Number Allowed to Diagnose (NAD), Number Needed to Misdiagnose (NNM) and Probability of Misclassification

Description

Habibzadeh reported the number needed to misdiagnose (NNM). Gerke and Vach suggested that the number allowed to diagnose (NAD) may be a better term to use.

NAD =

= 1 / (probability of misclassification)

The larger the NAD the better the test is for making a diagnosis (and avoiding a misdiagnosis).

Parameters:

(1) sensitivity (SE)

(2) specificity (SP)

(3) prevalence (P)

probability of misclassification =

= ((prevalence)*(1 - SE)) + ((1 - SP) * (1 - (prevalence)))

To read more or access our algorithms and calculators, please log in or register.