Background

In June 1989, a 71-year-old man (index patient) who lived in a local boarding home developed active tuberculosis. His condition was diagnosed promptly, and the health department was notified. The public health investigation of the patient's contacts identified the man who was later determined to be the source patient.

The source patient was a 32-year-old man who worked at the shipyard and lived in the boarding home with the index patient and seven other men. He had productive cough, fever, night sweats, and weight loss of 18.2 kg from November 1988 until active pulmonary tuberculosis was diagnosed and appropriate therapy was initiated 8 months later. Although the patient had repeatedly sought medical attention for his chronic cough, sore throat, and hoarseness, he had been treated only with tetracycline. A chest roentgenogram obtained in April 1989 at the emergency department of a local hospital was consistent with tuberculosis, but no follow-up took place. The patient returned to the same emergency department 1 month later because of persistent symptoms and was informed of the previous roentgenographic findings. No antituberculous medications were prescribed, and no report was submitted to the Maine Bureau of Health. The patient was advised by the emergency department physician to wear a mask while working in the shipyard to reduce the risk for infecting his coworkers.

The source patient's chest roentgenogram showed several large cavitary lesions in the left upper lobe. Sputum smears contained acid-fast bacilli that were too numerous to count. Treatment with isoniazid and rifampin was started on 18 July 1989. Direct laryngoscopy, done because of the patient's chronic sore throat and hoarseness, showed vocal cord edema and erythema. No biopsy was done. The patient's sputum cultures contained Mycobacterium tuberculosis that was susceptible to all antituberculous drugs. The medical records and the source patient's descriptions of the illness indicated that the source patient had probably had tuberculosis for many months before the index patient became ill.

Methods

Top Methods Results Discussion Author & Article Info References

We defined a patient with an active case of tuberculosis as a resident of the community or an employee of the shipyard who had sputum or tissue cultures that were positive for M. tuberculosis from June 1989 to May 1992. We defined a case of tuberculous infection as a positive tuberculin skin test result in persons from these populations who had had no previous documented positive test result. All tuberculin skin tests were administered with 5 tuberculin units by the Mantoux method and were read by trained public health nurses. Because this was an outbreak investigation, a tuberculin skin reaction with induration of at least 5 mm was considered positive [5, 6].

Persons in two exposure categories, social contacts and occupational contacts, were investigated. Social contacts of the source patient included friends, fellow boarding home residents, and persons who frequented two taverns where the source patient had spent many evenings during his infectious period (from the onset of cough until the initiation of antituberculous therapy). Occupational contacts were persons who worked at the shipyard.

Isolates of M. tuberculosis from the source patient, six outbreak-related cases, and three cases of tuberculosis in Maine that were not related to the outbreak were sent to the Mycobacteriology Laboratory of the Centers for Disease Control and Prevention (CDC) for phage typing [7].

Statistical analyses were done in a univariate manner using the Mantel-Haenszel chi-square test. Statistical testing was done with Epi Info, a software package developed by the CDC [8].

Results

Top Methods Results Discussion Author & Article Info References

From 1989 to 1992, 21 persons who lived in the small Maine community or worked in the local shipyard developed active tuberculosis. They ranged in age from 20 to 78 years (mean age, 45 years); 6 were younger than 35 years of age. Seventeen patients were men, and 19 were white. Eighteen had pulmonary tuberculosis, and 3 had extrapulmonary disease. Eleven patients were hospitalized; 4 died with tuberculosis (1 died of tuberculosis). Twelve patients were shipyard employees, and the others were social contacts of the source patient.

Community Investigation

Five of the nine men who lived in the boarding home (including the index patient and the source patient) developed active tuberculosis. One man had a positive skin test result and a negative chest roentgenogram, one had a negative skin test result, and two were lost to follow-up (Table 1). Family contacts of all boarding home residents with active disease had skin testing; none had a positive test result except those who had also had direct contact with the source patient.

One hundred fifteen social contacts of the source patient had skin testing (Table 1). Of these, 61 (53%) had a positive test result. Another 32 persons were identified as both social and work contacts of the source patient, and 25 (81%) of these had a positive test result. Thirty of 60 persons who were identified as employees or regular patrons of either of the two taverns that the source patient frequented were tested, and 3 (10%) had positive test results.

Shipyard Investigation

In the months before his diagnosis, the source patient worked during all three shifts and in all parts of the shipyard. Workstations on the ships were enclosed, cramped, and crowded. Initially, 184 of the patient's closest work contacts had skin tests; 49 (27%) of these had positive test results (Table 1). After these results were obtained, 1700 additional persons who worked in areas in which the source patient worked (intermediate contacts) were tested; 126 (7.4%) of these had positive test results. At first, this rate was thought to represent the expected background rate for this community. However, because all infected workers were born in the United States and resided in an area with a low risk for tuberculosis, and because most workers were young (53% were younger than 35 years of age), this infection rate was ultimately attributed to recent exposure during the outbreak. Therefore, attempts were made to test all remaining employees; among the 7861 additional persons tested, 453 (5.8%) infected workers were identified. In total, of 9745 workers who agreed to be tested, 628 (6.4%) had positive test results; 464 (74%) of these were older than 35 years of age.

Health personnel who administered the tuberculin skin tests were asked to record only whether a test result was positive (> 5 mm of induration), but some personnel also recorded the size of the skin test reaction. Of 461 shipyard employees whose positive skin test size was recorded, 386 (84%) had an induration of 10 mm or greater; 185 (40%) had an induration of 15 mm or greater; and 40 (9%) had an induration of 20 mm or greater. In 1993, more than 1 year after the outbreak was over, 3264 newly hired shipyard workers were tested; only 32 (0.98%) had positive test results.

Isolates

Seven isolates from outbreak-related active cases, including the isolate from the source patient, were phage type 1, auxiliary type 14. This phage type is detected in fewer than 1% of U.S. tuberculosis isolates [9]. The three strains from Maine that were unrelated to the outbreak did not have this phage type. All outbreak-related strains were susceptible to isoniazid, rifampin, and other antituberculous agents.

Prophylaxis Program

The Maine Bureau of Health Tuberculosis Control Program routinely recommends isoniazid prophylaxis (300 mg/d for 6 months) for persons younger than 35 years of age who have a positive tuberculin skin test result. Persons older than 35 years of age who have positive skin test results are not offered prophylaxis unless they have a previously documented negative test result. In this outbreak, a rationale based on national recommendations was advanced for providing prophylaxis to all exposed and infected persons regardless of age [5, 6]. However, many local physicians and public health officials considered the infection rate to be the expected "background" in this community and not the result of recent exposure. Ultimately, isoniazid prophylaxis was not routinely offered to infected contacts older than 35 years of age.

Among the 697 infected contacts of the source patient, only 341 (49%) completed isoniazid prophylaxis. The rate of completing prophylaxis varied according to the type of contact and age of the infected person. Of 69 infected social contacts, 52 (75%) completed isoniazid prophylaxis. Among the 628 infected shipyard workers, 289 (46%) completed prophylaxis. Of workers younger than 35 years of age, 71% (116 of 164) completed isoniazid prophylaxis compared with only 37% (173 of 464) of older workers (relative risk, 1.9 [95% CI, 1.63 to 2.21; P < 0.001]). Only one person developed tuberculosis despite completing isoniazid prophylaxis.

Discussion

Top Methods Results Discussion Author & Article Info References

A large outbreak of tuberculosis occurred in a small Maine community populated predominately by white, middle-class persons born in the United States. Twenty-one cases of active tuberculosis occurred in this community, which had reported fewer than 10 cases of tuberculosis in the decade before the outbreak. Six hundred ninety-seven tuberculous infections were identified among shipyard workers and local residents. A highly infectious source patient was identified. Closer contact with this patient was associated with both an increased risk for becoming infected and an increased risk for developing tuberculous disease. The highest infection rates (81%) were recorded in persons who had both work and social contact with the source patient. Most notably, 67% of the source patient's fellow boarding home residents who were available for follow-up developed active disease.

The number of cases of tuberculous infection and disease in this small community cannot be dismissed as the expected background rate in the area. The phage type of M. tuberculosis found in all seven of the outbreak-associated isolates that were tested has rarely been identified in the United States. The incidence of tuberculosis in Maine had been low and almost nonexistent in the affected town. Not surprisingly, surveys from 1980 to 1986 showed that only 0.79% of naval military recruits from the entire New England region had positive tuberculin skin test results [10]. Furthermore, in 1993, fewer than 1% of newly hired workers who had no previous exposure to the source patient had positive tuberculin skin test results. In this small town, infection rates greater than 5% are many times the currently expected rate and are probably the result of recent infection. Although the size of the indurated area was recorded only for 461 of the workers who were tested, nearly all of these indurations were greater than 10 mm; 40% were greater than 15 mm. Such large skin test reactions are consistent with recently acquired infection [11, 12].

In many cases, tuberculosis is a preventable disease. In the United States, patients with newly diagnosed tuberculosis transmit the disease to an average of only 1.5 other persons [13]. However, recent studies show that as many as 10% of patients with tuberculosis have been highly infectious, probably accounting for most instances of transmission [14]. Prompt diagnosis and treatment of the source patient would have limited the transmission of disease to others within the community and at the source patient's worksite. The outbreak would probably have been smaller if more infected persons had been given isoniazid prophylaxis.

Between 5% and 10% of newly infected persons who do not receive prophylaxis eventually develop tuberculosis [15-18]. Approximately half of these cases develop within 2 years of infection; the remainder occur in later years, often remote from the time of infection [15]. In the Maine outbreak, 356 infected persons did not complete isoniazid prophylaxis. It could have been predicted, therefore, that 18 active cases would occur before 1993 (20 were actually identified). Another 18 active cases are likely to occur at some time during the lives of persons exposed in this outbreak. Furthermore, a large cohort of untreated persons residing in the community could be the lurking nidus for future active cases in an area with a previous low incidence of tuberculosis.

Epidemic-related cases of active tuberculosis occurred in this community for more than 3 years after the outbreak was recognized. The outbreak shows that the resurgent national epidemic of tuberculosis can affect middle-class healthy persons in small towns. However, most of the attention devoted to tuberculosis in the United States has focused on other populations. Physicians in all settings need to become familiar with the recently revised recommendations for the treatment of both tuberculous infection and tuberculous disease. It is perhaps not surprising that physicians who rarely or never see a case of tuberculosis are not familiar with the most recent tuberculosis prevention and treatment guidelines [19, 20] or the special recommendations that pertain to outbreak settings. Health departments must make greater efforts to enlist the collaboration of private physicians in providing appropriate chemoprophylaxis to high-risk contacts of infectious cases if the increase in tuberculosis is to be curtailed.

Dr. Gensheimer and Ms. Lewis: Maine Bureau of Health, State House #11, Augusta, ME 04333.

Dr. Bloch: Division of Tuberculosis Elimination, Centers for Disease Control and Prevention, Atlanta, GA 30333.

Dr. Parrotte: Unit N-8, 20 Chestnut Street, Exeter, NH 03833.

Dr. Horan: Epidemiology Program Office, Centers for Disease Control and Prevention, Atlanta, GA 30333.

Dr. Schaffner: Department of Preventive Medicine, Vanderbilt University School of Medicine, Nashville, TN 37232.