An administrative data validation study of the accuracy of algorithms for identifying rheumatoid arthritis: the influence of the reference standard on algorithm performance

Table 1 Methods to computing measures of diagnostic accuracy

		Reference standard		$Pre - test prevalence = \frac{TP + FN}{TP + FP + FN + TN}$
		RA cases	Non-cases	$Pre - test prevalence = \frac{TP + FN}{TP + FP + FN + TN}$
Administrative data	Positive	True Positive (TP)	False Positive (FP)	$PPV = \frac{TP}{TP + FP}$
Administrative data	Negative	False Negative (FN)	True Negative (TN)	$NPV = \frac{TN}{FN + TN}$
		$Sensitivity = \frac{TP}{TP + FN}$	$Specificity = \frac{TN}{FP + TN}$	$Post - test Prevalence = \frac{TP + FP}{TP + FP + FN + TN}$

ISSN: 1471-2474