Skip to main content

Advertisement

Table 2 Connecting sensitivity, specificity, and prevalence, to predictive values and NNT values

From: Planning clinically relevant biomarker validation studies using the “number needed to treat” concept

   
Sensitivity 0.50 0.60 0.70 0.80 0.90 1.00  
Specificity 0.50 0.60 0.70 0.80 0.90 1.00
PPV 0.50 0.60 0.70 0.80 0.90 1.00 Prevalence = 0.5 (see Fig. 3a)
NPV 0.50 0.60 0.70 0.80 0.90 1.00 Prevalence = 0.5 (see Fig. 3a)
NNT Pos 2.0 1.7 1.4 1.3 1.1 1.00 Prevalence = 0.5 (see Fig. 3a)
NNT Neg 2.0 2.5 3.3 5.0 10.0 Inf Prevalence = 0.5 (see Fig. 3a)
PPV 0.05 0.07 0.11 0.17 0.32 1.00 Prevalence = 0.05 (see Fig. 3b, c)
NPV 0.95 0.97 0.98 0.99 0.99 1.00 Prevalence = 0.05 (see Fig. 3b, c)
NNT Pos 20.0 13.7 9.1 5.8 3.1 1.00 Prevalence = 0.05 (see Fig. 3b, c)
NNT Neg 20.0 29.5 45.3 77.0 172.0 Inf Prevalence = 0.05 (see Fig. 3b, c)
  1. Circled letters refer to points labeled in Fig. 3. Example: for column D, if the test sensitivity and specificity both equal 0.80, and the prevalence is 0.05, then the predictive values for the test are respectively 0.17 and 0.99 (point D graphed in Fig. 3b), and the NNT values in the positive and negative test groups are respectively 5.8 and 77.0 (point D graphed in Fig. 3c)