ROC curve analysis: predictive values

Command:

Statistics ROC curves Predictive values

Description

When you do have access to the raw data to perform ROC curve analysis, you can still calculate positive and negative predictive values for a test when the sensitivity and specificity of the test as well as the disease prevalence (or the pretest probability of disease) are known, using Bayes' theorem.

Required input

Enter the sensitivity and specificity of a test (expressed as percentages), and the disease prevalence (also expressed as a percentage).

Optionally you can enter the number of cases in the diseased and normal groups. These are the number of cases included in the study in which sensitivity and specificity were established. Input of these numbers will enable MedCalc to calculate 95% confidence intervals for the positive and negative predictive values.

When these data are entered click Test or press Enter to see the results.

Results

• Positive predictive value (PPV): probability that the disease is present when the test is positive (expressed as a percentage).
  $$ PPV = \frac {sensitivity \times prevalence } {sensitivity \times prevalence + (1-specificity)\times (1-prevalence) } $$
• Negative predictive value (NPV): probability that the disease is not present when the test is negative (expressed as a percentage).
  $$ NPV = \frac {specificity \times (1-prevalence) }{ (1-sensitivity) \times prevalence + specificity \times (1-prevalence) } $$
• When the sample sizes in the diseased and normal groups are known, MedCalc calculates confidence intervals for the predictive values using the standard logit method given by Mercaldo et al. 2007.

Literature

• Mercaldo ND, Lau KF, Zhou XH (2007) Confidence intervals for predictive values with an emphasis to case-control studies. Statistics in Medicine 26:2170-2183.

See also

• ROC curves: Introduction
• Interactive dot diagram
• Plot versus criterion values
• Interval likelihood ratios
• Comparison of ROC curves