Mycobacterium tuberculosis nucleic acid amplification tests reduce nosocomial tuberculosis exposure in intensive care...

This retrospective national surveillance study investigated the burden of and risk factors for nosocomial exposure of pulmonary tuberculosis (TB) in intensive care units.

Patients admitted to intensive care units were identified from the National Health Insurance Research Database. During 2004-2009, there were 1 387 707 intensive care unit admissions of 900 562 adult patients. Pulmonary tuberculosis association was considered if the patient was diagnosed with pulmonary tuberculosis during admission or within 3 months after discharge. Nosocomial transmissible period was calculated based on the length of anti-tuberculosis treatment and negative-pressure isolation during admission.

Pulmonary tuberculosis was associated with 1.20% of all intensive care unit admissions and 6731 (38.9%) started anti-TB treatment during admission. For the other 10 583 admissions, the diagnosis was made after discharge and anti-TB treatment was not prescribed during admission. The probability paralleled the regional tuberculosis incidence. On average, 2794 pulmonary tuberculosis associated intensive care unit admissions contributed to 42 999-44 062 days of nosocomial exposure per year. The length of nosocomial transmissible period decreased with the gradual implementation of Mycobacterium tuberculosis nucleic acid amplification tests in intensive care practice. Multivariate linear regression analysis revealed that the length of nosocomial transmissible period was inversely associated with male gender, airway symptoms prior to admission and performing M. tuberculosis nucleic acid amplification tests and mycobacterial culture.

Nosocomial tuberculosis exposure is not uncommon in intensive care units. Performing rapid molecular diagnostic tests in those suspected of tuberculosis is recommended to reduce the risk of nosocomial exposure.