usually present in patients with pneumonia. Relevant laboratory test results include a normal WBC count and a focal, usually subsegmental lower lobe infiltrate on chest radiograph; among older patients and persons with underlying diseases, the infiltrates can be extensive. Occasionally, patients may have coinfection with S. pneumoniae and, of course, these patients can be very ill. Case fatality rates for relatively fit persons with pneumonia are extremely low; however, cough and malaise can persist for weeks following resolution of the infection.
The isolation of C. pneumoniae requires tissue culture systems. Accordingly, serologic testing represents the usual method of diagnosis. The C. pneumoniae-specific microimmunofluorescence test appears to be the most reliably sensitive and specific assay. The serologic response to infection is variable and occurs slowly; accordingly, acute and convalescent phase sera should be secured 3 to 4 weeks apart, and the paired specimens should be submitted for testing. Because patients can experience either an acute primary infection or an acute reinfection, the patterns of serologic response will vary. In general, in acute primary infection, IgM antibodies appear about 2 to 3 weeks following the onset of illness, and IgG antibodies become detectable 6 to 8 weeks after the onset of disease; a fourfold antibody rise, an IgM titer at or above 1:16, or an IgG titer at or above 1:512 is considered diagnostic. In patients with reinfection, IgM antibodies may not appear, and a fourfold titer rise, which can occur in 1 to 2 weeks, is considered diagnostic. Finally, although employed primarily as a research tool, PCR has been utilized successfully to detect C. pneumoniae in clinical specimens, such as pharyngeal swabs and sputum.
Controlled trials concerning the therapy of lower respiratory tract infections caused by C. pneumoniae remain limited. In vitro, the microbe is susceptible to erythromycin, azithromycin, clarithromycin, doxycycline, and some of the fluoroquinolones; limited clinical data suggest that the macrolides and the azalide (azithromycin) possess comparable efficacy. The course of therapy should be 14 to 21 days; some authorities suggest that if cough or malaise persists, a second course of antimicrobials be administered, and in the absence of contraindications, doxycycline or tetracycline is recommended. Like other chlamydial organisms, the agent is not susceptible to beta-lactam antibiotics, such as amoxicillin.
C. pneumoniae has also been identified as a cause of other acute illnesses; these include pharyngitis, sinusitis, otitis media, and rarely, a systemic sepsis-like syndrome. Of special note is the fact that the microbe has been associated with several chronic conditions, most notably asthmatic bronchitis and coronary artery disease. The relationship between C. pneumoniae and coronary artery disease is based on seroprevalence studies, which have demonstrated that patients with coronary artery disease are more likely to have serologic evidence of prior infection, and on morphologic and immunologic investigations, in which the microbe has been visualized by electron microscopy or detected by PCR or immunocytochemical staining in atheromatous plaques from coronary arteries and other vessels. Although C. pneumoniae has been definitively associated with atheromatous lesions, the role of the microbe in the pathogenesis of vascular disease remains to be defined. Finally, a number of reports have indicated that reinfection may precipitate the onset of acute myocardial infarction. (A.L.E.)
Alexiou SD, et al. Isolation of Legionella pneumophila from hotels in Greece. Eur J Epidemiol 1989;5:47.
Legionella was recovered from water samples taken from hotels located in different areas of Greece that were associated with cases of Legionnaires’ disease.
Ampel NM, Rubin FL, Norden CW. Cutaneous abscess caused by Legionella micdadei in an immunosuppressed patient. Ann Intern Med 1985;102:630.
A patient receiving immunosuppressive drugs had a leg abscess complicate Legionella pneumonia.
Arnow PM, Boyko EJ. Perirectal abscess caused by Legionella pneumophila and mixed anaerobic bacteria. Ann Intern Med 1983;98:184.
Legionella pneumophila and anaerobic bacteria were cultured from a perirectal abscess in an immunosuppressed patient with Legionella pneumonia.
Bangsborg JM, et al. Legionellosis in patients with HIV infection. Infection 1990; 18:342.
Based on observations in 180 patients, the authors conclude that Legionnaires’ disease is uncommon in HIV-infected patients. They did identify, however, two patients infected concurrently with Legionella and Pneumocystis carinii.
Blatt SP, et al. Nosocomial Legionnaires’ disease: aspiration as a primary mode of disease acquisition. Am J Med 1993;95:16.
In this prospective, case-control study of 14 patients with nosocomial Legionnaires’ disease, the authors found evidence that the infections were caused by aspiration.
Dorman SA, Hardin NJ, Winn WC Jr. Pyelonephritis associated with Legionella pneumophila serogroup 4. Ann Intern Med 1980;93:186.
The authors report the isolation of Legionella from the lungs and kidneys of a patient with metastatic bladder cancer and fatal pneumonia.
Dowling JN, Saha AK, Glew RH. Virulence factors of the family Legionellaceae. Microbiol Rev 1992;56:32.
An excellent review of the molecular mechanisms that enable Legionella to persist and replicate within phagocytic cells.
Dowling JN, et al. Pneumonia and multiple lung abscesses caused by dual infection with Legionella micdadei and Legionella pneumophila. Am Rev Respir Dis 1983;127:121.
A patient immunosuppressed by medications and a splenectomy experienced a near-fatal infection caused by two distinct species of Legionella.
Edelstein PH. Antimicrobial therapy for Legionnaire’s disease: a review. Clin Infect Dis 1995;21 (Suppl 3):S265.
Based on an exhaustive review of the laboratory and clinical data concerning the activity and efficacy of macrolides, tetracyclines, and other classes of antimicrobials, this expert recommends that patients with Legionnaires’ disease who are immunocompromised or severely ill be treated with a fluoroquinolone rather than erythromycin.
Evans CP, Winn WC. Extrathoracic localization of Legionella pneumophila in Legionnaires’ pneumonia. Am J Clin Pathol 1981;76:813.
In 6 of 12 cases of fatal Legionnaires’ disease, the bacterium was isolated from extrapulmonary sites, including the spleen, liver, and kidney.
Fang GD, Yu VL, Vickers RM. Disease due to Legionellaceae (other than Legionella pneumophila). Medicine 1989;68:116.
An extensive review of the epidemiology, microbiology, and clinical diseases associated with this family of bacteria.
Fraser DW, et al. Legionnaires’ disease: description of an epidemic of pneumonia. N Engl J Med 1977;297:1189.
In the summer of 1976, American Legion conventioneers in Philadelphia experienced an outbreak of disease that was associated with pneumonia in 90% of those affected and with a case fatality rate of 16%.
Glick TH, et al. Pontiac fever: an epidemic of unknown etiology in a health department. I. Clinical and epidemiologic aspects. Am J Epidemiol 1978;107:149.
Serologic studies confirmed that the acute febrile illness experienced by 95 of the 100 persons employed in the county health department facility in Pontiac, Michigan, was caused by the bacterium responsible for Legionnaires’ disease.
Hamedani P, et al. The safety and efficacy of clarithromycin in patients with Legionella pneumonia. Chest 1991;100:1503.
A role for the use of clarithromycin in treating patients with Legionnaires’ disease is suggested by the results of this study.
Heath CH, Grove DI, Looke DFM. Delay in appropriate therapy for Legionella pneumonia associated with increased mortality. Eur J Clin Microbiol Infect Dis 1996; 15:286.
In this retrospective review of 39 cases of Legionnaires’ disease, the mortality rate was correlated with delays in initiating therapy with erythromycin.
Jaulhac B, et al. Detection of Legionella spp. in bronchoalveolar lavage fluids by DNA amplification. J Clin Microbiol 1992;30:920.
Polymerase chain reaction was more sensitive than bacterial cultures or serologic assays in detecting the presence of Legionella.
Kalweit WH, Winn WC. Hemodialysis fistula infections caused by Legionella pneumophila. Ann Intern Med 1982;96:173.
In two patients receiving long-term hemodialysis, infections developed at fistula sites as complications of Legionnaires’ pneumonia.
Lake KB, et al. Legionnaires’ disease and pulmonary cavitation. Arch Intern Med 1979;139:485.
Legionella has the potential to produce lung abscesses.
Lowry PW, et al. A cluster of Legionella sternal wound infections due to postoperative topical exposure to contaminated tap water. N Engl J Med 1991;324:109.
The authors describe three patients in whom sternal wound infections with Legionella developed as the result of contact with contaminated tap water.
Marrie TJ, et al. Control of endemic nosocomial Legionnaires’ disease by using sterile potable water for high-risk patients. Epidemiol Infect 1991;107:591.
Although Legionnaires’ disease is considered inhalation-based in origin, the authors’ data indicate that the infection can be aspiration-based in etiology. Immunosuppressed patients exposed to contaminated potable water through mechanical ventilation or nasogastric tubes were at greatest risk for the infection.
Maycock R, Skale B, Kohler B. Legionella pneumophila pericarditis proved by culture of pericardial fluid. Am J Med 1983;75:534.
Legionella pneumophila was isolated from the pericardial fluid of a patient with pericarditis complicating pneumonia.
Milder JE, Rough RR. Concurrent Legionnaires’ disease and active pulmonary tuberculosis. Am Rev Respir Dis 1982;125:759.
An elderly patient infected with Mycobacterium tuberculosis and Legionella pneumophila is described.
Passi C, Maddaluno R, Pastoris MC. Incidence of Legionella pneumophila infection in tourists: Italy. Public Health 1990;104:183.
Legionella was isolated from the water systems of 10 Italian hotels that were associated with cases of Legionnaires’ disease.
Plouffe JF, et al. Reevaluation of the definition of Legionnaires’ disease: use of the urinary antigen assay. Clin Infect Dis 1995;20:1286.
A single acute-phase titer above 1:256 did not discriminate between adults with Legionnaires’ disease and those with pneumonia caused by other microbes. The urinary antigen assay rarely gave a false-positive result (1:64) of antibodies to C. pneumoniae experienced a fourfold greater risk for an adverse cardiovascular event during follow-up than did patients without increased titers.
Gurfinkel E, et al. Randomised trial of roxithromycin in non–Q-wave coronary syndromes: ROXIS Pilot Study. Lancet 1997;350:404.
In this prospective, double-blinded study of patients with unstable angina or non–Q-wave myocardial infarction, the use of roxithromycin, a macrolide antibiotic, was associated with a reduction in the rates of recurrent ischemic events, including infarction and death.
Kuo CC, et al. Chlamydia pneumoniae (TWAR). Clin Microbiol Rev 1995;8:451.
An excellent review that focuses on the microbiology, epidemiology, and laboratory diagnosis of disease caused by C. pneumoniae.
Laurila AL, von Hertzen L, Saikku P. Chlamydia pneumoniae and chronic lung diseases. Scand J Infect Dis Suppl 1997;104:34.
C. pneumoniae has been associated with several chronic pulmonary diseases, including asthma, chronic obstructive pulmonary disease, sarcoidosis, and even lung cancer.
Mazzoli S, et al. Chlamydia pneumoniae antibody response in patients with acute myocardial infarction and their follow-up. Am Heart J 1998;135:15.
In contrast to healthy and matched controls, patients with acute myocardial infarction had high titers of antichlamydial antibodies and elevated levels of interleukin 6.
Troy CJ, et al. Chlamydia pneumoniae as a new source of infectious outbreaks in nursing homes. JAMA 1997;277:1214.
In this retrospective investigation of three nursing homes, the authors found that attack rates for respiratory tract infections caused by C. pneumoniae were high among residents and workers and that pneumonia in the debilitated patient was often fatal.