This ShinyApp can be used to estimate the sample size for a study where the aim is to test whether the performance of a diagnostic test is sufficient in terms of false positive (specificity) and true positive fraction (sensitivity).
The user needs to define values for the false positive fraction (FPF; specificity)
and true positive fraction (TPF; sensitivity) that are minimally acceptable in order to design the study.
The study will test the null hypothesis:
$$H_0 : \left[ \text{TPF} \leq \text{TPF}_0 \text{ or } \text{FPF} \geq \text{FPF}_0 \right] $$
From a study that rejects the null hypothesis it will be concluded that TPF and FPF meet the minimal criteria.
Sample sizes are calculated by using the formula based on asymptotic normal distribution theory as described in Pepe (section 8.2, 2003) or by simulations
using different exact or approximate confidence intervals for the difference of binomial
proportions as described in Agresti and Coull (1998).
Agresti, Alan, und Coull, Brent A.: Approximate is better than 'exact' for interval estimation of binomial proportions, The American Statistician 52, 119-126 (1998).
Margaret Sullivan Pepe.: The statistical evaluation of medical tests for classification and prediction,
Oxford University Press (2003).
Contact: m.sill(at)dkfz.de