da: Calculate diagnostic accuracy measures for binary measures...

In lbraglia/lbdiag: Diagnostic Accuracy Miscellaneous Utilities

Calculate diagnostic accuracy measures for binary measures (test, reference standard)

Description

Calculate diagnostic accuracy measures for binary measures (test, reference standard) and confidence intervals (two sided binomial for all the indexes except PPV and NPV, calculated via bdpv a-la Mercaldo)

Usage

da(
  test = NULL,
  refstd = NULL,
  alpha = 0.05,
  digits = 4L,
  positive_first = TRUE,
  ppv_npv_prev = NULL,
  ppv_npv_force_unadj = FALSE
)

Arguments

test	test variable (dichotomic factor or logical)
refstd	reference standard (dichotomic factor or logical)
alpha	type I error for two sided confidence interval
digits	rounding digits
positive_first	logical: display positive test and reference standard in first row column
ppv_npv_prev	prevalence adopted for ppv and npv confidence interval (if NULL it's estimated from sample)
ppv_npv_force_unadj	force unadjusted (standard logit) ppv npv estimates (and confidence interval) if at least one between NPV or PPV is 0 or 1

Examples

## CASS Example (Pepe pag 22)
db <- dadb(tn = 327, fn = 208, fp = 115, tp = 815)
with(db, da(test = test, refstd = refstd))

## Alzheimer Example (Mercaldo 2007)
db <- dadb(tn = 288, fn = 178 , fp = 87, tp = 240)
with(db, da(test = test, refstd = refstd, ppv_npv_prev = .03))

lbraglia/lbdiag documentation built on July 29, 2023, 3:35 p.m.