CIfinder

Computes confidence intervals for the positive predictive value (PPV) and negative predictive value (NPV) based on varied scenarios. In situations where the proportion of diseased subjects does not correspond to the disease prevalence (e.g. case-control studies), this package provides two types of solutions: 1) five methods for estimating confidence intervals for PPV and NPV via ratio of two binomial proportions including Gart & Nam (1988), Walter (1975), MOVER-J (Laud, 2017), Fieller (1954), and Bootstrap (Efron, 1979); 2) three direct methods that compute the confidence intervals including Pepe (2003), Zhou (2007), and Delta. In prospective studies where the proportion of diseased subjects is an unbiased estimate of the disease prevalence, this package provides several methods for calculating the confidence intervals for PPV and NPV including Clopper-Pearson, Wald, Wilson, Agresti-Coull, and Beta. See the Details and References sections in the corresponding functions.