- Email: [email protected]
Nosocomial transmission of tuberculosis from unsuspected disease
Nosocomial Transmission of Tuberculosis from Unsuspected Disease
HARVEY RITA SHARON
S. POBLETE, L.
North Chicago,
KANTOR,
M.D.
M.D.* PUSATERI,
M.S.N.
Illinois
From the Laboratory Service, Medical Service, and Nursing Service, Veterans Administration Medical Center, and the Division of infectious Diseases, Department of Medicine, and the Department of Pathology, University of Health Sciences/Chicago Medical School, North Chicago, Illinois. This work was presented in part at the 87th Annual Meeting of the American Society for Microbiology, Atlanta, Georgia, March 5, 1987. Requests for reprints should be addressed to Dr. Harvey S. Kantor, Veterans Administration Medical Center, Laboratory Service (113), North Chicago, Illinois 80084. Manuscript submitted April 6, 1987, and accepted in revised form March 4, 1988. Current address: University of Miami Hospital, 1776 N.W. 10th Avenue, Miami, Florida. l
Tuberculosis remains an occupational hazard for hospital employees. A case of acute generalized tuberculosis occurring in a patient with a malignancy who had received corticosteroids was undetected during life and during a gross autopsy examination. Histologic examination of tissue performed one month later was necessary to establish the diagnosis. Of susceptible hospital staff members who were exposed to the index case, infection developed in nine of 56 (16 percent) compared with three of 333 (0.9 percent) unexposed personnel with similar risk but no known exposure (p
May
1988
The American
Journal
of Medicine
Volume
84
833
NOSOCOMIAL
TRANSMISSION
OF TUBERCULOSIS-KANTOR
ET AL
CASE REPORTS
Secondary Case. Five months after being present at an autopsy, one of the staff members had gradual development of shortness of breath and right-sided chest pain, which persisted for one month. The staff member denied any cough, fever, night sweats, weight loss, or loss of appetite. Physical examination was only remarkable for lung dullness at the right lower lung base. A tuberculin skin test, given for the first time, was positive at 10 mm of induration. This staff member had given a history of receiving bacillus Calmette-Guerin vaccine (BCG) at the age of 10 and had been excluded from pre-employment and annual tuberculin skin testing. A chest radiograph revealed a right pleural effusion without any pulmonary infiltrates. Two liters of yellow, slightly cloudy fluid were obtained by thoracentesis. It was consistent with an exudate, with total protein of 5.5 g/dl, lactic dehydrogenase of 366 units/liter, glucose of 81 mg/dl, red blood cell count of 3,600/mm3, and white blood cell count of 4,500/mm3, with 83 percent mononuclear cells and 17 percent polymorphonuclear cells. Result of acid-fast smear of the fluid was negative, and culture grew M. tuberculosis. A pleural biopsy specimen revealed granulomas containing acid-fast bacilli; M. tuberculosis was identified on culture. There was no history of exposure to tuberculosis outside of the index case.
Index Case. A 50-year-old male patient was admitted to the hospital in late 1984, with a diagnosis of small cell carcinoma of the lung with brain metastasis. During the first two months, he received chemotherapy, steroids (dexamethasone, 4 mg orally, four times a day), and radiation with symptomatic improvement. During therapy with steroids, diabetes mellitus developed and was controlled with insulin. One month later, he began to complain of dry cough, shortness of breath, right pleuritic chest pain, and fever. He was found to have a right lower lobe empyema that necessitated a chest tube for three months. This was complicated by the development of a bronchopleural fistula. He was treated empirically with clindamycin, gentamicin, and amphoteritin. Hydrocortisone, 100 mg, was added to each infusion of amphotericin, which was continued for two weeks. Results of pleural fluid studies for bacteria, fungi, mycobacteria, and cytology were negative. He continued to receive radiation and chemotherapy. In June 1985, one month after his empyema resolved, he complained of weakness, loss of appetite, weight loss, shortness of breath, and a low-grade fever. Chest radiograph revealed right upper lobe and left lower lobe pneumonias. Treatment with tobramycin and clindamycin was started but symptoms were persistent. Sputum cultures showed mixed bacterial flora. One sputum acid-fast smear and culture gave negative results. Chest radiographic infiltrates remained unchanged and were thought to represent cancer. The patient’s condition continued to deteriorate. Throughout July 1985, he had extensive pulmonary congestion and produced large amounts of sputum, necessitating vigorous nasotracheal suctioning. Signs of pulmonary consolidation continued until his death in late July. At autopsy, gross examination revealed extensive bilateral necrotizing bronchopneumonia and metastatic tumor to the cerebellum. Tuberculosis was not suspected; cultures for mycobacteria were not performed and respiratory infection control precautions were not instituted. During the performance of the autopsy, an electric bone saw was used to prepare sections of vertebrae. On histologic examination, performed approximately one month later, lung sections showed acute tuberculous pneumonia with extensive destruction of the lung parenchyma due to large areas of acute necrotizing inflammation. The alveolar sacs around the necrotic areas were filled with fibrinous exudate that contained many inflammatory cells including histiocytes (Figure 1, top left). A large number of acid-fast bacilli were seen within the exudate (Figure 1, top right). The bacilli were small, had a beaded appearance, and were consistent with tubercle bacilli. Fungal stains were negative. Caseating granulomas containing few acid-fast bacilli were found on microscopic examination of the spleen and bone marrow (vertebral) (Figure 1, bottom left and bottom right). Death was attributed to tuberculous bronchopneumonia. A review of the patient’s hospital record revealed that he had a positive tuberculin skin test result in 1981. A postmortem review of the patient’s chest radiographs were reinterpreted and several cavities within infiltrates were seen that were not recognized on the initial review.
034
May
1988
The American
Journal
of Medicine
Volume
METHODS Epidemiologic Investigation. As part of their pre-employment physical examination, prospective hospital employees undergo a tuberculin skin test and chest radiography. Skin test conversion was defined as a change in tuberculin reaction from less than 10 mm in diameter to 10 mm or more in diameter and increases by at least 6 mm within a period of two years [ 131. Since 1982, two-step testing to detect boosted reactions has been used for persons whose initial skin test results are negative. A second skin test is given one week later. Persons with negative findings on skin tests are offered repeat tests at six- to IP-month intervals, depending on exposure. In 1984, tuberculin skin testing of BCG vaccine recipients became routine for new employees. Prior to this time they were not tested. Upon the diagnosis of pleural tuberculosis in the pathologist who had performed the autopsy on a patient with unsuspected generalized tuberculosis, a contact investigation was initiated [ 141. The study period was February 1986 through August 1986. Other hospital personnel who had shared air with this patient while he was alive were identified and included in our survey. The index case was in a single patient room and no hospital patients were found to be at risk. Questionnaires were sent to all involved. Each employee was asked to discuss contact in terms of duration of time and closeness of contact to the patient. Specific information about previous tuberculous disease, prior tuberculin skin test status, and recent symptoms compatible with tuberculosis was elicited. There were three histotechnologists who assisted the pathologist during this particular autopsy. They were also considered to be potentially infected contacts. All contacts were given an intracutaneous Mantoux test with the standard 5-tuberculin-unit dose of purified protein derivative-S, unless there was a confirmed, previously pos-
84
NOSOCOMIAL
TRANSMISSION
OF TUBERCULOSIS-KANTOR
ET
AL
Figure 1. Top /eff, acute tuberculous pneumonia. Lung architecture partially preserved with acute inflammation and caseous necrosis. No discrete granuloma. Histiocytes absent in this field (hematoxylin and eosin stain; magnification X 100). Top righf, numerous extracellular acid-fast bacilli in caseous necrotic lung exudate. Scattered neotrophils and nuclear fragments (Kinyoun stain; magnification X 1,000). Bottom left, well-defined granuloma with a multinucleated giant cell and caseous necrosis in vertebral bone marrow (hematoxylin and eosin stain; magnification X 700). Bottom right, rare acid-fast bacilli in vertebral bone marrow necrotic exudate (Kinyoun stain; magnification X 1,000).
known to react to tuberculin on prior testing (Figure 2). The secondary case, an autopsy staff member previously vaccinated with BCG at the age of IO years, had not received a tuberculin skin test prior to exposure to the index case. The development of tuberculous pleuritis five months after contact fits the current view of pleural space
itive tuberculin skin test reaction. A stabilized tuberculin (Aplisal; Parke-Davis and Company, Detroit, Michigan, 0.1 ml) was used throughout the study. Disposable tuberculin syringes (27-gauge needle) were filled immediately prior to use. Induration was recorded on the third day after application. For evaluating the results of the Mantoux test among tuberculosis contacts, a cut point of greater than or equal to 5 mm of induration was used [13]. All contacts with a productive cough of recent onset had sputum samples submitted for mycobacterial smear and culture, and chest radiography to exclude active pulmonary tuberculosis. A chi-square analysis was used to determine the statistical significance of between-group differences.
localization
occurring
with primary
infection
[15]. Of the
remaining 55 susceptible hospital staff members, including three involved in the autopsy, eight (14.5 percent) had conversion on tuberculin skin testing (Tables I and II). All of those in whom conversion occurred were aSymptOtEitic except for two. The first patient identified had a pleural effusion and negative pleural fluid culture. The second had radiologically typical apical cavitary disease with sputum culture confirmation. Among the 24 employees with previous positive tuberculin test results, six had productive coughs. All had negative chest radiographic findings and four of six had
RESULTS A total of 80 hospital employees were considered in the surveillance, including four members of the autopsy staff; 55 were previously known to be nonreactive to tuberculin
and were considered susceptible whereas 24 were
May
1988
The American
Journal
of Medicine
Volume
84
835
NOSOCOMIAL
TRANSMISSION
OUTCOME
OF TUBERCULOSIS-KANTOR
OF CONTACT
ET AL
negative sputum culture results (two were non-compliant). During this contact investigation, nine of 56 (16 percent) susceptible hospital staff who were exposed to the index case became infected compared with three of 333 (0.9 percent) unexposed personnel with similar risk but no known exposure (p
INVESTIGATION
r
I 80 employees
exposed
to index
case
1 1 I
I
I I Post-exposure given
1 employee with ECG history
Mantoux
I
I
test
I
4 employees with productive coughs: Negative Sputum cultures 8 chest x-rays
18 employees asymptomatic
Figure 2. Outcome of contact investigation conducted between February 1986 and August 1986 among hospital employees exposed to the index case. TB = tuberculosis; BCG = bacillus Calmette-Guerin.
Conversions Number Percent
Personnel Exposed (n = 55) Unexposed (n = 333)
II
Subjects
14.5 0.9
8 3
with Tuberculin
Professional Employment Position Place
Subiect
Conversion Skin Test Reaction (mm induration)
ar ar ar mw mw mw x-ray mw a = autopsy staff; autopsy room: mw
836
Within the past three decades, the presentation of generalized tuberculosis has changed radically. Its occurrence as a clinical complication of primary childhood infection has diminished, whereas its cryptic emergence in a much older age group is on the increase [9,16]. Moreover, in many of these aged patients, the diagnosis is made only at autopsy, since either concomitant illness masked the clinical picture, or the disease presented in an insidious fashion, unassociated with the typical hallmarks of miliary tuberculosis [7-9,16- 191. Of 100 autopsies in patients who died of this entity at Johns Hopkins Hospital between 1937 and 1959, 20 percent of patients exhibited no constitutional symptoms prior to hospitalization. A history of tuberculosis often was lacking, fever curves and hematologic findings commonly were nonspecific, monocytosis frequently was absent, chest radiographs were nondiagnostic in about 50 percent of the cases, and anergy occurred, particularly in the elderly [7]. Tuberculosis may even be missed during the gross autopsy and histologic examination may be necessary for the diagnosis [ 161, as occurred in this case. The present case study demonstrated nosocomial transmission of tuberculosis from patient to staff. Airborne transmission from the index patient to the four autopsy staff members probably resulted from inhalation of aerosolized droplet nuclei generated by compression of the lungs during the postmortem examination, given the large numbers of acid-fast bacilli seen on histologic examina-
Tuberculin Skin Test Conversions as a Function of Exposure to Unsuspected Tuberculosis
TABLE I
TABLE
COMMENTS
May
1988
n = nursing staff; = medical ward.
The American
16 15 15 15 20 20 15 5 r = radiology
Journal
of Medicine
staff;
ar =
Volume
84
NOSOCOMIAL
tion. Whether an infectious aerosol was also contributed by the cutting of vertebral sections by an electric bone saw cannot be ruled out [20]. It appears that during the time the index case was on the medical ward, he infected five persons. His respiratory problems, including pulmonary congestion, copious sputum production, and vigorous naso-tracheal suctioning, were likely to have resulted in frequent aerosols of infected sputum. The patient appeared to be highly infectious, yet a single sputum smear and culture specimen obtained one month before his death showed no growth of mycobacteria. It is possible that this false-negative smear and culture resulted from analysis of a salivary specimen rather than from a sputum sample [21]. Mycobacteria may be released from the lung irregularly, resulting in cultures that may be negative one day and positive the next. For this reason, a minimum of three to five specimens should be obtained from patients suspected of having pulmonary tuberculosis [22]. Riley and colleagues [23] have shown that the infectivity of tuberculosis was related to the number of air changes per hour in the room. Air control is necessary in rooms of known or suspected transmitters and in places where persons with untreated tuberculosis might contaminate the air [24]. Infection of susceptible personnel in contact with patients with unrecognized active tuberculosis can be prevented by the correct use of mechanical ventilation and ultraviolet irradiation of the air [25]. Patients’ rooms should have fresh air introduced through a ventilatory system and preferably exhausted to the outside. United States Public Health Service standards require air disinfection in hospital areas housing immunocompromised patients [26], and call for a minimum ventilation rate of six air changes per hour in isolation rooms [25]. In areas where the nature of the work is exceptionally hazardous, such as autopsy rooms, 15 to 25 air changes per hour [27,28] have been recommended. However, even high-
TABLE III
Year 1979 1980
1981 1982 1983 1984 1985 1986+
TRANSMISSION
OF TUBERCULOSIS-KANTOR
ET AL
Annual Tuberculin Skin Test Conversion Rate in Hospital Personnel from 1979 to 1988”
Number Tested
Conversions Number
Percent
4 2 5 3 3 2 9 3
0.9 0.5 1.5 0.8 0.8 0.5 1.45 0.9
458 429 333 389 359 380 a20 333
Mean skin test conversion rate + January 1986 to June 1966. l
= 0.92
percent.
change-rate ventilating systems may not be adequate to remove “burst source” contaminants before they disperse. Local exhaust and primary barrier systems are preferred because they control the contaminants at their source and minimize exposure to personnel [29]. At present, there are no uniform infection control guidelines for handling suspected or confirmed cases of tuberculosis at autopsy [30]. Each institution has had to develop its own policies based upon the recommended guidelines in the use of air control and containment of infectious aerosols. The index case had not received isoniazid chemoprophylaxis. As a patient with known tuberculin reaction who had received systemic corticosteroids and immunosuppressive therapy for several weeks, he fits the special clinical situations considered to clearly benefit from preventive therapy with isoniazid [3 11. ACKNOWLEDGMENT We thank Lorna Waiters, M.P.H., for her technical assistance, and Patricia W. Wellman for secretarial assistance.
REFERENCES 1. 2.
3.
4. 5. 6.
Myers JA: Tuberculosis: a half century of study and conquest. St. Louis: Warren H. Green, 1970; 166-199. Ehrenkranz NJ, Kicklighter L: Tuberculosis outbreak in a general hospital: evidence for airborne spread. Ann Intern Med 1972; 77: 377-382. Craven RB, Wenzel RP, Atuk N: Minimizing tuberculosis risk to hospital personnel and students exposed to unsuspected disease. Ann Intern Med 1975; 82: 628-632. Catanzaro A: Nosocomial tuberculosis. Am Rev Respir Dis 1982; 125: 559-562. Hershfield ES: Tuberculosis in the world. Chest 1979; 765: 8055-8115. Edlin GP: Active tuberculosis unrecognized until necropsy. Lancet 1978; I: 650-652.
7.
8.
9. 10. 11.
Slavin RE, Walsh TJ, Pollack AD: Late generalized tuberculosis: a clinical and pathologic analysis and comparison of 100 cases in the preantibiotic and antibiotic eras. Medicine (Baltimore) 1980; 59: 352-366. Enarson DA, Grzybwsk S, Dorken E: Failure of diagnosis as a factor in tuberculosis mortality. Can Med Assoc J 1978; 118: 1520-1522. Proudfoot AT, Akhtar AJ, Douglas AC, Horne NW: Miliary tuberculosis in adults. Br Med J 1969; 2: 273-276. Reid DD: Incidence of tuberculosis among workers in medical laboratories. Br Med J 1957; 2: 10-14. Harrington JM, Shannon HS: Incidence of tuberculosis, hepatitis, brucellosis and shigellosis in British medical laboratory workers. Br Med J 1976; 1: 759-762.
May 1988 The American Journal of Medicine Volume 84
837
NOSOCOMIAL
12.
13. 14.
15.
16. 17.
18. 19.
20.
21.
22.
838
TRANSMISSION
OF TUBERCULOSIS-KANTOR
ET Al
Pike RM: Laboratory-associated infections: incidences, fatalities, causes, and prevention. Annu Rev Microbial 1979; 33: 41-66. American Thoracic Society: The tuberculin skin test. Am Rev Respir Dis 1981: 124: 356-363. American Thoracic Society: Guidelines for the investigation and management of tuberculosis contacts. Am Rev Respir Dis 1976; 114: 459-463. Des Prez RM, Goodwin RA: Mycobacterium tuberculosis. In: Mandell GL, Douglas RG, Bennett JE, eds. Principles and practices of infectious disease, 2nd ed. New York: John Wiley, 1985; 1383-1406. Treip C, Meyers D: Fatal tuberculosis in a general hospital. Lancet 1959; I: 164-167. Weir MR, Thornton GF: Extra pulmonary tuberculosis. Experience of a community hospital and review of the literature. Am J Med 1985; 79: 467-478. Babrowitz ID: Active tuberculosis undiagnosed until autopsy. Am J Med 1982; 72: 650-658. Makela V, Ala-Kulju K, Hulst J, Siljander T: Tuberculosisdiagnosed and undiagnosed-as a cause of death. Autopsy observations. Stand J Respir Dis 1971; 52: 13-18. Mackel DC, Forney JE: Overview of the epidemiology of laboratory-acquired infections. In: Miller BM, ed. Laboratory safety: principles and practices. Washington: American Society for Microbiology, 1986; 37-42. Isaac-Renton JL, Puselja BB, Allen EA, Grzybowski S, Black WA: Microscopic evaluation of sputum specimens submitted for Mycobacterium tuberculosis culture. Am J Clin Pathol 1985; 84: 361-363. Sommers HM, McClatchy JK: Cumitech 16, laboratory diagnosis of the mycobacterioses. Morello JA, coordinating ed. Washington: American Society for Microbiology,
May
1988
The
American
Journal
of Medicine
Volume
23.
24.
25.
26.
27.
28.
29.
30. 31.
84
1983. Riley RL, Mills CC, O’Grady F, et al: infectiousness of air from a tuberculosis ward. Ultraviolet irradiation of infected air: comparative infectiousness of different patients. Am Rev Respir Dis 1962; 85: 51 l-525. Centers for Disease Control: Guidelines for the prevention of TB transmission in hospitals (publication no. 82-8371). Atlanta: U.S. Department of Health and Human Services, Public Health Service, October 1982. Centers for Disease Control: Notes on air hygiene. Summary of conference on air disinfection. Arch Environ Health 197 1; 23: 473-474. Reiley RL: Principles of UV air disinfection. U.S. Department of Health, Education, and Welfare, Public Health Service publication no. 00-2215, January 1972. McCracken RC: Facility engineering, planning and construction letter (publication no. IL lOA4-82-23). Departments of Medicine and Surgery, Veterans Administration. Heating, Ventilation, Air Conditioning Criteria, November 1982. U.S. Department of Health, Education and Welfare: Minimum requirements of construction and equipment for hospitals and medical facilities (DHEW publication no. 79-14500). Washington: U.S. Government Printing Office, 1979. West DL, Chatigny MA: Design of microbiological and biomedical research facilities. In: Miller BM, ed. Laboratory safety: principles and practices. Washington: American Society for Microbiology, 1986; 124-137. Orenstein JM: Guidelines for high risk or potentially high risk autopsy cases. Pathologist 1984; 38: 33-34. National Consensus Conference on Tuberculosis. Preventive treatment of tuberculosis. Chest 1985; 87: 128S1325.
HARVEY RITA SHARON
S. POBLETE, L.
North Chicago,
KANTOR,
M.D.
M.D.* PUSATERI,
M.S.N.
Illinois
From the Laboratory Service, Medical Service, and Nursing Service, Veterans Administration Medical Center, and the Division of infectious Diseases, Department of Medicine, and the Department of Pathology, University of Health Sciences/Chicago Medical School, North Chicago, Illinois. This work was presented in part at the 87th Annual Meeting of the American Society for Microbiology, Atlanta, Georgia, March 5, 1987. Requests for reprints should be addressed to Dr. Harvey S. Kantor, Veterans Administration Medical Center, Laboratory Service (113), North Chicago, Illinois 80084. Manuscript submitted April 6, 1987, and accepted in revised form March 4, 1988. Current address: University of Miami Hospital, 1776 N.W. 10th Avenue, Miami, Florida. l
Tuberculosis remains an occupational hazard for hospital employees. A case of acute generalized tuberculosis occurring in a patient with a malignancy who had received corticosteroids was undetected during life and during a gross autopsy examination. Histologic examination of tissue performed one month later was necessary to establish the diagnosis. Of susceptible hospital staff members who were exposed to the index case, infection developed in nine of 56 (16 percent) compared with three of 333 (0.9 percent) unexposed personnel with similar risk but no known exposure (p
May
1988
The American
Journal
of Medicine
Volume
84
833
NOSOCOMIAL
TRANSMISSION
OF TUBERCULOSIS-KANTOR
ET AL
CASE REPORTS
Secondary Case. Five months after being present at an autopsy, one of the staff members had gradual development of shortness of breath and right-sided chest pain, which persisted for one month. The staff member denied any cough, fever, night sweats, weight loss, or loss of appetite. Physical examination was only remarkable for lung dullness at the right lower lung base. A tuberculin skin test, given for the first time, was positive at 10 mm of induration. This staff member had given a history of receiving bacillus Calmette-Guerin vaccine (BCG) at the age of 10 and had been excluded from pre-employment and annual tuberculin skin testing. A chest radiograph revealed a right pleural effusion without any pulmonary infiltrates. Two liters of yellow, slightly cloudy fluid were obtained by thoracentesis. It was consistent with an exudate, with total protein of 5.5 g/dl, lactic dehydrogenase of 366 units/liter, glucose of 81 mg/dl, red blood cell count of 3,600/mm3, and white blood cell count of 4,500/mm3, with 83 percent mononuclear cells and 17 percent polymorphonuclear cells. Result of acid-fast smear of the fluid was negative, and culture grew M. tuberculosis. A pleural biopsy specimen revealed granulomas containing acid-fast bacilli; M. tuberculosis was identified on culture. There was no history of exposure to tuberculosis outside of the index case.
Index Case. A 50-year-old male patient was admitted to the hospital in late 1984, with a diagnosis of small cell carcinoma of the lung with brain metastasis. During the first two months, he received chemotherapy, steroids (dexamethasone, 4 mg orally, four times a day), and radiation with symptomatic improvement. During therapy with steroids, diabetes mellitus developed and was controlled with insulin. One month later, he began to complain of dry cough, shortness of breath, right pleuritic chest pain, and fever. He was found to have a right lower lobe empyema that necessitated a chest tube for three months. This was complicated by the development of a bronchopleural fistula. He was treated empirically with clindamycin, gentamicin, and amphoteritin. Hydrocortisone, 100 mg, was added to each infusion of amphotericin, which was continued for two weeks. Results of pleural fluid studies for bacteria, fungi, mycobacteria, and cytology were negative. He continued to receive radiation and chemotherapy. In June 1985, one month after his empyema resolved, he complained of weakness, loss of appetite, weight loss, shortness of breath, and a low-grade fever. Chest radiograph revealed right upper lobe and left lower lobe pneumonias. Treatment with tobramycin and clindamycin was started but symptoms were persistent. Sputum cultures showed mixed bacterial flora. One sputum acid-fast smear and culture gave negative results. Chest radiographic infiltrates remained unchanged and were thought to represent cancer. The patient’s condition continued to deteriorate. Throughout July 1985, he had extensive pulmonary congestion and produced large amounts of sputum, necessitating vigorous nasotracheal suctioning. Signs of pulmonary consolidation continued until his death in late July. At autopsy, gross examination revealed extensive bilateral necrotizing bronchopneumonia and metastatic tumor to the cerebellum. Tuberculosis was not suspected; cultures for mycobacteria were not performed and respiratory infection control precautions were not instituted. During the performance of the autopsy, an electric bone saw was used to prepare sections of vertebrae. On histologic examination, performed approximately one month later, lung sections showed acute tuberculous pneumonia with extensive destruction of the lung parenchyma due to large areas of acute necrotizing inflammation. The alveolar sacs around the necrotic areas were filled with fibrinous exudate that contained many inflammatory cells including histiocytes (Figure 1, top left). A large number of acid-fast bacilli were seen within the exudate (Figure 1, top right). The bacilli were small, had a beaded appearance, and were consistent with tubercle bacilli. Fungal stains were negative. Caseating granulomas containing few acid-fast bacilli were found on microscopic examination of the spleen and bone marrow (vertebral) (Figure 1, bottom left and bottom right). Death was attributed to tuberculous bronchopneumonia. A review of the patient’s hospital record revealed that he had a positive tuberculin skin test result in 1981. A postmortem review of the patient’s chest radiographs were reinterpreted and several cavities within infiltrates were seen that were not recognized on the initial review.
034
May
1988
The American
Journal
of Medicine
Volume
METHODS Epidemiologic Investigation. As part of their pre-employment physical examination, prospective hospital employees undergo a tuberculin skin test and chest radiography. Skin test conversion was defined as a change in tuberculin reaction from less than 10 mm in diameter to 10 mm or more in diameter and increases by at least 6 mm within a period of two years [ 131. Since 1982, two-step testing to detect boosted reactions has been used for persons whose initial skin test results are negative. A second skin test is given one week later. Persons with negative findings on skin tests are offered repeat tests at six- to IP-month intervals, depending on exposure. In 1984, tuberculin skin testing of BCG vaccine recipients became routine for new employees. Prior to this time they were not tested. Upon the diagnosis of pleural tuberculosis in the pathologist who had performed the autopsy on a patient with unsuspected generalized tuberculosis, a contact investigation was initiated [ 141. The study period was February 1986 through August 1986. Other hospital personnel who had shared air with this patient while he was alive were identified and included in our survey. The index case was in a single patient room and no hospital patients were found to be at risk. Questionnaires were sent to all involved. Each employee was asked to discuss contact in terms of duration of time and closeness of contact to the patient. Specific information about previous tuberculous disease, prior tuberculin skin test status, and recent symptoms compatible with tuberculosis was elicited. There were three histotechnologists who assisted the pathologist during this particular autopsy. They were also considered to be potentially infected contacts. All contacts were given an intracutaneous Mantoux test with the standard 5-tuberculin-unit dose of purified protein derivative-S, unless there was a confirmed, previously pos-
84
NOSOCOMIAL
TRANSMISSION
OF TUBERCULOSIS-KANTOR
ET
AL
Figure 1. Top /eff, acute tuberculous pneumonia. Lung architecture partially preserved with acute inflammation and caseous necrosis. No discrete granuloma. Histiocytes absent in this field (hematoxylin and eosin stain; magnification X 100). Top righf, numerous extracellular acid-fast bacilli in caseous necrotic lung exudate. Scattered neotrophils and nuclear fragments (Kinyoun stain; magnification X 1,000). Bottom left, well-defined granuloma with a multinucleated giant cell and caseous necrosis in vertebral bone marrow (hematoxylin and eosin stain; magnification X 700). Bottom right, rare acid-fast bacilli in vertebral bone marrow necrotic exudate (Kinyoun stain; magnification X 1,000).
known to react to tuberculin on prior testing (Figure 2). The secondary case, an autopsy staff member previously vaccinated with BCG at the age of IO years, had not received a tuberculin skin test prior to exposure to the index case. The development of tuberculous pleuritis five months after contact fits the current view of pleural space
itive tuberculin skin test reaction. A stabilized tuberculin (Aplisal; Parke-Davis and Company, Detroit, Michigan, 0.1 ml) was used throughout the study. Disposable tuberculin syringes (27-gauge needle) were filled immediately prior to use. Induration was recorded on the third day after application. For evaluating the results of the Mantoux test among tuberculosis contacts, a cut point of greater than or equal to 5 mm of induration was used [13]. All contacts with a productive cough of recent onset had sputum samples submitted for mycobacterial smear and culture, and chest radiography to exclude active pulmonary tuberculosis. A chi-square analysis was used to determine the statistical significance of between-group differences.
localization
occurring
with primary
infection
[15]. Of the
remaining 55 susceptible hospital staff members, including three involved in the autopsy, eight (14.5 percent) had conversion on tuberculin skin testing (Tables I and II). All of those in whom conversion occurred were aSymptOtEitic except for two. The first patient identified had a pleural effusion and negative pleural fluid culture. The second had radiologically typical apical cavitary disease with sputum culture confirmation. Among the 24 employees with previous positive tuberculin test results, six had productive coughs. All had negative chest radiographic findings and four of six had
RESULTS A total of 80 hospital employees were considered in the surveillance, including four members of the autopsy staff; 55 were previously known to be nonreactive to tuberculin
and were considered susceptible whereas 24 were
May
1988
The American
Journal
of Medicine
Volume
84
835
NOSOCOMIAL
TRANSMISSION
OUTCOME
OF TUBERCULOSIS-KANTOR
OF CONTACT
ET AL
negative sputum culture results (two were non-compliant). During this contact investigation, nine of 56 (16 percent) susceptible hospital staff who were exposed to the index case became infected compared with three of 333 (0.9 percent) unexposed personnel with similar risk but no known exposure (p
INVESTIGATION
r
I 80 employees
exposed
to index
case
1 1 I
I
I I Post-exposure given
1 employee with ECG history
Mantoux
I
I
test
I
4 employees with productive coughs: Negative Sputum cultures 8 chest x-rays
18 employees asymptomatic
Figure 2. Outcome of contact investigation conducted between February 1986 and August 1986 among hospital employees exposed to the index case. TB = tuberculosis; BCG = bacillus Calmette-Guerin.
Conversions Number Percent
Personnel Exposed (n = 55) Unexposed (n = 333)
II
Subjects
14.5 0.9
8 3
with Tuberculin
Professional Employment Position Place
Subiect
Conversion Skin Test Reaction (mm induration)
ar ar ar mw mw mw x-ray mw a = autopsy staff; autopsy room: mw
836
Within the past three decades, the presentation of generalized tuberculosis has changed radically. Its occurrence as a clinical complication of primary childhood infection has diminished, whereas its cryptic emergence in a much older age group is on the increase [9,16]. Moreover, in many of these aged patients, the diagnosis is made only at autopsy, since either concomitant illness masked the clinical picture, or the disease presented in an insidious fashion, unassociated with the typical hallmarks of miliary tuberculosis [7-9,16- 191. Of 100 autopsies in patients who died of this entity at Johns Hopkins Hospital between 1937 and 1959, 20 percent of patients exhibited no constitutional symptoms prior to hospitalization. A history of tuberculosis often was lacking, fever curves and hematologic findings commonly were nonspecific, monocytosis frequently was absent, chest radiographs were nondiagnostic in about 50 percent of the cases, and anergy occurred, particularly in the elderly [7]. Tuberculosis may even be missed during the gross autopsy and histologic examination may be necessary for the diagnosis [ 161, as occurred in this case. The present case study demonstrated nosocomial transmission of tuberculosis from patient to staff. Airborne transmission from the index patient to the four autopsy staff members probably resulted from inhalation of aerosolized droplet nuclei generated by compression of the lungs during the postmortem examination, given the large numbers of acid-fast bacilli seen on histologic examina-
Tuberculin Skin Test Conversions as a Function of Exposure to Unsuspected Tuberculosis
TABLE I
TABLE
COMMENTS
May
1988
n = nursing staff; = medical ward.
The American
16 15 15 15 20 20 15 5 r = radiology
Journal
of Medicine
staff;
ar =
Volume
84
NOSOCOMIAL
tion. Whether an infectious aerosol was also contributed by the cutting of vertebral sections by an electric bone saw cannot be ruled out [20]. It appears that during the time the index case was on the medical ward, he infected five persons. His respiratory problems, including pulmonary congestion, copious sputum production, and vigorous naso-tracheal suctioning, were likely to have resulted in frequent aerosols of infected sputum. The patient appeared to be highly infectious, yet a single sputum smear and culture specimen obtained one month before his death showed no growth of mycobacteria. It is possible that this false-negative smear and culture resulted from analysis of a salivary specimen rather than from a sputum sample [21]. Mycobacteria may be released from the lung irregularly, resulting in cultures that may be negative one day and positive the next. For this reason, a minimum of three to five specimens should be obtained from patients suspected of having pulmonary tuberculosis [22]. Riley and colleagues [23] have shown that the infectivity of tuberculosis was related to the number of air changes per hour in the room. Air control is necessary in rooms of known or suspected transmitters and in places where persons with untreated tuberculosis might contaminate the air [24]. Infection of susceptible personnel in contact with patients with unrecognized active tuberculosis can be prevented by the correct use of mechanical ventilation and ultraviolet irradiation of the air [25]. Patients’ rooms should have fresh air introduced through a ventilatory system and preferably exhausted to the outside. United States Public Health Service standards require air disinfection in hospital areas housing immunocompromised patients [26], and call for a minimum ventilation rate of six air changes per hour in isolation rooms [25]. In areas where the nature of the work is exceptionally hazardous, such as autopsy rooms, 15 to 25 air changes per hour [27,28] have been recommended. However, even high-
TABLE III
Year 1979 1980
1981 1982 1983 1984 1985 1986+
TRANSMISSION
OF TUBERCULOSIS-KANTOR
ET AL
Annual Tuberculin Skin Test Conversion Rate in Hospital Personnel from 1979 to 1988”
Number Tested
Conversions Number
Percent
4 2 5 3 3 2 9 3
0.9 0.5 1.5 0.8 0.8 0.5 1.45 0.9
458 429 333 389 359 380 a20 333
Mean skin test conversion rate + January 1986 to June 1966. l
= 0.92
percent.
change-rate ventilating systems may not be adequate to remove “burst source” contaminants before they disperse. Local exhaust and primary barrier systems are preferred because they control the contaminants at their source and minimize exposure to personnel [29]. At present, there are no uniform infection control guidelines for handling suspected or confirmed cases of tuberculosis at autopsy [30]. Each institution has had to develop its own policies based upon the recommended guidelines in the use of air control and containment of infectious aerosols. The index case had not received isoniazid chemoprophylaxis. As a patient with known tuberculin reaction who had received systemic corticosteroids and immunosuppressive therapy for several weeks, he fits the special clinical situations considered to clearly benefit from preventive therapy with isoniazid [3 11. ACKNOWLEDGMENT We thank Lorna Waiters, M.P.H., for her technical assistance, and Patricia W. Wellman for secretarial assistance.
REFERENCES 1. 2.
3.
4. 5. 6.
Myers JA: Tuberculosis: a half century of study and conquest. St. Louis: Warren H. Green, 1970; 166-199. Ehrenkranz NJ, Kicklighter L: Tuberculosis outbreak in a general hospital: evidence for airborne spread. Ann Intern Med 1972; 77: 377-382. Craven RB, Wenzel RP, Atuk N: Minimizing tuberculosis risk to hospital personnel and students exposed to unsuspected disease. Ann Intern Med 1975; 82: 628-632. Catanzaro A: Nosocomial tuberculosis. Am Rev Respir Dis 1982; 125: 559-562. Hershfield ES: Tuberculosis in the world. Chest 1979; 765: 8055-8115. Edlin GP: Active tuberculosis unrecognized until necropsy. Lancet 1978; I: 650-652.
7.
8.
9. 10. 11.
Slavin RE, Walsh TJ, Pollack AD: Late generalized tuberculosis: a clinical and pathologic analysis and comparison of 100 cases in the preantibiotic and antibiotic eras. Medicine (Baltimore) 1980; 59: 352-366. Enarson DA, Grzybwsk S, Dorken E: Failure of diagnosis as a factor in tuberculosis mortality. Can Med Assoc J 1978; 118: 1520-1522. Proudfoot AT, Akhtar AJ, Douglas AC, Horne NW: Miliary tuberculosis in adults. Br Med J 1969; 2: 273-276. Reid DD: Incidence of tuberculosis among workers in medical laboratories. Br Med J 1957; 2: 10-14. Harrington JM, Shannon HS: Incidence of tuberculosis, hepatitis, brucellosis and shigellosis in British medical laboratory workers. Br Med J 1976; 1: 759-762.
May 1988 The American Journal of Medicine Volume 84
837
NOSOCOMIAL
12.
13. 14.
15.
16. 17.
18. 19.
20.
21.
22.
838
TRANSMISSION
OF TUBERCULOSIS-KANTOR
ET Al
Pike RM: Laboratory-associated infections: incidences, fatalities, causes, and prevention. Annu Rev Microbial 1979; 33: 41-66. American Thoracic Society: The tuberculin skin test. Am Rev Respir Dis 1981: 124: 356-363. American Thoracic Society: Guidelines for the investigation and management of tuberculosis contacts. Am Rev Respir Dis 1976; 114: 459-463. Des Prez RM, Goodwin RA: Mycobacterium tuberculosis. In: Mandell GL, Douglas RG, Bennett JE, eds. Principles and practices of infectious disease, 2nd ed. New York: John Wiley, 1985; 1383-1406. Treip C, Meyers D: Fatal tuberculosis in a general hospital. Lancet 1959; I: 164-167. Weir MR, Thornton GF: Extra pulmonary tuberculosis. Experience of a community hospital and review of the literature. Am J Med 1985; 79: 467-478. Babrowitz ID: Active tuberculosis undiagnosed until autopsy. Am J Med 1982; 72: 650-658. Makela V, Ala-Kulju K, Hulst J, Siljander T: Tuberculosisdiagnosed and undiagnosed-as a cause of death. Autopsy observations. Stand J Respir Dis 1971; 52: 13-18. Mackel DC, Forney JE: Overview of the epidemiology of laboratory-acquired infections. In: Miller BM, ed. Laboratory safety: principles and practices. Washington: American Society for Microbiology, 1986; 37-42. Isaac-Renton JL, Puselja BB, Allen EA, Grzybowski S, Black WA: Microscopic evaluation of sputum specimens submitted for Mycobacterium tuberculosis culture. Am J Clin Pathol 1985; 84: 361-363. Sommers HM, McClatchy JK: Cumitech 16, laboratory diagnosis of the mycobacterioses. Morello JA, coordinating ed. Washington: American Society for Microbiology,
May
1988
The
American
Journal
of Medicine
Volume
23.
24.
25.
26.
27.
28.
29.
30. 31.
84
1983. Riley RL, Mills CC, O’Grady F, et al: infectiousness of air from a tuberculosis ward. Ultraviolet irradiation of infected air: comparative infectiousness of different patients. Am Rev Respir Dis 1962; 85: 51 l-525. Centers for Disease Control: Guidelines for the prevention of TB transmission in hospitals (publication no. 82-8371). Atlanta: U.S. Department of Health and Human Services, Public Health Service, October 1982. Centers for Disease Control: Notes on air hygiene. Summary of conference on air disinfection. Arch Environ Health 197 1; 23: 473-474. Reiley RL: Principles of UV air disinfection. U.S. Department of Health, Education, and Welfare, Public Health Service publication no. 00-2215, January 1972. McCracken RC: Facility engineering, planning and construction letter (publication no. IL lOA4-82-23). Departments of Medicine and Surgery, Veterans Administration. Heating, Ventilation, Air Conditioning Criteria, November 1982. U.S. Department of Health, Education and Welfare: Minimum requirements of construction and equipment for hospitals and medical facilities (DHEW publication no. 79-14500). Washington: U.S. Government Printing Office, 1979. West DL, Chatigny MA: Design of microbiological and biomedical research facilities. In: Miller BM, ed. Laboratory safety: principles and practices. Washington: American Society for Microbiology, 1986; 124-137. Orenstein JM: Guidelines for high risk or potentially high risk autopsy cases. Pathologist 1984; 38: 33-34. National Consensus Conference on Tuberculosis. Preventive treatment of tuberculosis. Chest 1985; 87: 128S1325.