Why the Same Blood Test at Different Labs Gives Different-Looking Results

A patient has an HbA1c of 6.9 at one lab and 7.2 at another the same week. Both labs are NABL-accredited. The doctor is asked which is right. The answer is that both probably are — but the reasons the numbers look different are worth understanding, because they show up on every patient's records eventually and often cause avoidable worry.

The five reasons lab values differ between labs

• Method differences. Two labs may use different analytical techniques for the same test, with slightly different sensitivities.
• Instrument calibration. Different lab equipment reads slightly differently even when using the same method.
• Reference ranges. Each lab establishes its own reference range based on its local population — the same value may be 'normal' at one lab and 'flagged high' at another.
• Sample handling. Time between draw and analysis, storage temperature, transport conditions all affect results.
• Biological variability. The same person's value can fluctuate by 5-10% day to day on many tests, independent of the lab.

The specific tests where differences are most pronounced

• Test: HbA1c · Typical inter-lab variation: ±0.3 percentage points · Note: Different methods (HPLC vs immunoassay)
• Test: TSH · Typical inter-lab variation: ±10-15% · Note: Different reference ranges especially
• Test: Vitamin D · Typical inter-lab variation: ±20% · Note: Notorious for inter-lab variability
• Test: Iron studies · Typical inter-lab variation: ±15% · Note: Multiple parameters, each with variation
• Test: Lipid profile · Typical inter-lab variation: ±5-10% · Note: Relatively consistent across labs
• Test: Creatinine · Typical inter-lab variation: ±5% · Note: Well-standardised globally
• Test: Basic haemogram · Typical inter-lab variation: ±3-5% · Note: Most consistent
• Test: Coagulation (PT/INR) · Typical inter-lab variation: ±10-15% · Note: Requires standardisation to reference thromboplastin

What to do about it

Three specific habits:

• For ongoing monitoring, stick with one lab. The absolute number matters less than the trend, and one-lab trends are cleaner.
• When you do use different labs, note which lab each result came from and interpret against that lab's range, not the other's.
• For clinically important decisions on borderline values, a repeat at the same lab a week or two later is often more informative than a chase across labs.

The reference range trap in detail

Reference ranges are not universal truths. They are statistical descriptions of what values look like in a specific population studied by the specific lab. Two implications:

• A value flagged 'High' at one lab may be 'Normal' at another, based purely on where the reference range was drawn.
• What is normal for one demographic may not be normal for another. Some Indian labs use ranges validated in Indian populations; some use imported ranges from other populations.

Do not obsess about the lab flag. Focus on the clinical significance of the value in your specific context, ideally as interpreted by your doctor.

The specific tests to be careful about

For a small number of tests, the choice of lab and method really matters:

• Thyroid function during pregnancy — narrow trimester-specific ranges; consistency of lab matters more.
• Anticoagulation monitoring (PT/INR) — international standards but lab differences still exist.
• Tumour markers post-cancer treatment — track at the same lab if possible.
• HIV viral load — the specific assay method matters.
• Cardiac troponin — highly sensitive assays differ meaningfully from older ones.

The one-question script for your doctor

Any time a lab result seems inconsistent with expectation:

• 'This result seems different from my last one. Is it worth repeating, or is it within the expected variation?'

A good doctor knows the specific variability for the specific test and will tell you whether to worry, wait, or retest. This is a specific piece of judgement that AI or lookup tools do not fully substitute for — the doctor's answer incorporates knowledge of your body and the lab's history.

References

• IFCC — lab standardisation for HbA1c
• NABL — India accreditation for labs
• PMC — inter-laboratory variation and clinical decisions