A test with 90% sensitivity and 90% specificity sounds reliable. But use it for screening, where 1 in 100 people has the disease, and fewer than 1 in 10 positive results will be true positives. Our new Predictive Value Explorer shows why.

Sensitivity and specificity describe how well a test detects disease in samples with a known diagnosis. Patients and doctors ask a different question: if the result is positive, how likely is it that I have the disease? The positive and negative predictive values (PPV and NPV) answer that question. They depend on a third number: the prevalence, or pre-test probability, of the disease in the people being tested.

When the pre-test probability is low, as in screening, even a good test produces many false positives, and only a very high specificity keeps the PPV useful. When the pre-test probability is high, the NPV drops instead. This is also why predictive values from 50:50 case–control studies rarely carry over to clinical practice.

With more than 13 years in biomarker discovery and diagnostic test development, we work with these numbers every day at TAmiRNA, and we know how easy it is to lose track of how they interact. We built the Predictive Value Explorer to make the relationships visible:

  • Enter sensitivity, specificity and prevalence
  • See PPV, NPV and likelihood ratios update instantly
  • Read the expected results per 10,000 people tested
  • Follow the calculation on a live Fagan nomogram
  • Compare screening, symptomatic and case–control settings with one-click presets

Use it to plan biomarker validation and clinical performance studies, to interpret published accuracy data, or to define the intended purpose of an IVD under the IVDR.

[Try the Predictive Value Explorer →]