Why More Testing Produces More False Positives

Comprehensive panels are marketed on the reasoning that more information is better. The arithmetic of reference ranges works against that assumption in a specific and predictable way.

The arithmetic is straightforward

If a reference range covers the middle ninety-five percent of a healthy population, then one in twenty healthy people falls outside it on any single test.

Run twenty independent tests on a healthy person and the expected number of abnormal results is one. Run more and the expected number rises proportionally.

The probability of at least one abnormal result across a large panel approaches certainty. This follows from how the ranges were defined, not from any measurement error.

Correlation between tests changes the numbers

Many measurements on a panel are not independent. Related markers move together, which reduces the effective number of separate opportunities for a false flag.

Correlation cuts both ways. When related markers are abnormal together, they may all reflect one underlying deviation rather than several separate problems.

Reading a panel therefore involves looking at patterns across related measures rather than counting individually flagged values.

Biological variation adds a second layer

Individual values fluctuate over time around a personal set point, driven by daily rhythms, recent meals, hydration, activity and illness.

For some measures this within-person variation is large relative to the width of the reference range, meaning a single sample may land outside on one day and inside on another.

This is why repeating an unexpected abnormal result before acting on it is standard practice. Repetition separates persistent deviation from momentary fluctuation.

Consequences fall on the person tested

An abnormal result creates concern that persists through the period of investigation, and that period can extend for weeks depending on what follow-up is required.

Follow-up investigations carry their own burdens, including procedures with genuine though small risks, and further tests with their own false positive rates.

Where investigation ends without a finding, the original result is usually never explained, which is an unsatisfying outcome that the testing decision made likely.

How clinical practice manages the problem

Ordering tests in response to a specific question raises the prior likelihood, which improves the meaning of a positive result substantially.

Restraint in test ordering is therefore a considered position rather than a limitation of resources, and guidelines increasingly discourage untargeted panels.

Anyone holding a panel with flagged results should have them interpreted by a clinician, who can distinguish which warrant attention and which are statistically expected.