1 table omitted
Tuberculosis (TB) is a substantial health concern in correctional facilities;
inmates and employees are at high risk, and TB outbreaks can lead to transmission
in surrounding communities.1- 3 The
Advisory Council for the Elimination of Tuberculosis (ACET) recommends that
all correctional facilities have a written TB infection–control plan
(TBICP).4 In September 2002, after diagnosis
of smear-positive pulmonary TB in a prison inmate, the Kansas TB Control Program,
with assistance from CDC, initiated a 6-month contact investigation. This
report summarizes the results of that investigation, which determined that,
while symptomatic for TB, the inmate had resided in three different jails
and a state prison, placing hundreds of employees and other inmates at risk
for TB infection. The circumstances of this case underscore the need for effective
TBICPs to be implemented by trained employees in jails and prisons and for
establishment of mechanisms to facilitate information-sharing between correctional
facilities and local and state health departments.
In October 2001, a U.S.-born man aged 36 years who was living temporarily
in a California homeless shelter had a productive cough with hemoptysis. In
December 2001, a physician at the shelter examined the man, suspected TB or
neoplasia, and recommended a chest radiograph; however, the man did not follow
In January 2002, the man returned to his residence in Kansas. Shortly
after arrival, he turned himself in to police on an outstanding arrest warrant.
He was held in jail A for 3 days before being transferred to jail B. While
in jail B, he received a medical evaluation, and bronchial asthma was diagnosed.
A tuberculin skin test (TST) was not administered, nor was a chest radiograph
obtained. After 7 weeks in jail B, the man was released in March 2002.
In June 2002, the man was convicted of a crime and again placed in jail
A for 3 days. He was then transferred to jail C, a large overflow facility,
pending sentencing. During the 6 weeks the man was in jail C, 125 transferred
inmates passed through the facility.
In August 2002, after being held for a total of 14 weeks in three jails,
none of which had a TBICP or had provided TB screening for inmates or employees,
the man was sentenced and transported to a Kansas state prison. During routine
processing of entering inmates, he answered affirmatively to six of seven
questions regarding TB symptoms. The state prison had a TBICP in place, and
a medical evaluation was indicated on the basis of the man’s answers;
however, he was not referred for medical evaluation. Following the prison’s
TBICP procedure for entering prisoners who are to be serially TB screened,
medical staff performed a two-step TST, which was read as 0 mm induration
on both occasions. The new inmate was then placed among the general prison
Medical staff at the state prison did not see the inmate again until
4 weeks later, when he was scheduled to receive chronic care for asthma. At
this medical examination, he received a chest radiograph that showed a cavitary
lesion of the right lower lobe. Despite having TB symptoms, he was placed
back with the general prison population and scheduled for a computerized tomography
(CT) scan 2 weeks later to rule out neoplasia. After the CT scan indicated
cavitary lesions consistent with TB, the man, now the TB index patient, was
placed in airborne infection isolation (AII), and sputum samples were collected.
The AII room was newly constructed and in working condition, according to
maintenance and monitoring documentation. However, because the recommended
N95 respirators5 were not available, prison
health staff used surgical masks when in the AII room with the index patient.
The first laboratory result from the index patient of 4+ smear-positive Mycobacterium tuberculosis was reported in late September,
6 weeks after he had arrived at the prison facility.
A contact investigation conducted in Kansas and Missouri identified
318 of an estimated 800 possible contacts of the index patient during the
infectious period, defined as the time from symptom onset to diagnosis, October
2001–September 2002. Of these contacts, two (0.6%) received a diagnosis
of TB disease. These two patients had been cellmates of the index patient,
one in jail A and the other in jail C. Tests of samples from these patients
and the index patient determined they had M. tuberculosis isolates with a matching 10-band restriction fragment-length polymorphism
pattern. The three isolates also had matching spacer oligonucleotide typing
and mycobacterial interspersed repetitive unit patterns.
Of 318 contacts identified, 256 were tested, and 47 (19.1%) of those
received diagnoses of latent TB infection (LTBI); 60 contacts could not be
located or refused follow-up. Two (4.1%) had a previously documented positive
TST. Sixty (23.4%) contacts had a previously documented negative TST, and
six (10.0%) of these had a positive TST during investigation screening. Among
196 contacts with no previously documented TST, 41 (20.9%) had a positive
TST during the investigation screening. The majority of infections among jail
and prison employees occurred in jail B (TST reaction rate: eight of 36 [22.2%])
and jail C (TST reaction rate: five of 32 [15.6%]), compared with jail A (TST
reaction rate: one of 14 [7.1%]) and the state prison (TST reaction rate:
one of 58 [1.7%]). All three jails had an open-cell design with multiple inmates
per cell; the state prison had single-occupancy cells with solid walls and
P Griffin, D McClenahan, J VandeVelde, Kansas Tuberculosis Control Program;
G Pezzino, MD, Kansas Dept of Health and Environment. R Funk, DVM, M Kitt,
MD, EIS officers, CDC.
During 1992-2002, the TB rate in Kansas increased from 2.2 per 100,000
population to 3.3, the largest increase among all 50 states and the District
of Columbia; in the majority of states, the TB rate declined.6 Although
the contribution of correctional facilities to the TB burden in Kansas is
unknown, a study in Tennessee reported that 43% of persons identified with
TB in the city of Memphis had previous contact with a single urban jail and
no other identified common exposure.3
As a result of the investigation findings described in this report,
the Kansas TB Control Program worked with prisons, jails, and local health
departments to provide guidance for developing or improving TBICPs and providing
TB education and baseline TSTs for all correctional employees. This guidance
has improved communication among all agencies to coordinate the return to
the community of inmates receiving TB medications.
Outbreak investigators were limited in their ability to determine the
extent of TB transmission directly attributable to the index patient because
of lack of previously documented TST results and the large number of contacts
who could not be located or refused follow-up.
Compared with the general population, inmates have higher TB prevalence,
associated with their higher prevalence of human immunodeficiency virus, increased
illicit substance use, and lower socioeconomic status [SES]. The risk for
TB is known to increase with lower SES, with crowded living conditions having
the greatest impact.7 Overcrowding enhances
the likelihood of infectious droplet nuclei transmission and has been correlated
with TST conversion in the Maryland state correctional system.8 Cell
design and overcrowding might have been factors in TB transmission in these
three Kansas jails. However, the impact of overcrowding and ventilation could
not be assessed directly in this investigation because facility surveys were
Early identification and treatment of persons with TB disease remains
the most effective means of preventing disease transmission. With the assistance
of state and local health departments, correctional facilities should develop
formal TBICPs (Box). Health departments
should provide assistance to correctional facilities in developing TBICPs
and conducting contact investigations, thereby controlling transmission within
facilities and the surrounding communities. Employee education and continuous
monitoring and evaluation of these policies should be part of every TBICP.
In addition, correctional facilities should maintain a tracking system for
inmate TB screening and treatment and establish a mechanism for sharing this
information with local and state health departments and other correctional
The findings in this report are based in part on contributions from
county health departments, prison health services, Kansas Dept of Corrections.
Missouri Tuberculosis Control Program.
Tuberculosis Transmission in Multiple Correctional Facilities—Kansas, 2002-2003. JAMA. 2004;292(13):1543-1545. doi:10.1001/jama.292.13.1543