CHDS faculty Nicolas Menzies and third-year doctoral student in the Department of Global Health and Population, Sarah Baum, co-authored a study published in PLOS Computational Biology developing a modeling approach to correct for potential biases in estimates of drug resistant Tuberculosis (DR-TB) prevalence estimated from rapid diagnostic tests (RDTs).
While DR-TB data may be routinely collected by National TB Control Programs RDTs, these data streams may not be fully utilized for surveillance where low testing coverage may bias inferences due to systematic differences in RDT access.
The authors collaborated with colleagues from the Ministry of Health in Brazil to estimate the prevalence of rifampicin resistance (RR-TB) – one of the first-line TB drugs — using national case-level notification data. Since 2015, Brazil has expanded RDTs into routine TB care, but testing is not yet universal. Modeled estimates of RR-TB prevalence among notified TB cases were substantially higher for new cases (up to 44%) and previously treated cases (up to 17%) compared to naïve estimates. Modeled estimates were also more precise compared to existing estimates of RR-TB prevalence from the World Health Organization.
In addition to being valuable in settings where testing coverage is low or variable, this approach may also be useful in settings with high testing coverage that could benefit from additional statistical correction.
Learn more: Read the paper, Surveillance for TB Drug Resistance Using Routine Rapid Diagnostic Testing Data: Methodological Development and Application in Brazil
Read more: Cost-Effectiveness of Malaria Testing and Response
Read more: Risk Factors Underlying Disparities in TB Outcomes