Bayesian Decision Tree Model
The model simulates a hypothetical cohort of 1,000 participants for every algorithm, country, and population setting. Diagnostic performance parameters use beta distributions derived from posterior distributions estimated in the diagnostic…
1 sources - 5 claims
The model simulates a hypothetical cohort of 1,000 participants for every algorithm, country, and population setting. Diagnostic performance parameters use beta distributions derived from posterior distributions estimated in the diagnostic performance analysis. The study uses a Bayesian decision tree model to simulate costs and confirmed TB detections for all pre-selected algorithms, built separately for facility-based and community-based case finding in Amua version 0.3.1. Effectiveness is measured as the number of people with confirmed TB detected by each algorithm. A total of 22 independent model runs will be performed across settings, including HIV-status stratification for aggregate analyses.