Diagnostic accuracy calculator and teaching tool

Diagnostic Accuracy Calculator

Enter raw values and prevalence to calculate sensitivity, specificity, PPV, NPV, LR+, and LR-. Export results as PDF, explore prevalence effects, use a Fagan nomogram, and calculate ROC and AUC from data.

Same test. Same sensitivity. Same specificity. Different prevalence. Different clinical meaning.
Will this paper work in my hospital?

Disclaimer

Educational use only. This calculator supports learning and journal club interpretation. It does not replace clinical judgment, local guidelines, laboratory validation, or patient-specific decision-making.

Default demonstration

The visual lesson should land before the formulas do.

Start by calculating the usual diagnostic accuracy measures from your own data. Then use the slider and nomogram tools to see what those numbers mean clinically.

If you wish, you can also load the NGAL example to see the prevalence problem play out as a guided demonstration.

Sensitivity, Specificity, PPV, NPV
LR+ and LR-
Prevalence Slider
Fagan Nomogram
ROC and AUC
NGAL Example Preset
PDF Export
Learning Booklet

Flagship Experience

Guided Custom Demonstration

Load the NGAL preset or enter your own sensitivity, specificity, and prevalence values. The walkthrough will teach the same prevalence lesson either way.

Scenario Builder

No disease-specific preset is loaded. Enter your own values here.

Disclaimer

Educational use only. This calculator supports learning and journal club interpretation. It does not replace clinical judgment, local guidelines, laboratory validation, or patient-specific decision-making.

Step 1

Published Study Results

Sensitivity

85%

Looks strong.

Among patients who truly have target disease, most test positive.

Specificity

80%

Reasonably useful.

Among patients without target disease, many test negative, though false positives still occur.

PPV

59%

If Your test is positive in this published study population, about 59% of positive tests truly represent target disease.

At first glance, the test looks clinically useful.

NPV

94%

A negative result looks reassuring.

In this population, a negative test usually points away from target disease.

LR+

4.26

A positive result raises suspicion.

This looks like a modestly useful positive test.

LR-

0.19

A negative result lowers suspicion.

This looks like a useful rule-out contribution.

Headline

Looks like a useful test.

Published PPV 59%. Published NPV 94%.

Comparison

Published Study vs Your Ward

A side-by-side comparison that keeps sensitivity and specificity fixed while letting the clinical meaning shift.

FeaturePublished StudyYour Ward
Prevalence25%25%
Sensitivity85%85%
Specificity80%80%
PPV59%59%
NPV94%94%

Interactive Lab

Manual 2x2 Calculator

This is a diagnostic accuracy calculator. Enter raw values to calculate sensitivity, specificity, PPV, NPV, LR+, LR-, and related measures.

Enter study data

Use your own 2x2 table.

Sensitivity = TP / (TP + FN)

Specificity = TN / (FP + TN)

Predictive value always depends on disease prevalence in the population where you apply the test.

Sensitivity

85.2%

If 100 diseased patients are tested, about 85 will be detected.

This describes how often the test catches true disease.

Specificity

80.0%

If 100 patients do not have disease, about 80 will test negative.

This describes how often the test avoids false alarms.

PPV

58.7%

In this sample, a positive test means there is about a 59% chance of true disease.

This is the number most likely to collapse when prevalence falls.

NPV

94.2%

In this sample, a negative test means there is about a 94% chance of being disease-free.

This often rises when prevalence falls.

LR+

4.26

How much a positive test should raise suspicion.

Likelihood ratios belong to the test, not the patient population.

LR-

0.19

How much a negative test should lower suspicion.

Use this when asking how much reassurance a negative test really buys you.

Disclaimer

Educational use only. This calculator supports learning and journal club interpretation. It does not replace clinical judgment, local guidelines, laboratory validation, or patient-specific decision-making.

Prevalence Slider

Adjust prevalence and watch PPV and NPV change

Now you can adjust prevalence with the slider and see how PPV and NPV change despite the same sensitivity and specificity.

Prevalence25%
PPV
59%
NPV
94%

Disclaimer

Educational use only. This calculator supports learning and journal club interpretation. It does not replace clinical judgment, local guidelines, laboratory validation, or patient-specific decision-making.

Fagan Nomogram

Estimate post-test probability with Fagan's nomogram

Use the pre-test probability and likelihood ratios to understand what a positive or negative result means after testing.

Pre-test probability
25%
After positive test
59%
After negative test
6%

Disclaimer

Educational use only. This calculator supports learning and journal club interpretation. It does not replace clinical judgment, local guidelines, laboratory validation, or patient-specific decision-making.

Journal Club Mode

Will This Paper Work In My Hospital?

Enter the published numbers and your local prevalence. The app translates the paper into your clinical population.

Inputs

The test performance does not change. The patient mix does.

Published population

Published PPV
59%
Published NPV
94%

Your local population

Local PPV
32%
Local NPV
98%

The test performance did not change.

The patient population changed.

Published interpretation: a positive test means about 59% probability of disease. In your population, that becomes 32%.

Disclaimer

Educational use only. This calculator supports learning and journal club interpretation. It does not replace clinical judgment, local guidelines, laboratory validation, or patient-specific decision-making.

ROC and Cutoffs

ROC and AUC Calculator

This is a ROC and AUC calculator. Please enter data to calculate.

ROC input

Paste a CSV with `Disease_Status` and `Test_Value`. An example dataset is included.

ROC and AUC

AUC 1.00 (Perfect), 95% CI 1.00 to 1.00

Threshold explorer: Infinity
Sensitivity 0.0%
Specificity 100.0%
Youden Index 0.00
Optimal cutoff 6.20

Disclaimer

Educational use only. This calculator supports learning and journal club interpretation. It does not replace clinical judgment, local guidelines, laboratory validation, or patient-specific decision-making.

Reports and Precision

PDF Export and Confidence Intervals

Summarise the lesson and the supporting numbers in a publication-ready report.

Publication-quality PDF report

Download the current demonstration state or copy the summary.

Advanced statistics mode

Confidence intervals show how precise the 2x2 estimates are.

Sensitivity 95% CI: 80.3% to 89.1%
Specificity 95% CI: 77.0% to 82.7%
PPV 95% CI: 53.5% to 63.6%
NPV 95% CI: 92.1% to 95.8%
LR+ 95% CI: 3.66 to 4.96
LR- 95% CI: 0.14 to 0.25

Disclaimer

Educational use only. This calculator supports learning and journal club interpretation. It does not replace clinical judgment, local guidelines, laboratory validation, or patient-specific decision-making.

Free Booklet

Free one-page booklet on diagnostic accuracy

Click below to download a free booklet explaining the meaning of all major diagnostic accuracy terms along with a nomogram.

Author: Manjunath Kulkarni, Nephrologist, Father Muller Medical College, Mangaluru

Includes sensitivity, specificity, PPV, NPV, LR+, LR-, prevalence, and a simple nomogram guide.

Disclaimer

Educational use only. This calculator supports learning and journal club interpretation. It does not replace clinical judgment, local guidelines, laboratory validation, or patient-specific decision-making.

Learning Centre

What Clinicians Should Remember

The educational spine of the platform keeps returning to one central insight.

PPV and NPV change when disease prevalence changes, even if the test itself is unchanged.

A dramatic fall in prevalence usually makes false positives more common among all positive tests.

Likelihood ratios stay the same when the population changes. Post-test probability does not.

The visual memory to keep: as prevalence falls, the orange false-positive squares spread across the grid.

Study results can look strong in the paper but weaker in your hospital if the patient mix is different.

Sensitivity and specificity describe the test. PPV and NPV describe what the result means in a particular population.

Disclaimer

Educational use only. This calculator supports learning and journal club interpretation. It does not replace clinical judgment, local guidelines, laboratory validation, or patient-specific decision-making.