afdx

Diagnosis performance using attributable fraction

This R-package help on the estimation of diagnosis performance (Sensitivity, Specificity, Positive predictive value, Negative predicted value) of a diagnostic test where the golden standard can’t be measured but can be estimated using the attributable fraction

Two methods are presented with examples for Malaria diagnosis, using a maximum likelihood estimated logistic exponential model and using a bayesian latent class model.

To install the package from github use:

devtools::install_github("johnaponte/afdx", build_manual = T, build_vignettes = T)