# A reference interval is not always a decision threshold

The shaded band beside a laboratory result looks authoritative. It is authoritative for the particular report, but it does not necessarily mark the boundary between health and disease or the point where treatment begins. Before reading the colour, it helps to understand what kind of boundary the report is showing.

A reference interval describes results from a selected reference population using a particular approach. A clinical decision limit serves a different purpose: it supports a decision about investigation, risk or management. A previous result from the same person can provide another comparison again. These are three different questions, even when all are expressed as numbers.

[MedlinePlus's guide to laboratory results](https://medlineplus.gov/lab-tests/how-to-understand-your-lab-results/) explains that healthy people can have results outside a reference range and people with a health problem can have results inside it. It also notes that laboratories may use different methods and ranges. A flag is useful information, but it does not establish the clinical conclusion by itself.

## A deliberately fictional example

Imagine a made-up measurement called Test Z. Its report displays a reference interval of 10 to 20 units. A result of 21 is outside that interval; a result of 19 is inside. Those statements are enough to describe the flags. They tell us nothing about Test Z's purpose, the reference population or whether either person needs a clinical action.

Now imagine that a separate, properly established decision rule for this fictional test uses another boundary for a specific investigation. That rule would need its own evidence and scope. We cannot invent it from the interval printed beside the result. No real analyte, diagnostic cutoff or treatment recommendation is represented by this example.

A value moving from 11 to 19 would still be inside the same interval. Whether that change matters is another question. It cannot be settled by the lack of a flag, nor can it be declared significant merely because the increase looks large on a phone screen. The appropriate comparison depends on the measurement and the person's circumstances.

## Where the interval comes from

[CLSI's EP28 overview](https://clsi.org/shop/standards/ep28/) describes a laboratory standard concerned with establishing and verifying reference intervals. The public overview explains its scope; it is not the full standard, and this article does not claim to have inspected every method it contains. The existence of that work is a reminder that an interval is constructed and checked, not discovered as a universal number independent of the laboratory.

The methodological article [“Are my Laboratory Results Normal?”](https://pmc.ncbi.nlm.nih.gov/articles/PMC4975205/) distinguishes reference intervals from decision limits and from comparison with an individual's previous results. Published in 2008, it is used here for that conceptual distinction, not as a current source of disease thresholds or clinical treatment rules.

This matters when someone posts an “optimal” range online. A narrower band might be asserted without identifying a reference population, a laboratory method or a demonstrated clinical decision. Replacing the report's interval with that band does not automatically improve interpretation. Ask what the band represents and which decision it is meant to support.

## Read the boundary's label

A report may display a reference interval, a target, a comment or a disease-specific decision limit. Their labels matter. If you cannot tell which one you are seeing, asking the requester is more useful than searching for a single universal range. Keep the exact wording rather than copying every boundary into a column called “normal”.

For an editorial example, a reader might write: “The portal flags this value against its interval. Is that interval a population comparison, or is there a separate decision rule relevant to why I was tested?” That question does not dismiss the flag. It asks how the flag will be used.

Do not make an urgent instruction wait for an explanation of statistical terminology. If a service contacts you about a result or provides a next step, follow that instruction. Equally, a value inside an interval should not be used to disregard symptoms or cancel an agreed clinical review.

The practical next step is to locate the label beside one result and ask what it compares you with. Understanding that label will not diagnose anything, but it prevents a common error: asking a population band to answer every clinical question.

[The DomDNA quiz](https://domdna.com/quiz) is a separate educational lifestyle questionnaire. It performs no DNA analysis, interprets no laboratory boundaries and cannot determine whether a result requires treatment.

Draft • Research checked 4 October 2026 • AI-assisted DomDNA editorial content. No independent clinical review or publication is claimed. General education, not individual medical advice.