Skip to contents

Four tumor marker dataset with 220 patients for evaluating marker panel performance without gold standard. Markers: CA125, HE4, CEA, AFP with varying sensitivity (0.75-0.68) and specificity (0.88-0.85).

Usage

nogoldstandard_tumormarker

Format

A data frame with 220 rows and 7 variables:

patient_id

Character: Patient identifier (PT001-PT220)

CA125

Factor: CA125 level ("Normal", "Elevated"), Sens=0.75, Spec=0.88

HE4

Factor: HE4 level ("Normal", "Elevated"), Sens=0.70, Spec=0.85

CEA

Factor: CEA level ("Normal", "Elevated"), Sens=0.68, Spec=0.90

AFP

Factor: AFP level ("Normal", "Elevated"), Sens=0.72, Spec=0.87

age

Numeric: Patient age in years (mean 62, SD 10)

risk_category

Factor: Risk level (Low, Moderate, High)

Source

Generated test data for ClinicoPath package

Details

Simulated with 20% cancer prevalence (screening context). Multiple markers enable composite reference and latent class analysis comparisons.

Examples

data(nogoldstandard_tumormarker)
nogoldstandard(data = nogoldstandard_tumormarker,
               test1 = "CA125", test1Positive = "Elevated",
               test2 = "HE4", test2Positive = "Elevated",
               test3 = "CEA", test3Positive = "Elevated",
               test4 = "AFP", test4Positive = "Elevated",
               clinicalPreset = "tumor_markers")
#> Error in nogoldstandard(data = nogoldstandard_tumormarker, test1 = "CA125",     test1Positive = "Elevated", test2 = "HE4", test2Positive = "Elevated",     test3 = "CEA", test3Positive = "Elevated", test4 = "AFP",     test4Positive = "Elevated", clinicalPreset = "tumor_markers"): argument "test5Positive" is missing, with no default