NTCA - A Tuberculosis Outbreak in Homeless Shelter -- Illinois 2007-2011

This prolonged outbreak demonstrates the potential value of identification and treatment of latent tuberculosis cases early in an outbreak. 


Michael F. Arbisi BS(MT)MS1, Claire Dobbins, MS RN2 , Elaine Darnall, RN, BSN, CIC1, Julia Howland MPH CPH1, Kate Marishta, MPH2, Sara Boline, MPH2
1 Illinois Department of Public Health, 2 Kane County Department of Public Health

Background: Despite declining national rates, homeless shelters continue to be high-risk environments for tuberculosis (TB). Conditions in homeless shelters and characteristics of the homeless population make TB control and treatment difficult. From 2007-2011, an outbreak of TB with a matching genotype pattern occurred among residents of a homeless shelter in Illinois.

Methods: Outbreak associated cases were defined as shelter staff, residents or contacts who developed tuberculosis from November 2007 – present, with a genotype match based on PCR-based MIRU1 and spoligotyping analysis. Culture negative cases with an epidemiologic link to the shelter were also classified as outbreak-associated. Cases were identified through three mass screening events, active surveillance by shelter staff, collaboration with local providers and contact investigations.

Results: In November 2007, one shelter resident was found to have smear positive, cavitary disease. A contact investigation was done and no additional cases were located until November 2009. To date, 1184 potentially exposed staff and residents have been identified and 406 staff and residents have been screened. Three mass screening events in 2009 and 2010 used symptom screen, IGRA, chest x-rays and sputum smear and culture to test staff and residents. Local providers were informed of the outbreak and were asked to report suspected cases of tuberculosis to the local health department. Twenty-two outbreak-associated cases have been identified at the shelter. Nineteen cases have a PFGE pattern identical to the index case and two are clinical cases with a strong epidemiologic link.

Conclusion: Screening for active cases conducted in 2007, 2009 and 2010 did not interrupt transmission. Almost two-thirds of the exposed residents have not yet been tested. Shelter populations require multiple and varied opportunities for TB testing. In congregate settings, one case of active tuberculosis should prompt regular testing of staff and symptomatic residents even if an initial contact investigation does not yield additional cases. This prolonged outbreak demonstrates the potential value of identification and treatment of latent tuberculosis cases early in an outbreak. Developments in LTBI treatment may address barriers to treatment in the homeless.