Reference Intervals and Decision Thresholds

#An interval describes a reference population

A reference interval describes a distribution of results in a selected reference population. Many conventional intervals contain the central 95% of that distribution, leaving about 5% outside by construction. This is not a claim that 5% of the population has a disease, and not every interval uses that exact convention.

The population and collection conditions matter. Selection of reference individuals, age groups, methods and statistical choices affects the limits. MedlinePlus explains that healthy people can have a result outside a reference range, while a person with a health problem can have a result inside it. A flag is information for interpretation, not a diagnosis.

Evidence: Ozarda: reference interval methods and limits / MedlinePlus: understanding lab results

#A decision limit answers a clinical question

A clinical decision limit is a threshold linked to a diagnostic question, risk or a proposed medical decision. It is not simply the high or low end of the distribution in a healthy reference group. Evidence from clinical studies and clinical guidance can support a decision limit; describing typical values in a population answers a different question.

Before interpreting a report, establish which kind of limit it presents. The same numerical format can hide that difference. This page gives no disease-specific cut-offs and does not say what action should follow a particular result. Those decisions depend on the test, purpose, clinical context and applicable guidance.

Evidence: Ozarda: reference interval methods and limits

#The method and units belong with the number

A value is not interpretable without its units and the relevant laboratory information. MedlinePlus notes that laboratories may use different methods and reference ranges, and that ranges can differ between groups such as children and adults. A range found elsewhere is not automatically the range for the report in front of you.

Transferring an interval also requires checking that it fits local testing. CLSI EP28IG describes verification of a preestablished interval from studies, literature or a developer for a laboratory's similar quantitative tests. Its public summary supports the need for verification, not any claim that Mynd has verified a laboratory or supplied a suitable interval.

Evidence: MedlinePlus: understanding lab results / CLSI EP28IG: verifying an interval for local use

#A population interval does not capture every change

People differ from one another, and results within one person also vary. Ozarda's methods review describes situations in which a person's usual values occupy only part of the wider population interval. A change can therefore need attention to previous results rather than a simple check of whether both values lie inside the printed range.

Measurement variation, collection conditions and biological variation all matter when assessing serial results. They do not turn every small difference into a meaningful clinical change. MedlinePlus recommends considering results with other health information. This page does not calculate a personal baseline, a reference change value or a treatment threshold.

Evidence: Ozarda: reference interval methods and limits / MedlinePlus: understanding lab results

Source note

The sections above were checked against the linked sources. No clinical review has been performed. This is general research education, not a clinical guideline.