Tuberculosis in asymptomatic HIV-infected adults with abnormal chest radiographs screened for tuberculosis prevention.

BACKGROUND: Isoniazid preventive therapy (IPT) prevents tuberculosis (TB) in people living with HIV (human immunodeficiency virus, PLWH). Symptom screening without chest radiographs (CXRs) was established as the strategy for excluding TB disease among PLWH seeking IPT in Botswana's 2001 pilot project. This strategy was evaluated in 2004-2006 among candidates screened for an IPT clinical trial. METHODS: PLWH referred from clinics and HIV testing centers were screened for TB symptoms. All asymptomatic candidates received CXRs; those with abnormal CXRs were investigated further. RESULTS: Among 2732 asymptomatic candidates screened, 302 (11%) had abnormal CXRs potentially compatible with TB; TB disease was diagnosed in 43 of these 302 (14%), or 43 (1.6%) of the 2732 asymptomatic candidates. While not associated with CD4 lymphocyte counts < 200 cells/mm(3), TB was associated with a positive tuberculin skin test (relative risk 2.1, 95%CI 1.1-4.0). IPT was initiated in 113 (62%) of 182 asymptomatic PLWH with abnormal CXRs; 8/113 (7%) subsequently developed TB, and 7/8 (88%) successfully completed anti-tuberculosis treatment. CONCLUSIONS: The prevalences of abnormal CXRs and TB were respectively 2.6- and 8.9-fold higher among asymptomatic PLWH screened for the trial than in the pilot. A cost-effectiveness analysis is needed to determine whether the benefits of symptom screening alone are offset by the risk of inducing INH resistance by excluding CXRs during screening.

Effects of alcohol reduction on blood pressure: a meta-analysis of randomized controlled trials.

Alcohol drinking has been associated with increased blood pressure in epidemiological studies. We conducted a meta-analysis of randomized controlled trials to assess the effects of alcohol reduction on blood pressure. We included 15 randomized control trials (total of 2234 participants) published before June 1999 in which alcohol reduction was the only intervention difference between active and control treatment groups. Using a standard protocol, information on sample size, participant characteristics, study design, intervention methods, duration, and treatment results was abstracted independently by 3 investigators. By means of a fixed-effects model, findings from individual trials were pooled after results for each trial were weighted by the inverse of its variance. Overall, alcohol reduction was associated with a significant reduction in mean (95% confidence interval) systolic and diastolic blood pressures of -3.31 mm Hg (-2.52 to -4.10 mm Hg) and -2.04 mm Hg (-1.49 to -2.58 mm Hg), respectively. A dose-response relationship was observed between mean percentage of alcohol reduction and mean blood pressure reduction. Effects of intervention were enhanced in those with higher baseline blood pressure. Our study suggests that alcohol reduction should be recommended as an important component of lifestyle modification for the prevention and treatment of hypertension among heavy drinkers.