A clinician uses a new HIV screening test not yet FDA-approved. You test positive. Which statistic would help you figure your probability of being HIV positive?

• A. Sensitivity
• B. Specificity
• C. Predictive value positive
• D. Predictive value negative

Answer: (C)

Positive predictive value is the statistic that tells you the probability you actually have HIV given a positive test result. It answers the question “What is the chance I’m infected if the test is positive?”

This value depends on three things: how good the test is at detecting disease when it’s present (sensitivity), how good it is at ruling out disease when it’s absent (specificity), and how common the disease is in the population being tested (prevalence). The formal relationship is P(Disease | Positive) = [Sensitivity × Prevalence] / [Sensitivity × Prevalence + (1 − Specificity) × (1 − Prevalence)]. So even a test with decent sensitivity and specificity can yield a relatively low probability of disease after a positive result if the disease is rare in the tested group.

For example, with a sensitivity of 90%, specificity of 95%, and a prevalence of 1%, the probability you truly have HIV after a positive result is about 15%. If prevalence were higher, this probability would be higher as well. In practice, positive results from a screening test—especially one not FDA-approved—are typically followed by confirmatory testing to establish the true diagnosis.

So the statistic needed to figure the probability of being HIV positive after a positive test is the predictive value positive.