Mycobacterium tuberculosis Septic Shock

Background: Septic shock due to Mycobacterium tuberculosis (MTB) is an uncommon but well-recognized clinical syndrome. The objective of this study was to describe the unique clinical characteristics, epidemiologic risk factors, and covariates of survival of patients with MTB septic shock in comparison with other bacterial septic shock. Methods: A retrospective nested cohort study was conducted of patients given a diagnosis of MTB septic shock derived from a trinational, 8,670-patient database of patients with septic shock between 1996 and 2007. Results: In the database, 53 patients had been given a diagnosis of MTB shock compared with 5,419 with septic shock associated with isolation of more common bacterial pathogens. Patients with MTB and other bacterial septic shock had in-hospital mortality rates of 79.2% and 49.7%, respectively (P < .0001). Of the cases of MTB shock, all but five patients had recognized respiratory tract involvement. Fifty-five percent of patients (29 of 53) were documented (by direct culture or stain) as having disseminated extrapulmonary involvement. Inappropriate and appropriate initial empirical therapy was delivered in 28 patients (52.8%) and 25 patients (47.2%); survival was 7.1% and 36.0%, respectively (P = .0114). Ten patients (18.9%) did not receive anti-MTB therapy; all died. The median time to appropriate antimicrobial therapy for MTB septic shock was 31.0 h (interquartile range, 18.9-71.9 h). Only 11 patients received anti-MTB therapy within 24 h of documentation of hypotension; six of these (54.5%) survived. Only one of 21 patients (4.8%) who started anti-MTB therapy after 24 h survived (P = .0003 vs <24 h). Survival differences between these time intervals are not significantly different from those seen with bacterial septic shock due to more common bacterial pathogens. Conclusions: MTB septic shock behaves similarly to bacterial septic shock. As with bacterial septic shock, early appropriate antimicrobial therapy appears to improve mortality.