During 1995-1996, staff from the California departments of corrections and health services and local health departments investigated two outbreaks of drug-susceptible tuberculosis (TB). The outbreaks occurred in two state correctional institutions with dedicated HIV housing units. In each outbreak, all cases were linked by IS6110-based DNA fingerprinting of Mycobacterium tuberculosis isolates. This report describes the investigations of both outbreaks; the findings indicated that M. tuberculosis can spread rapidly among HIV-infected inmates and be transmitted to their visitors and prison employees, with secondary spread to the community.
In both of the investigations, a positive tuberculin skin test (TST) was defined as an induration of greater than or equal to 5 mm in contacts and/or HIV-infected persons. A TST conversion in a contact was defined as an increase of greater than or equal to 5 mm from a documented negative to a positive TST within the previous 2 years. Only culture-positive pulmonary cases were considered infectious, and the infectious period was considered to begin 6 weeks before the date the culture-positive specimen was obtained (if the patient was asymptomatic) or the date of onset of symptoms consistent with TB.
On entry to the 500-person prison HIV housing unit in May 1995, the index case-patient was asymptomatic and anergic with a negative TST, and had a CD4 count of 6 cells/uL, and a 1-cm calcified nodule on chest radiograph. Three sputum specimens, routinely collected on entry of all inmates into the housing unit, were smear- and culture-negative. Isoniazid (INH) was not prescribed because of baseline liver function test abnormalities. During the next 3 months he was treated with several courses of antibiotics, initially for laboratory-confirmed Pneumocystis carinii pneumonia (PCP) and then for episodic fever and cough. Each time his symptoms decreased, and one chest radiograph showed a new infiltrate that resolved with antibiotic treatment. In late August 1995, a chest radiograph revealed a new infiltrate, and sputum specimens were smear-positive for acid-fast bacillus (AFB). The patient was isolated and started on multidrug therapy for TB.
During September 1995-April 1996, drug-susceptible TB was diagnosed in 14 other inmates (including three parolees) and the HIV-infected wife of the index case-patient. Their M. tuberculosis isolates matched the isolate from the index case-patient by DNA fingerprint analysis. All inmates with TB resided on the same wing when one or more persons with TB with the outbreak strain had infectious cases. Of the 312 inmates who resided at least 1 day on the same wing as case-patients, 185 were available for screening in December; three had TST conversions but no disease. Inmates with TB disease were isolated and treated, and the proportion of the approximately 150 contacts in the wing receiving directly observed INH preventive therapy was increased from 14% in October 1995 to 60% in January 1996.
In January 1995, the index case-patient had a positive TST and received 6 months of preventive therapy while in a state prison. In December 1995, he was sent from the prison to a community hospital with cough, fever, a chest radiograph with infiltrate on the right, AFB smear-negative sputum specimens, and a newly diagnosed immuno-deficiency (i.e., low CD4 count). He was empirically treated for PCP with trimethoprim/sulfamethoxazole, but his fever persisted. After the addition of prednisone, his fever resolved. On January 6, 1996, he was transferred from the hospital into an 180-person HIV housing unit in a different prison (prison B). The community hospital staff indicated that no respiratory isolation was necessary. A chest radiograph on January 11, was normal. By January 19, cultures from sputum specimens and bone marrow aspirate obtained while he was at the community hospital (December 23, 1995) grew M. tuberculosis; he was placed in respiratory isolation, had a chest radiograph with a diffuse infiltrates bilaterally, and was started on multidrug therapy for TB. He died from miliary TB on January 20. None of his sputum specimens obtained on January 8, January 10, and January 11 were AFB smear-positive.
During January-August 1996, drug-susceptible TB was diagnosed in 15 other inmates (including six parolees). The DNA fingerprints of M. tuberculosis isolates from all 15 matched the fingerprint of the isolate of the index case-patient. Analysis of sputum specimens from all 140 inmate contacts in the facility at the time of the investigation identified seven secondary case-patients whose chest radiographs were normal at the time of screening; five were asymptomatic. Screening of inmate contacts also detected 25 (18%) asymptomatic TST convertors who did not have TB disease. These 25 received preventive therapy.
In both prisons, during the 4-month intervals between identification of the index case-patients and chest radiograph screening of all the contacts remaining in the housing unit, 190 inmates had been released. Of 56 (29%) who were reincarcerated in prisons or jails before they had had health evaluations in the community, follow-up information was available for the eight who were reincarcerated in jails; none had TB disease, and six accepted preventive therapy. The remaining 134 were referred to 22 local health jurisdictions. Of these 134, 76 (57%) were assessed; nine (12%) had culture-positive TB (three from prison A and six from prison B), each with the same outbreak strain of M. tuberculosis as found in the originating prison.
Secondary transmission may have occurred from both prison outbreaks to the community. The HIV-infected wife of the index case-patient in Prison A visited her husband for 4 hours per day on the 3 days before his placement in AFB isolation. Two months later, she developed smear- and culture-positive pulmonary TB with the outbreak strain. Her daughter, whose TST result was 0 mm on school entry in 1994, had a 28-mm reaction; she had not visited her father during his infectious period. An adult and two children aged less than 5 years who lived with a parolee from prison B while he was symptomatic all had TST results greater than 10 mm but had no prior baseline.
Among prison employees who had contact with case-patients, TST conversions occurred in nine (2.8%) of 319 in prison A and 11 (4.9%) of 223 in prison B. All 20 had had two documented negative TSTs during the previous 2 years, 19 had a baseline TST result of 0 mm, and 18 had a positive TST result of greater than 10 mm. No employees had TB attributable to either outbreak strain.
Reported by: T Prendergast, MD, B Hwang, MD, R Alexander, San Bernardino County Health Dept, San Bernardino; T Charron, MD, E Lopez, MD, Solano County Health Dept, Vallejo; J Culton, MD, J Bick, MD, M Shalaby, MD, D Dewsnup, DO, H Meyer, MD, E Horowitz, MD, N Khoury, MD, California Dept of Corrections; J Mohle-Boetani, MD, S Royce, MD, D Chin, MD, S Petrillo, V Miguelino, E Desmond, PhD, R Harrison, MD, J Cone, MD, C Greene, M Joseph, S Waterman, MD, State Epidemiologist, California Dept of Health Svcs. Div of Tuberculosis Elimination, National Center for HIV, STD, and TB Prevention, CDC.