Sensitivity and specificity calculator

FAQ

How are sensitivity and specificity calculated from a 2×2 table?

Sensitivity (true-positive rate) is TP / (TP + FN). Specificity (true-negative rate) is TN / (TN + FP). They describe how the index test behaves among people who do and do not have the condition according to the reference standard.

Why do PPV and NPV depend on prevalence?

Positive predictive value is TP / (TP + FP) and negative predictive value is TN / (TN + FN) inside the table you entered. Those counts embed that sample’s prevalence. The same sensitivity and specificity yield different PPV/NPV in a low-prevalence screening population than in a high-prevalence referral clinic. Use a prevalence-adjusted calculator when you need to project to a different setting.

What are LR+ and LR−?

The positive likelihood ratio is sensitivity / (1 − specificity). The negative likelihood ratio is (1 − sensitivity) / specificity. Likelihood ratios are a prevalence-independent way to update pre-test odds; they still assume the 2×2 estimates are stable.

Does this replace a CLSI or FDA diagnostic validation study?

No. Clinical performance for IVDs and many diagnostic devices follows study designs and reporting expectations in FDA guidance and standards such as CLSI documents for precision, LoD and method comparison. This page only does the arithmetic on counts you already have.

What if a cell is zero?

Division by zero makes a metric undefined (shown as a dash). Sparse tables also produce unstable estimates; confidence intervals and larger samples belong in the analysis plan, not in this helper.

How should I pick the reference standard?

The reference standard defines “truth” for TP/FP/FN/TN. A poor or discrepant reference biases every metric. Document the clinical reference standard, adjudication rules and any imperfect-gold-standard methods in the protocol before locking the 2×2.