© 2005 American Thoracic Society
Risk Factors for Tuberculosis among HIV-infected Patients Receiving Antiretroviral TreatmentFrom the Authors:Lawn and colleagues raise a number of crucial points. However, we disagree with them on most of these issues. They state that, in our study, the rate of tuberculosis (TB) in patients with a past history of TB (11.3/100 person-years) represents an "unusually high recurrence rate." In a recent report by Grant and coworkers (1), the incidence of TB in non–HAART-treated South African adults with a past history of TB was estimated at 19.3/100 person-years. In Grant and coworkers' study, as in ours, the "rate of TB in persons with a past history of TB" is not a "recurrence rate" but the rate of TB after the date of study enrollment, which was not the date of the end of the previous TB episode. In our patients with a history of TB, the mean time since the last TB episode was 2.8 yr. Including this time in the period at risk would obviously have led to an estimate of recurrence rate quite lower than the TB rate after enrollment. In the study by Grant and coworkers, the rate of TB in non–HAART-treated patients with no TB history was half that of patients with TB history (1). We observed the same phenomenon in patients on HAART. Lawn and colleagues report that they did not, although they do not provide sufficient data to allow us to comment on the strength and limits of their analysis. We would be interested in seeing their data on TB incidence in patients with and without TB history (even if not significantly different in their sample), and in the procedures used to document a TB history at study entry. Documenting smear- and culture-negative (SCN) TB requires standardized diagnosis criteria and systematic review by an event documentation committee. Both conditions were applied in our study. In HIV-infected patients, SCN TB has long been recognized as a frequent problem, even in industrialized countries (2, 3). In reports from sub-Saharan settings, the percentage of SCN episodes which are diagnosed as "possible TB" by experienced infectious diseases specialists is frequently high (4). Where low rates are reported, one expects that patients with SCN TB died before being diagnosed. SCN TB is frequently extrapulmonary and found in patients with low CD4 count. Therefore, SCN TB and culture-positive TB may be associated with different risk factors. Contrary to Lawn and colleagues, we don't believe that SCN TB should be excluded from TB risk factor studies. Finally, we agree with Lawn and colleagues on one point: studies performed in sub-Saharan Africa frequently include patients with low CD4 count. Some of them occasionally give negative conclusions on patients with high CD4 count, despite insufficient power (5). In our reduced sample of patients with low CD4 count, CD4 was not significantly associated with TB occurrence, but we carefully stated that this did not mean that a low baseline CD4 was not a risk factor for TB occurrence in the overall population of HIV-infected adults who begin HAART.
Programme PAC-CI, Abidjan, Ivory Coast FOOTNOTES Conflict of Interest Statement: None of the authors have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. REFERENCES
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
