We read with interest the article by Pechère and colleagues1 on Mycobacterium genavense infection in patients with human immunodeficiency virus infection. Based on a comparison of survival with CD4 cell count—matched controls, the authors concluded that M genavense infection lessens survival. Clearly, there are biological reasons to believe that this disease could lessen survival. However, we suggest caution in attributing causality to M genavense infection and its treatment in the interpretation of the statistical relationships of these observational data.
An alternative explanation for the association of M genavense infection with decreased survival is that the mycobacterial infection is a marker of more severe immunosuppression that is not accounted for by CD4 cell count alone. Although controls were matched on CD4 cell count, this measure may be an incomplete marker of degree of immunosuppression in Pechère and colleagues' study as it is in other settings.2,3 Data from the Multicenter