How Is Negative Predictive Value Calculated

Negative predictive value (NPV) is a key metric in medical testing and diagnostics that measures the probability a test result is accurate when the test is negative. This guide explains how to calculate NPV, its formula, interpretation, and practical applications.

What Is Negative Predictive Value?

Negative predictive value (NPV) is a statistical measure that answers the question: "If a test is negative, how likely is it that the person actually doesn't have the condition?"

NPV is particularly important in medical testing where false negatives (missing a condition) can have serious consequences. It's one of four key metrics in a 2×2 contingency table used to evaluate diagnostic tests:

True negatives (TN)
False negatives (FN)
True positives (TP)
False positives (FP)

NPV is different from negative predictive power, which refers to the effect size of a negative result rather than its probability.

NPV Formula

Negative Predictive Value = (True Negatives) / (True Negatives + False Negatives)

Mathematically: NPV = TN / (TN + FN)

The formula shows NPV is calculated by dividing the number of true negatives by the sum of true negatives and false negatives. This gives the proportion of negative test results that are actually correct.

How to Calculate NPV

To calculate NPV, you need the following data from a diagnostic test:

Number of true negatives (TN)
Number of false negatives (FN)

Using these values, you can apply the formula:

NPV = TN / (TN + FN)

For example, if a test has 900 true negatives and 50 false negatives, the NPV would be calculated as:

NPV = 900 / (900 + 50) = 900 / 950 = 0.947 or 94.7%

Interpreting NPV Results

NPV results are interpreted as probabilities:

A NPV of 95% means there's a 95% chance the person doesn't have the condition when the test is negative.
A higher NPV indicates the test is more reliable for identifying true negatives.
NPV is particularly important when false negatives are costly (e.g., missing a disease).

NPV should be considered alongside other metrics like sensitivity (true positive rate) and specificity (true negative rate) for a complete picture of test performance.

Worked Example

Let's calculate NPV for a hypothetical HIV test:

True negatives: 1,200
False negatives: 20

NPV = 1,200 / (1,200 + 20) = 1,200 / 1,220 ≈ 0.984 or 98.4%

This means when the test is negative, there's a 98.4% probability the person doesn't have HIV. The test has excellent NPV, meaning it's very reliable for identifying people who don't have the disease.

FAQ

What's the difference between NPV and specificity?

NPV measures the probability a negative test result is correct, while specificity measures the proportion of true negatives among all actual negatives. NPV is a probability, while specificity is a proportion.

How does NPV relate to false negatives?

NPV is directly affected by false negatives. A higher number of false negatives will decrease NPV, as it means more people with the condition are incorrectly tested as negative.

Can NPV be 100%?

Yes, NPV can be 100% if there are no false negatives (FN = 0). This would mean every negative test result is correct, though this is rare in practice.