-The prevalence of latent tuberculosis infection (LTBI) among prison inmates is four times higher than the prevalence in the general population.
-The prevalence of LTBI among jail inmates is 17 times higher than the prevalence in the general population.
-More than 500,000 inmates with LTBI are released nationwide every year.
-The rate of tuberculosis (TB) infection in jails is 15 times that seen in the general population.
-One-third of those with active TB in this country have been recently incarcerated^1,2.

While most prisons and jails are vigilant when it comes to screening for TB infection, some correctional facilities are not attentive to LTBI treatment completion, thereby providing an ideal condition in which TB outbreaks may occur. When a TB outbreak occurs, public health officials should initiate an investigation of the circumstances related to the outbreak and try to interrupt further transmission of Mycobacterium tuberculosi (M. tuberculosis). This report of a recent TB outbreak in two Florida correctional facilities illustrates the complexities of TB control in congregate settings and highlights the need for further improvements in TB control measures in prisons and jails.

The TB Outbreak Setting
The local Health Department, in consultation with the Florida Bureau of TB and Refugee Health (FBTBRH), and the Florida Department of Corrections, investigated an outbreak of drug-susceptible TB that occurred among staff at two closely situated correctional facilities (facilities A and B) during the period April -September 2004.

Facility A has an inmate population of 875 and 343 work camp inmates. Correctional personnel working at the facility include 363 correctional staff, 26 medical personnel, and 13 contracted food workers. Facility B is located ten miles south of facility A. This facility incarcerates 1,361 inmates, and 216 correctional staff and 38 medical personnel work at the facility. Some of the correctional facility employees periodically rotate between the two facilities.

Approach to the Outbreak Investigation
The investigation of the outbreak was conducted in 2004 and standard TB case-based follow-up methods were used. The index patient was defined as the first patient to receive a diagnosis of TB, regardless of location of that case. The outbreak period was defined as the time period commencing when the index case reported TB-associated symptoms, and concluding one week after the last infectious patient was placed in respiratory isolation. The potentially infectious period for each patient was defined as commencing on the date of onset of symptoms consistent with TB or if no symptoms, 12 weeks from the date of diagnosis, and concluding when the patient was isolated from further contact with others. M. tuberculosis isolates from case patients were fingerprinted by IS-6110 restriction fragment length polymorphism (RFLP) to determine if cases were caused by the same strain. A case was included in the outbreak if there was a history of close contact, defined as shared cells or work area with another case linked to the outbreak, and/or if the subject's M. tuberculosis isolate had a matching RFLP fingerprint.ⁱ

All inmates and staff were questioned regarding TB risk factors and tuberculin skin tests (TSTs) were administered to all inmates and staff, excluding members of the staff who had a previously documented positive test or those who had a negative test within the past three months.ⁱⁱ (see TB 101, page 7.)

A positive TST was defined as induration greater or equal to 5 mm. Inmates who had positive TST results or symptoms suggestive of TB, regardless of TST results, were evaluated with a chest radiograph. (see IDCR-o-gram, page 6) The chest radiographs were conducted at the on-site medical units. Correctional staff members who had a positive TST or symptoms suggestive of TB were referred to the local DOH TB clinic for chest radiograph, medical evaluation, and treatment.

Three sputum samples were obtained from every inmate or staff member who demonstrated signs and symptoms suggestive of TB. Sputum smears were examined for acid-fast bacilli (AFB), were cultured for mycobacteria, and the Mycobacterium tuberculosis direct (MTD) test was performed.ⁱⁱⁱ Correctional staff members who had a positive TST greater than or equal to 5MM or symptoms suggestive of TB were sent to the local health department for evaluation. Inmates who were suspected of having TB were placed in negative airborne infection isolation (AII) rooms at the facility.

Results of the 2004 Outbreak Investigation
The 2004 outbreak investigators discovered that over a period of two and a half years (May 2001-September 2004), five cases of TB were reported among correctional staff members working at facilities A or B. Of these five cases, four cases of TB were identified among the correctional staff members at facility A and one case of TB was identified in a correctional staff member who worked at facility B. Cases 2 and 4 from facility A, and Case 3 from facility B were linked to the index case by RFLP fingerprinting of isolates and contact exposure history. Case 5 from facility A was epidemiologically linked to the other four cases, however the RFLD fingerprint did not match with the other cases.

The index case, identified as the source case, was an HIV-infected staff member initially diagnosed with extrapulmonary TB in May 2001. This patient was employed as a secretary in the medical unit in facility A and had frequent contact with coworkers and correctional officers who were involved in the transportation of inmates. A private physician managed TB treatment, TB medications were self-administered, and the patient was found to be non-adherent with the medications^iv. No contact investigation was performed when the patient was first diagnosed in 2001, since the patient was felt to have only extrpulmonary TB and therefore, was considered to be "non-infectious".^v However, in September 2002, despite the fact that the patient originally was noted to have had a cough, a sputum specimen was found to be positive for AFB on smear and M. tuberculosis in culture. At this time, pulmonary TB was diagnosed (see Table 1, page 3.) and a contact investigation in facility A was conducted. A review of the patient's medical records revealed that sputum specimens were not obtained prior to September 2002.

Case 2, an HIV-uninfected correctional transportation officer employed at facility A, was identified during the September 2002 contact investigation. This individual had a previously documented positive TST and a previous history of TB that could no be verified. Symptoms suggestive of TB were not recorded in 2002, but a private physician who was providing treatment ordered a chest radiograph.^vi The chest radiograph demonstrated an infiltrate and sputum specimens were positive for AFB on smear and M. tuberculosis in culture. Additional investigations were not conducted, as the contact investigation around the index case was deemed sufficient to capture this case's workplace contacts.

Case 3, an HIV-uninfected correctional transportation officer employed at facility B, was diagnosed with pulmonary TB in October 2002. This individual was responsible for transporting inmates to and from the medical units of facilities A and B. Case 3's sputum specimen was positive for both AFB on smear and M. tuberculosis in culture.

Case 4, an HIV-uninfected correctional transportation officer employed at facility A, was diagnosed with TB in March 2004. A private physician monitoring this subject for a history of asthma ordered a chest radiograph in 2004 for reasons unrelated to the TB outbreak. The chest radiograph revealed a 1 cm nodule in the right upper lobe. Tissue culture obtained following excision of the upper lobe nodule was positive for M. tuberculosis. Case 4 was initially identified as a contact to the index case but did not follow through with medical evaluation and treatment. Additionally, this subject was a TST converter; his baseline TST measured 0 mm in 2002 and 10 mm in August 2003.^vii The RFLP pattern confirmed the link between this case, the index case, and cases 2 and 3.^viii

Case 5, an HIV-uninfected correctional officer employed at facility A was diagnosed with pulmonary TB in April 2004. This case's RFLP fingerprint did not match the RFLP fingerprint obtained for the index case or cases 2, 3, or 4, but an epidemiological link was identified as this case had close social contact with Case 4.^ix

Records on skin tests and/or TST records, results, and chest radiographs were not available for the majority of correctional staff at both facilities.

The clinical characteristics of the index patient and the secondary cases (Cases 2, 3, 4 and 5) are listed in (see Table 1, page 3.)

Outbreak Investigation
During the 2004 outbreak investigation, all movement in and out of both facilities, including visitation, was halted for one week, to allow initial screening of all inmates and correctional officers. During that one week 78 staff received TSTs and all staff and inmates were screened using a symptom screen and risk assessment.^x Of the staff screened/tested, 54 of these individuals were referred to the local health department TB clinic for evaluation of symptoms suggestive of TB or for follow-up evaluation of a positive TST. TSTs were performed on all inmates with a history of a negative TST and chest radiographs were obtained for 30 inmates to further evaluate symptoms suggestive of TB or because they had a positive TST. Sputum specimens were obtained from 18 inmates who had abnormal chest radiographs and/or symptoms suggestive of TB.

Follow up
As a result of this outbreak, the state and local health departments recommended implementation of an electronic database for tracking serial TB screenings so that TST conversions and appropriate recommendation of follow-up medical care will be documented. This is the recommended practice in all high risk settings, such as medical and correctional facilities³. Quarterly skin testing of inmates and correctional staff at these facilities will continue until no further conversions are identified. At the time of this publication, no further active cases of TB disease have been identified.

Discussion and Recommendations for Florida Outbreak
Several lines of evidence suggest that M. tuberculosis was transmitted from the index patient to correctional staff at facilities A and B. The index case was a medical staff member employed in the health care unit at facility A, who had been diagnosed with extrapulmonary TB in 2001. This patient was not adherent with self-administered treatment and worked with several correctional transportation officers. Since an evaluation for pulmonary TB was not performed in May 2001 (the time of initial diagnosis), it is not clear if this individual was infectious at this time. This individual developed symptoms of pulmonary TB in 2002, and may have infected other correctional staff from January 2001 to June 2002.

The long duration of exposure to the source (index) case is also attributable to a failure to monitor the patient's adherence with anti-TB treatment. Noncompliance with TB treatment is a well-known problem⁴. All TB medications should be administered by directly observed therapy and adherence documented, as this is the current standard of care3. Public health consultants and staff are available to assist with TB case evaluation and adherence with anti-TB medication.

Furthermore, since records on TSTs and/or chest radiographs were not available for the majority of the correctional staff, it was difficult to ascertain correctional staff compliance with annual TSTs. TST conversions could not clearly be related to the period of exposure. Obtaining baseline TSTs in higher risk settings, such as medical and correctional facilities, is generally recommended,ⁱⁱⁱ as it provides a point of reference for measurement of TST conversion rates.

Although mandatory screening and testing of all employees had been implemented three years prior to this outbreak, several correctional staff members did not comply. Correctional supervisors must be able to identify employees who are not compliant to such policies and implement appropriate education and corrective actions if these policies are not adhered to. Measures need to be supported (with collaboration to help design, implement, and monitor) by public health programs to address conditions, which foster the transmission of M. tuberculosis, identify corrective measures, and implement and monitor all follow-up.

Employers are unlikely to be aware of the HIV status of their employees. Educational programs for both inmates and staff are essential for the protection of all persons at the facility. All staff (and therefore the potentially unknown HIV-infected or immunosuppressed staff) should be repeatedly advised of the increased risk with regard to TB in those with HIV infection, the importance of TB skin testing and completion of LTBI treatment, and should be taught to recognize symptoms of TB. Practitioners treating correctional personnel and inmates should have a high index of suspicion for TB and obtain appropriate smears and cultures whenever possible.

General Discussion and Recommendations
There is a dangerous synergy between HIV and TB. Prisoners, who have long been known to have disproportionately high rates of TB disease and TB infection,^v also have more than five times the general population's rate of AIDS, and between four and 10 times the general population's rate of HIV infection⁶. The single most potent factor affecting the risk of progression from LTBI to TB disease is HIV coinfection⁷. It is often more difficult to detect TB in persons with HIV/AIDS because they may not respond to the TST, and may present with atypical or negative findings on chest radiographs⁸. It is therefore recommended that HIV-infected patients with respiratory symptoms undergo a sputum analysis in addition to a chest radiograph. Furthermore, significant drug interactions may complicate the concurrent treatment of HIV and TB⁹. In short, HIV increases the risk of progression from TB infection to disease, makes screening for TB more difficult, and complicates the treatment of TB.

As recommended by the National Commission on Correctional Health Care (NCCHC), all inmates should receive TB symptom screening on intake; anyone with TB symptoms (chronic productive cough, fever, weight loss, night sweats) should immediately be moved to a negative pressure respiratory isolation room and evaluated for TB disease. TSTs should be administered to all inmates and correctional staff members who have not had a previous documented positive TST result. (see IDCR-o-gram, page 6.) Any patient whose TST indicates TB infection should receive a chest radiograph. TST may fail to identify TB infection in high-risk patients including inmates¹⁰ - TB control officers may consider using on-site chest radiography to screen all inmates at entry.

Inmates and correctional staff who have documented LTBI should complete a course of treatment and adherence should be monitored³. Those with a positive TST who cannot complete treatment for LTBI should receive regular screening for TB symptoms; any such patient with significant immune compromising factors should be scrutinized for TB symptoms even more frequently. Inmates who do not present with TB upon initial intake screening should be evaluated annually for TST conversion, and more frequently if there is evidence of recent transmission of M. tuberculosis in the facility, or if inmates with HIV are housed together¹¹.

As an airborne infection, TB presents one of the most prescient threats not only to inmates, but also to correctional staff, health care providers, visitors, and others who come in close contact with TB patients. This case illustrates that point - one more time.

Disclosures:
*Nothing to dislcose.

Notes:
i. RFLP analysis is a DNA fingerprinting method that allows public health officials to distinguish the transmission of specific strains of tuberculosis during an outbreak. This method is based on the detection of the copy number and location of the mobile genetic element IS6110 in the M. tuberculosis genome. Enzymatic digestion of the DNA produces fragments, which can be separated by electrophoresis. The fragments are immobilized on a nylon membrane, and a specific chemoluminescence-labeled DNA probe is used to reveal the pattern bands on x-ray film.
ii. Best practice would be to test every individual who had a negative PPD test at the time of the outbreak investigation. However, this practice can be difficult to enforce with correctional staff. Additional education regarding TB exposure, testing and management could improve TST uptake among correctional personnel.
iii. MTD tests are isothermal transcription-mediated amplification assays used in the rapid identification of M. tuberculosis in respiratory samples. These tests produce results within two to seven hours after sputum processing.
iv. Best practice would be to have this patient take TB medications under directly observed therapy, especially given the patient's history of immunosuppression.
v. HIV-infected subjects who have extrapulmonary tuberculosis can have undetected pulmonary infection. In general, no patient with suspected tuberculosis can be considered non-infectious until sputum smears are also found to be negative; in the case of HIV-infected patients, culture for TB should also be performed, as sputum smears may be negative, despite the presence of active pulmonary tuberculosis.
vi. Any individual who has a positive TST, with or without the presence of symptoms suggestive of TB, should have a baseline chest radiograph.
vii. Best practice would have been to perform a chest radiograph at the time of TST conversion.
viii. Active tuberculosis should be considered whenever any individual who works in a higher risk setting, such as a medical facility or correctional facility, presents with a pulmonary infiltrate. TST should be performed and sputum specimens should be obtained for AFB smear and culture.
ix. There can be "background" cases even in the face of an outbreak. These types of cases could possibly be the next source case of a future outbreak. Vigilance for anyone with symptoms suggestive of TB need to be identified and evaluated early.
x. Best practice is to perform risk assesment, symptom screening and TSTs on all staff. However, staff have the right to refuse testing on-site and may receive TSTs from private physicians.

References:
1. National Commission on Correctional Health Care. "The Health Status Of Soon-To-Be-Released Inmates". A Report to Congress. 2002. Full document accessible at: http://www.ncchc.org/pubs/pubs_stbr.vol1.html
2. Health Standards of the American Correctional Association. "Standards for Adult Correctional Institutions, 4th Edition". 2003.
3. CDC. MMWR 2000; 49(No. RR-6): 1-39.
4. Sbarbaro, John, Burman William, Cornelis Rietmeijer, et al. Chest 1997; 111(5): 1151-3.
5. CDC. MMWR 1996; 45(No. RR-8): 1-27.
6. Hammett, TM; Rhodes, W; Harmon P. National HIV Prevention Conference 1999. Atlanta, GA. Abstract 571.
7. CDC. MMWR 2000; 49(46): 1041-4.
8. Burman WJ, Gallicano K, Peloquin C. Clin Infect Dis 1999 28(3): 419-29.
9. Spradling P, McLaughlin S, Drociuk D, et al. XIII International AIDS Conference 2000. Durban, South Africa. Abstract ThPeB5188.
10. Zoloth SR, Safyer S, Rosen J, et al. Am J Public Health 1993; 83(5): 749-51.
11. Perlman DC, el-Sadr WM, Nelson ET, et al. Clin Infect Dis 1997; 25: 242-6.

Back to Top