Diagnostic Accuracy

The Diagnostic Accuracy calculator evaluates how well a diagnostic test identifies a condition, given the four cells of a confusion matrix.


Step 1 — Enter the confusion matrix

Cell Meaning
True Positives (TP) Test positive, disease present
False Positives (FP) Test positive, disease absent
False Negatives (FN) Test negative, disease present
True Negatives (TN) Test negative, disease absent

Also supports Excel import and Sample Data Generator.


Step 2 — Calculate

Click Calculate Diagnostic Accuracy.


Step 3 — Read the results

All metrics with 95% CI:

  • Sensitivity — TP / (TP + FN)
  • Specificity — TN / (TN + FP)
  • Positive Predictive Value (PPV) — TP / (TP + FP)
  • Negative Predictive Value (NPV) — TN / (TN + FN)
  • Positive Likelihood Ratio (LR+) — Sensitivity / (1 - Specificity)
  • Negative Likelihood Ratio (LR-) — (1 - Sensitivity) / Specificity
  • Accuracy — (TP + TN) / N
  • Prevalence — (TP + FN) / N
  • Youden's Index — Sensitivity + Specificity - 1
  • Diagnostic Odds Ratio (DOR) — (TP * TN) / (FP * FN)

Additional panels:

  • Predictive Value Graphs — PPV/NPV vs prevalence
  • Fagan's Nomogram — pre-test to post-test probability
  • Custom Prevalence Input — Bayes-theorem based PPV/NPV at any prevalence
  • Materials and Methods — auto-generated text for manuscripts

Step 4 — Multiple analyses history

Every calculation is logged in an Analysis Summary Table. Export all to Excel or clear history.

A UniversalChatBot is available below.


Statistical methods used

Confidence intervals — Wilson Score Method

All proportion-based metrics use Wilson Score CI:

denominator = 1 + z^2/n
centre = (p + z^2/(2n)) / denominator
margin = z * sqrt(p(1-p)/n + z^2/(4n^2)) / denominator

Likelihood ratio CIs — computed on log scale and back-transformed.

Custom prevalence (Bayes theorem)

PPV = (Se * prev) / (Se * prev + (1-Sp) * (1-prev))
NPV = (Sp * (1-prev)) / ((1-Se) * prev + Sp * (1-prev))

Clinical utility tiers

LR+ LR- Label
> 10 < 0.1 Excellent
> 5 < 0.2 Good
> 2 < 0.5 Fair
Otherwise Limited