Tuberculosis (TB) remains a major global health challenge, with 9.0 million incident cases and 1.5 million deaths in the world in 2013. However, the current trend of TB incidence revealed a slow decline of 2% per year despite the implementation of directly observed treatment, short-course (DOTS) strategy. Thus beyond active case detection and treatment, methods to interrupt disease transmission is also important. We hypothesis that healthcare environment a possible site for TB transmission. First, delay in TB diagnosis is commonly observed worldwide, range from 44-51 days among studies in Taiwan, North America and Australia. Our prior project (NSC 101-2314-002-124) revealed increased risk of active TB with increased frequency of out-patient department (OPD) visiting, even after adjusting underlying diseases. Thus we aim to conduct a molecular epidemiology study via a retrospective cohort of TB exposed patients in OPD. The index cases were the patients who diagnosed as pulmonary TB at a tertiary teaching hospital with culture available. The exposed cohort were patients who attend the same clinic with the index case, 3 month before TB diagnosis. The molecular typing from the index cases and subsequent TB cases among the exposed cohort would be analyzed by mycobacterial interspersed repetitive unit -variable number tandem repeats (MIRU-VNTR) and spacer oligonucleotide genotyping. If cluster in the OPD setting is identified, we will further analysis the risk factors for acquiring active TB, including the host factors (e.g. age, underlying diseases, concurrent medication, et al) and environmental factors (e.g. out-patient department categories and ventilation system)