Reference ranges and verification in a small clinic lab

A result without a range is not a result

Haemoglobin of 11.8 means nothing on its own. For a 40-year-old man it is low; for a 30-year-old woman it is within range; for a child it depends on age. A lab report that prints a number without the range it should be read against has moved the interpretation onto whoever reads it next — usually a doctor who is looking at forty other things.

What ranges vary by

• Age. Continuous, not categorical. Neonatal, paediatric, adult and geriatric ranges differ for many analytes, and the boundaries are not round numbers.
• Sex. Haemoglobin, creatinine, ferritin, urate and others.
• Method and analyser. This is the one small labs most often get wrong. A range from a textbook may not apply to the assay you actually run. The correct source is your kit insert or your analyser's documentation.
• Physiological state. Pregnancy shifts several ranges materially.

Store ranges against the test with the age band, sex and method that produced them. A single global range per test is simpler and quietly wrong.

Flagging belongs in the system

Out-of-range values should be flagged at the moment of entry, automatically, against the range that applies to this patient's age and sex. Not by the technician reading the number and remembering.

Two levels are worth distinguishing:

• Out of range — flag it on the report. High or low, marked so a doctor's eye finds it.
• Critical — a value that needs somebody telephoned now, not collected tomorrow. Potassium of 6.9 is not an abnormal result, it is an event. Define your critical list, and define who makes the call.

Verification

Nothing leaves the lab until a qualified person has looked at it. That is both a regulatory expectation and the last defence against a transposed digit or a sample run on the wrong patient.

What verification means in practice:

1. The result is entered but not released.
2. A named person with the authority to verify reviews it.
3. They check it against the patient's context, not just the range — a haemoglobin of 8 in a known chronic case reads differently from one in a routine check.
4. On verification the report is released and frozen.

Why frozen matters

A verified report has left the building. Somebody may have acted on it. Editing it afterwards leaves no trace that a different number was once acted upon — which is exactly the record you need if a decision is later questioned.

So a correction is a new, dated, amended report that supersedes the first, with both retained. Same principle asa cancelled invoice keeping its number: the correction is recorded, not disguised.

What belongs on the printed report

• Patient name, age, sex and identifier
• Sample type, collection date and time, and receipt time
• Test, result, unit
• Reference range for this patient's age and sex
• A flag on anything out of range
• Method, where it affects interpretation
• Name and signature of the person who verified it
• Report date, and an amendment note if it is one

Turnaround, measured honestly

Measure from sample collection to report release, not from when the analyser finished. The gap between those two is where small labs lose most of their turnaround time, and it is almost always waiting for verification.

Clinikr resolves the reference range from the patient's age and sex at entry, flags out-of-range values automatically, and freezes a report on verification so a correction is a new dated document.