A 54-year-old man with hypertension and hyperlipidemia who came to the emergency department with chest pain wants to know if he is having a heart attack. Test A is newly available for diagnosing myocardial infarction (MI). In a recent study, the results of test A (compared to a gold standard diagnosis of MI) were as follows:
MI No MI Test A positive 200 50 Test A negative 120 80
The patient has a positive result on test A. Assuming his pre-test probability is equivalent to the prevalence of MI in the study, what is the probability that the patient has an MI?
The positive predictive value (PPV) of a diagnostic test answers the following question: Given a positive test result, what is the probability that a patient has the disease? PPV corresponds to the number of people with the disease who test positive among all those who test positive. Using a standard contingency (2 × 2) table, PPV = a / (a + b).
Show Explanatory Sources
Unlike specificity and sensitivity, PPV varies with disease prevalence. If disease prevalence increases, PPV increases; similarly, PPV decreases with decreasing prevalence. In this question, the patient's pre-test probability (which takes into account clinical judgment regarding how likely it is that he has an MI) is assumed to be equivalent to the prevalence of MI in the study, making the results directly translatable. Therefore, PPV = 200 / (200+50) = 200/250 = 0.8 or 80%. In other instances, clinicians may assume that pre-test prevalence equals disease prevalence in the population.
Educational objective:
Positive predictive value represents the probability of truly having a disease given a positive test result. It increases with increasing disease prevalence and decreases with decreasing disease prevalence.