Impact of a New Molecular Tuberculosis (TB) Test on TB/HIV Outcomes Among HIV-

HIV and tuberculosis (TB) are, along with malaria, the leading infectious causes of death worldwide and in sub-Saharan Africa.1 In 2008, there were 2 million deaths from AIDS and 1.8 million deaths from TB worldwide.2, 3 The HIV epidemic has fueled an increase in the incidence of, prevalence of and mortality due to TB in the past 3 decades. 15% of TB cases are HIV co-infected worldwide.34 78% of HIV-infected TB cases are in Africa.3 TB is the leading cause of death and opportunistic infection among persons living with HIV/AIDS (PLWHA). The HIV epidemic has challenged time-tested TB control methods that are now failing in high HIV prevalence settings.5 The World Health Organization (WHO) has recommended that the Three I's - intensified case-finding for TB (ICF), isoniazid preventive therapy (IPT) and infection control for TB - be targeted at PLWHA.6 ICF and IPT have not been adequately implemented in part due to the absence of sensitive, specific and rapid TB tests. The Cepheid GeneXpert System, a new diagnostic test for TB, is a self-contained sputum-processing and real-time PCR system to detect the M. tuberculosis complex as well as rifampin resistance.78 The GeneXpert is rapid, highly sensitive and specific, can be used as a point-of-care test, and has low human resource, laboratory and infection control requirements.7, 8 The WHO Strategic and Advisory Group for TB has endorsed recommendations for widespread use of the GeneXpert.9 WHO strongly recommended that the GeneXpert should be used as the initial diagnostic test in HIV-infected TB suspects and multidrug-resistant TB (MDR-TB) suspects.9 WHO also recommended that implementation of the GeneXpert be phased in within the context of comprehensive national and MDR-TB strategic plans.9 WHO recognized that several operational conditions need to be met for successful implementation, including but not limited to stable electrical supply, security against theft, trained personnel and annual calibration of the instrument by a commercial supplier.9 WHO also noted that it is important to document the impact and cost-effectiveness of the GeneXpert for TB case detection.9

Trial Concept The overall objective of this proposal is to conduct a cluster-randomized trial of the relative impact and cost-effectiveness of a routine TB screening algorithm -- symptom screening and point-of-care LED fluorescence sputum smear microscopy - versus a GeneXpert-based TB screening algorithm - symptom screening and point-of-care GeneXpert testing - on reducing morbidity and mortality due to TB among HIV-infected Malawians. 12 public sector clinics in southern Malawi will be randomized to 1 of 2 algorithms for TB case detection as part of ICF and for exclusion of TB prior to IPT and antiretroviral therapy (ART) initiation. In the clinics assigned to the GeneXpert algorithm, newly diagnosed HIV-infected patients will be screened for symptoms of TB and, if symptomatic, will provide sputum for GeneXpert point-of-care TB testing. Under the routine screening algorithm, patients at a clinic will be screened for symptoms of TB and, if symptomatic, will provide sputum for point-of-care LED fluorescence smear microscopy. The current standard of care in Malawi and most of Africa is symptom screening and sputum smear microscopy using Ziehl-Neelsen stain (not LED fluorescence microscopy) alone. There is not sufficient evidence at this time to demonstrate whether point-of-care LED microscopy versus GeneXpert testing is likely to be superior with respect to clinical impact and/or cost-effectiveness. Outcomes will be measured at the clinic level.