Positive predictive value (PPV) is calculated by dividing the number of true positives by the total number of positive test results. The formula is PPV = True Positives / (True Positives + False Positives). This gives you the probability that a person who tested positive actually has the condition. It’s one of the most practical measures in diagnostic testing because it answers the question patients and clinicians care about most: if my test came back positive, how likely is it that I’m actually sick?
The Basic Formula
The core calculation is straightforward. You need two numbers from your test results: true positives (people who tested positive and really do have the condition) and false positives (people who tested positive but don’t actually have it). Divide the true positives by the sum of both:
PPV = True Positives / (True Positives + False Positives)
Say a screening test flags 100 people as positive. Of those 100, 80 truly have the disease and 20 don’t. The PPV is 80 / (80 + 20) = 0.80, or 80%. That means each positive result carries an 80% chance of being correct. The remaining 20% are false alarms.
The Bayesian Formula Using Sensitivity, Specificity, and Prevalence
You won’t always have raw counts of true and false positives. Sometimes all you know about a test is its sensitivity (how well it catches true cases), its specificity (how well it correctly clears people without the disease), and the prevalence of the condition in the population you’re testing. In that case, you can calculate PPV using a version of Bayes’ theorem:
PPV = (Sensitivity × Prevalence) / [(Sensitivity × Prevalence) + ((1 − Specificity) × (1 − Prevalence))]
This looks more complex, but each piece has a clear role. The numerator captures the proportion of the population that is both sick and correctly identified. The denominator adds in the proportion that is healthy but incorrectly flagged. The ratio between these two gives you the same answer as the basic formula, just derived from different starting information.
A Worked Example
Imagine a test with 90% sensitivity and 95% specificity, used in a population where 2% of people have the disease. Plugging into the formula:
PPV = (0.90 × 0.02) / [(0.90 × 0.02) + ((1 − 0.95) × (1 − 0.02))]
That simplifies to 0.018 / (0.018 + 0.049) = 0.018 / 0.067 = 0.269, or about 27%. So even with a test that sounds impressive on paper, only about 1 in 4 positive results is a true case when the disease is rare. This surprises most people, and it’s the single most important thing to understand about PPV.
Why Prevalence Changes Everything
Sensitivity and specificity are fixed properties of a test. They don’t change based on who you’re testing. PPV is different. It shifts dramatically depending on how common the condition is in the group being screened.
When prevalence is low, false positives pile up. Even a small false positive rate, applied to a huge number of healthy people, generates a lot of incorrect positive results. Those false positives flood the denominator of the PPV formula, dragging the value down. This is exactly why breast cancer screening by clinical exam has a PPV of only about 1.5% to 4%, depending on the age group. Cancer is relatively uncommon in the screened population, so most positive findings turn out to be benign.
As prevalence rises, the math flips. When you test a population where the disease is common (a hospital ward during an outbreak, for instance), there are far more true positives relative to false positives, and PPV climbs. The same test with the same sensitivity and specificity can have a PPV of 27% in one setting and 90% in another, purely because of prevalence.
PPV vs. Sensitivity and Specificity
A common mistake is assuming that a highly sensitive test will have a high PPV. Sensitivity tells you how many sick people the test catches, but it says nothing about how many healthy people it incorrectly flags. A test can detect 99% of true cases and still have a low PPV if it also produces many false positives.
Specificity has a more direct relationship with PPV. Higher specificity means fewer false positives, which tightens the denominator and pushes PPV up. If you’re trying to improve a test’s PPV without changing the population you’re testing, increasing specificity is the lever that matters most.
Think of it this way: sensitivity and specificity describe how a test performs in the lab. PPV describes what a positive result means for a real person in a real population. It’s the bridge between test performance and clinical reality.
PPV vs. Negative Predictive Value
Negative predictive value (NPV) is the mirror image of PPV. While PPV answers “if I tested positive, do I really have it?”, NPV answers “if I tested negative, am I really clear?” The formula is:
NPV = True Negatives / (True Negatives + False Negatives)
The two values move in opposite directions as prevalence changes. When a disease is rare, NPV tends to be very high (a negative result is almost certainly correct, because most people don’t have the condition anyway). PPV, meanwhile, tends to be low. When a disease is common, PPV rises but NPV falls.
In practice, this means different tests serve different purposes. A screening test designed to catch every possible case (high sensitivity, high NPV) is useful for ruling out a disease. A confirmatory test designed to minimize false positives (high specificity, high PPV) is useful for ruling one in. Knowing which value matters depends on the clinical question being asked.
How to Build a 2×2 Table
If you’re calculating PPV from scratch, organizing your data into a 2×2 table makes the process much easier. The rows represent test results (positive or negative) and the columns represent actual disease status (present or absent).
- Cell A (top left): True positives. Tested positive, disease present.
- Cell B (top right): False positives. Tested positive, disease absent.
- Cell C (bottom left): False negatives. Tested negative, disease present.
- Cell D (bottom right): True negatives. Tested negative, disease absent.
PPV uses only the top row: A / (A + B). NPV uses only the bottom row: D / (C + D). Sensitivity uses the left column: A / (A + C). Specificity uses the right column: D / (B + D). Once you fill in the four cells, every major test statistic falls out with simple division.
To populate the table when you only have sensitivity, specificity, and prevalence, start with a hypothetical population (1,000 or 10,000 makes the math clean). Multiply by prevalence to get the number of truly sick people. Apply sensitivity to that group to get true positives (Cell A) and false negatives (Cell C). Apply specificity to the healthy group to get true negatives (Cell D) and false positives (Cell B). Then calculate PPV from the completed table.
Practical Interpretation
PPV is a conditional probability. It tells you the chance that disease is present, given that the test was positive. This is exactly the information a person needs after receiving a positive result. A PPV of 30% means that for every 10 people who get a positive result, roughly 3 actually have the condition and 7 do not.
Context matters enormously when interpreting PPV. The same test can be highly reliable in a high-risk population and nearly useless as a broad screen. This is why many screening programs use a two-step process: a sensitive initial test to cast a wide net, followed by a specific confirmatory test to weed out false positives and raise the effective PPV before anyone receives a diagnosis.