Specialty › Tests on summarized data

Diagnostic test (sens/spec/prev → PPV/NPV/LR)

The diagnostic calculator turns sensitivity, specificity and pre-test prevalence into PPV, NPV, LR+ and LR−.

What is Diagnostic test (sens/spec/prev → PPV/NPV/LR)?

Sensitivity / specificity are *test* properties — they don't depend on prevalence. PPV / NPV are *population* properties — they slide with prevalence. This calculator gives the conversion from one to the other for any prevalence you specify.

Likelihood ratios are prevalence-independent: LR+ = Sens / (1 − Spec); LR− = (1 − Sens) / Spec. LR+ > 10 (or LR− < 0.1) is the rule-of-thumb for strong evidence; the post-test odds = pre-test odds × LR.

When should I use Diagnostic test (sens/spec/prev → PPV/NPV/LR)?

  • Bedside diagnostic calculations.
  • Comparing test performance across populations with different prevalence.

What data does it need?

Sensitivity, specificity, prevalence (all 0 < x < 1).

What does it report?

PPV, NPV, LR+, LR− at the chosen prevalence.

Formula

PPV = Sens · prev / (Sens · prev + (1 − Spec) · (1 − prev))
NPV = Spec · (1 − prev) / ((1 − Sens) · prev + Spec · (1 − prev))

How do I interpret the result?

PPV is what the patient wants to know: 'I tested positive — what's the probability I actually have it?'. NPV is the reassurance after a negative test.

See also

References

  • Sackett et al. (2000). Evidence-Based Medicine: How to Practice and Teach EBM.