Blood culture for suspected disseminated tuberculosis.

Sir, Although blood culture was included in the 'work up' of the patient who presented with a month's history of low-grade fever and slowly progressive disturbed consciousness, 1 there was no mention of whether or not some of the blood cultures were specifically for the purpose of identifying bacteraemic Mycobacterium tuberculosis, for which human im-munodeficiency virus (HIV) infection is a risk factor requiring specific testing as well if disseminated tuberculosis is in the differential diagnosis. Mycobacterium tuberculosis was eventually detected by polymerase chain reaction analysis and culture of induced sputum and stomach fluid, 1 but the identification of M. tuberculosis by blood culture would have generated a more plausible explanation for the miliary dissemination detected by imaging. Accordingly, in the patient in whom fever of unknown origin and neurological compromise might be attributable to disseminated tuberculosis and in whom both initial imaging and cerebrospinal fluid analysis yield falsely negative results, 1 much might be subsequently gained from an initial work up that includes both serological testing for HIV infection and blood cultures for M. tuberculosis; the latter, in one study, tested positive in 15 out of 39 patients. 2 In a scenario such as the one in the clinical vignette, a positive blood culture generates a more direct explanation for disseminated tuberculosis than would be the case with culture-positive gastric juice. et al. Bacteremic disseminated tuberculosis in sub-Saharan Africa: a prospective cohort study.