Background: Tuberculosis (TB) is a public health concern in correctional fa
cilities. High turnover of inmate populations may preclude timely diagnosis
of TB, so that unrecognised transmission may be common. Objective: To dete
rmine the proportion of inmates with new skin test conversions who had iden
tifiable exposure to diagnosed cases of TB in the correctional system, and
to test the hypothesis that source cases of TB may be undiagnosed during in
carceration. Setting: Maryland Division of Corrections, USA. Subjects: All
inmates whose skin test converted from negative to positive at annual scree
ning. Design: All cases of TB in inmates, diagnosed in the prisons during t
he relevant time period, were identified. Movements of skin test converters
and potential source cases within the prisons were matched. We then matche
d all inmates discharged from the prison system with all new cases of tuber
culosis notified to the Maryland Department of Health & Mental Hygiene tube
rculosis registry in 1994. Results: The inmate turnover was 21% per year. P
robable exposure to a diagnosed source case was found in 13% of converters,
possible exposure in 10% and no exposure in 72%. In a further 5% exposure
status could not be determined. We identified four cases of pulmonary tuber
culosis notified to the state in 1994, within 3 months of discharge in rele
ased inmates, who were not known to have tuberculosis whilst incarcerated.
Conclusions: Significant transmission of TB due to undiagnosed index cases
may occur in prisons due to high population turnover. New skin test convers
ions in inmates should be treated as new infection, even in the absence of
identifiable exposure to TB.