Risk factors for active tuberculosis in adults on highly active antiretroviral therapy in Africa

Abstract

In a recent limited article we reported that a past history of tuberculosis (TB) at HAART initiation was significantly associated with the risk of active TB in adults on HAART in Cote dIvoire. In another recent excellent study Lawn and colleagues did not find such an association in South African adults. This striking difference deserves further discussion. In our study 25% of patients had a past history of TB at HAART initiation and 50% of the previous TB episodes were diagnosed more than 3 years prior to enrolment. In the study by Lawn and colleagues only 14% of patients had a TB history at baseline and 50% of the previous TB episodes were diagnosed within the 13 months prior to enrolment. Thus we wonder if some previous episodes of TB could have been missed particularly among the remotest ones. In sub-Saharan Africa underreport of previous episodes is a well-known cause of TB recurrence underestimation. As TB recurrence may vary over time under-documenting the oldest pastepisodes compared with the most recent ones could introduce a major bias in the analyses looking at the association between TB history and TB occurrence during HAART. (excerpt)