Investigating Tb Clusters In High-Density Institutes And Health Care Settings

Background: Investigations for tuberculosis (TB) outbreak would be carried out once notifying. The working definition of suspicious TB cluster is defined as one confirmed TB case with subsequent reported TB case within 1 year. Methods: We analyzed suspicious TB clusters during 6 months of surveillance national-wide. The analysis focused on institutes with high density population and health-care settings. Results: Two out of 23 clusters (8.7%) are true TB outbreaks with molecular evidence. The first one is an outbreak with 2 nurses in a TB isolation ward. The second one is an outbreak with 2 intubated cases in a respiratory care ward. Another two are inconclusive because of no available strain for comparison but obvious epidemiologic link. All the other 19 are excluded from outbreak and seven (36.8%) of them were non-TB Mycobacteria (NTM) related. Among the 56 cases with smear positive, only 3 were positive for NATT for MTB complex. The positive predictive value for NATT (if culture is gold standard) is 67% (2/3) and the negative predictive value is 100% (31/31). The result of NATT from 5 specimens initially smear negative were all negative and culture result for Mycobacteria Tuberculosis (MTB) complex were negative, too. 34 (81%) of 42 intubated patients in respiratory care ward were smear positive. NATT had been performed in 22(52%) of the specimens, and only 2(9%) were positive, which were culture positive, too. Conclusion: Although there are increasing trend of NTM related smear positive clusters, 2 true outbreaks were noted during the 6 month survey. The result revealed that the working definition of cluster is useful to detect true outbreaks. Since early and restricted intervention, the outbreak was small and confined to 2 cases in each. For smear positive cases, the negative predictive value of NATT is excellent.