HIV testing, care, and treatment
Johns Hopkins University
Perinatal HIV Research Unit, University of the Witwatersrand
Matlosana sub-district health services
Matlosana District, Northwest Province, South Africa
Richard Chaisson, Johns Hopkins University
Nicole Salazar-Austin, Johns Hopkins University

Preventing Tuberculosis among Pregnant Women and Child Contacts under Five Years in Matlosana, South Africa

The issue
Approximately 15 million children worldwide are household contacts of adults with tuberculosis (TB) disease. HIV-positive adults are the source case for a large proportion of pediatric TB cases, but few contacts of these adult cases are screened. Although prevention of TB with isoniazid preventive therapy (IPT) for children less than five years old has been a longstanding recommendation, implementation of such programs has been hampered by a number of challenges at the patient, provider, program, and policy levels.
Our approach
Project SOAR seeks to improve rates of pediatric IPT for child contacts of mostly HIV-positive TB index cases by integrating a nurse-led, symptom-based screen into decentralized TB services at community clinics in South Africa. The intervention allows nurses to start IPT for asymptomatic patients and refer symptomatic patients to a pediatrician for evaluation of TB disease and possible anti-tuberculosis therapy. The intervention is being assessed using a cluster-randomized trial design where 16 primary health clinics in the Matlosana sub-district of North West Province are stratified and randomized to either the current standard of care, tuberculin skin test (TST)-based screening, or symptom-based screening, the newly recommended but not implemented standard of care in South Africa. Nurses initiate TST-negative or asymptomatic children on IPT, monitor them throughout treatment, and measure adherence to therapy using a newly designed pediatric contact register.
Preliminary results indicate that up to 50 percent of child contacts may not be identified, when compared to a historical home-based contact tracing study performed in Matlosana in 2009, and an additional 50 percent of those identified children are never screened.
In phase two of this study, Project SOAR is evaluating the effectiveness of home-based versus clinic-based child contact evaluation including contact tracing, symptom screening, and IPT initiation, using outreach teams of retired professional nurses and community health workers. Again using a cluster-randomized design, we are re-randomizing the 16 clinics to either a clinic-based or home-based evaluation, comparing the proportion of child contacts that initiate and complete IPT. We expect home-based symptom-based screening of child TB contacts will be feasible and cost-effective for both patients and the health care system. In addition, SOAR is also evaluating an integrated approach to TB evaluation in six antenatal care clinics where we are systematically screening all pregnant women, identifying and treating active TB disease, and promoting prevention through initiation of IPT among pregnant women living with HIV. Study outcomes include the proportion of pregnant women screened for TB, proportion of pregnant women with a positive TST, proportion of pregnant women living with HIV initiated on IPT, and treatment outcomes including completion of therapy, early termination due to side effects, hepatotoxicity or personal preferences, and loss to follow up.
The impact
Evidence from the study is particularly important for preventing pediatric TB and screening pregnant women for TB in South Africa and other high-burden settings, given the country’s co-epidemics of HIV and TB and low levels globally of IPT initiation.