Meera had her thyroid test done at a neighbourhood diagnostics centre which showed TSH: 3.8 mIU/L. She consulted a doctor who said everything looked fine. Three months later, her company's panel hospital ran the same test and returned 4.8 mIU/L. Her HR department flagged it as abnormal. She spent two weeks convinced her thyroid had worsened. It had not.
This situation is common across India every day. Patients switch laboratories when they move cities, change insurance panels, or visit a different collection centre. The numbers that come back often look different, sometimes by enough to cause genuine alarm. What nobody explains is that in many such cases, the underlying biology may not have changed at all. What has changed is the measurement context: the laboratory platform, the reference boundary, the time of draw, or a combination of all three.
This article breaks down the five most important reasons blood test results differ between labs in India, the tests most vulnerable to variation, what clinical biochemists use to distinguish a true change from analytical noise, and a practical checklist to protect the accuracy of your next test.
When a laboratory prints a reference range next to your result, that range is not a universal biological standard. It is a statistical boundary, constructed per the CLSI guideline from a minimum of 120 apparently healthy reference individuals which projects values of 95% of the population. The lower bound of 120 subjects is the minimum needed to establish the 2.5th and 97.5th percentiles with 90 percent confidence intervals. The remaining 5 percent of entirely healthy individuals will fall outside this boundary by mathematical necessity, not because anything is wrong with them.
The critical issue: different laboratories use different reagent manufacturers. One NABL-certified lab might use a TSH reference range of 0.4 to 4.0 mIU/L. Another, running a different immunoassay platform, might use 0.3 to 5.5 mIU/L. TSH reference range differences of this magnitude are normal across certified platforms. Meera's result of 4.8 sat above one boundary and below another. The test did not change. The boundary did.
The Clinical Laboratory Standards Institute (CLSI) and the Indian Council of Medical Research both acknowledge that reference intervals should ideally be validated for local Indian populations. This is one of the least-discussed drivers of lab test variation in India. In routine practice, most labs inherit the ranges that came with their reagent kit, calibrated on a population with different demographics, different diet patterns, and different disease prevalence than urban or rural India.
This is a structural feature of laboratory medicine that every clinician and every patient should understand.
Most laboratory tests involve a chemical or immunological reaction. The outcome depends on the reagents that trigger reaction, temperature, what sequence the reagents are added, and on which analyzer platform the sample is run. The same molecule, measured by two chemically different routes, routinely produces two numerically different results. This is the core reason blood test results differ between labs even when both labs are fully accredited.
A College of American Pathologists (CAP) proficiency survey tested an identical standardised blood sample across more than 2,600 certified laboratories using 27 different NGSP-certified analyser methods. The true value was 6.00 percent HbA1c. Results ranged from approximately 5.6 percent to 6.4 percent. CAP accreditation permits HbA1c results to fall within a ±6 percent tolerance of the target value. All the labs that participated were certified and all the results were within acceptable limits. Yet the same sample spanned a range such that, a patient’s value near the diagnostic threshold, could mean the difference between a diabetes diagnosis and a normal result. This data is drawn from peer-reviewed proficiency testing literature and represents routine real-world variation.
Many Indian labs calculate LDL cholesterol rather than measure it directly. Two widely used calculation methods, the Friedewald equation and the Sampson-NIH equation, can produce meaningfully different values from the same triglyceride and cholesterol inputs. This is one of the more consequential lab test variations in India that goes unnoticed: in patients with elevated triglycerides or low cholesterol levels, Friedewald tends to underestimate LDL, potentially misclassifying a high-risk patient as borderline or normal. The same blood sample, calculated with different equations, ends up in different cardiovascular risk categorisation.
TSH and Free T3/T4: Multiple immunoassay platforms with documented differences of 20 to 30 percent on the same sample
Vitamin D (25-OH-D): Immunoassay and LC-MS/MS (mass spectrometry) methods vary significantly. A 2019 study documented inter-method differences of 10 to 40 percent
LDH: IFCC and DGKC methods use different reaction directions and produce different reference ranges. This means a result of 250 U/L on one platform is not equivalent to 250 U/L on another
Creatinine: The Jaffe kinetic and enzymatic methods differ by 10 to 20 percent on the same sample. This is clinically significant in Indian patients where lower muscle mass already tends to underestimate kidney function via eGFR calculations
Tumour markers (CEA, CA-125, PSA): No universal calibrator exists across manufacturers. Results from different platforms are not interchangeable and should never be compared directly for trend monitoring
The pre-analytical phase covers everything from the moment the blood is drawn to the moment analysis begins. A study done at a tertiary care hospital in India estimates that over 60 percent of laboratory errors happened in the pre-analytical phase. In the Indian context home blood collection, transport of samples to long distances, high staff turnover at collection centres are common. All of these can increase the risk of pre-analytical errors.
Blood glucose is often considered as a stable analyte, but this is only true when the sample is handled correctly. If the whole blood is not centrifuged and left at room temperature, a process called glycolysis continues after collection. Red blood cells and leucocytes present in the sample, keep metabolising glucose at a rate of approximately 5 to 7 percent per hour. A sample that is uncentrifuged for two to three hours in transit (common in home blood collection services), can lose 10 to 20 mg/dL of glucose. This can easily miss the diagnosis of gestational diabetes or falsely reassure a patient whose fasting glucose is borderline..
The correct procedure for glucose is in a sodium fluoride tube (grey cap), which inhibits glycolysis, followed by prompt centrifugation. When this does not happen it leads to a a false report.
Time of collection: TSH follows a diurnal rhythm, peaking in the early morning and falling in the afternoon. A TSH drawn at 7 am versus 11 am on the same day can legitimately differ by 0.5 to 1.0 mIU/L in healthy individuals.
Menstrual cycle phase: in women, TSH, LH, FSH, oestradiol, and progesterone all vary across the cycle. A result drawn on day 3 versus day 21 is not comparable
Biotin (Vitamin B7) supplementation: high-dose biotin supplements (above 5 mg/day, common in hair and skin products) directly interfere with biotin-streptavidin immunoassay platforms used for TSH, Free T4, troponin, and certain hormone tests. The interference falsely lowers TSH and falsely elevates Free T4, mimicking hyperthyroidism in a euthyroid patient. The US FDA has issued a safety advisory stating that patients should stop high-dose biotin for at least 72 hours before thyroid, hormone, or cardiac biomarker testing
Tourniquet application: A tourniquet held for more than 60 seconds causes haemoconcentration, falsely raising values of proteins, calcium, potassium, and certain enzymes
Fasting duration: Triglycerides are especially sensitive to food intake and can rise 20 to 50 percent within hours of eating; lipid panels, insulin, and glucose require verified fasting, not just self-reported fasting
The question is whether the difference between them reflects a true change in the patient's biology or whether it falls within the range of expected variation from the measurement process itself.
In routine laboratory practice, this is managed through a quality control tool called the delta check. Most hospital biochemistry laboratories, including NABL-accredited labs in India, run delta checks automatically as part of their laboratory information system. When a new result arrives for a patient who has been tested before, the system compares the new value against the previous one. If the difference exceeds a predefined threshold for that specific test, the result is cross-checked before it reaches the clinician, prompting the biochemist to review it for a possible sample mix-up, pre-analytical error, or genuine clinical change.
Delta check thresholds are test-specific and set by each laboratory based on the known biological and analytical variation of that analyte. For sodium, a sudden shift of more than 10 mmol/L between consecutive results in a stable patient would trigger an alert. For TSH, a jump from 2.0 to 6.0 mIU/L without a corresponding clinical change would prompt a review before the report is released.
The limitation that matters for patients comparing reports from two different laboratories is this: delta checks only work within a single laboratory's own records. When you take your previous report from Lab A to Lab B, Lab B has no access to your history. Its delta check system sees you as a new patient. The inter-laboratory difference is invisible to both quality control systems, which is precisely why the instruction to repeat at the same lab has a practical and scientific basis, it keeps your results within one system's quality net.
For tests where serial monitoring matters: HbA1c in diabetes, TSH in thyroid disease, tumour markers in cancer follow-up; staying with the same NABL-accredited laboratory means every new result is automatically compared against your personal history within a system that knows your baseline.
Not every test requires strict lab consistency. For a routine complete blood count or a urine routine examination, any NABL-accredited lab will typically give clinically equivalent results. The tests where TSH reference range differences, method variation, and pre-analytical errors are most likely to affect your care are those tied to ongoing treatment decisions or narrow diagnostic thresholds:
Before Your Test
Confirm fasting requirements: for lipid profiles and fasting glucose, verified 8 to 10 hours of fasting; for most other tests, ask specifically whether fasting is required
Stop high-dose biotin (Vitamin B7) supplements for at least 72 hours before thyroid, hormone, troponin, or fertility tests. This includes hair and skin supplements marketed for "biotin" even at lower doses
Book your thyroid test between 7 am and 9 am: TSH peaks in early morning; afternoon draws give consistently lower values; standardise the collection time for repeat monitoring
For women, note the day of your menstrual cycle on the requisition form for any hormonal panel (LH, FSH, oestradiol, progesterone, AMH); results vary by cycle phase and are only interpretable in that context
Choose and stay with one NABL-accredited laboratory for any test you monitor over time: HbA1c, TSH, tumour markers, Vitamin D, creatinine
At the Collection Centre
For home collection, confirm the collector is using a fluoride (grey cap) tube for glucose and that the sample will reach the lab within two hours; if uncertain, visit the lab directly for glucose testing
Tell the phlebotomist about all supplements and medications, including over-the-counter vitamins, herbal formulations, and biotin-containing products
Note the exact time of collection on your copy so your doctor can interpret time-sensitive tests correctly
When Reading Your Report
Always compare your result to the reference range on that specific report, not to range from a previous lab or values found on the internet
For follow-up tests, bring your previous reports to your doctor and let them assess whether any difference falls within expected analytical and biological variation
If a result is unexpectedly abnormal, repeat at the same NABL-accredited lab before acting rather than switching labs
A difference between labs is not itself an emergency. Seek immediate medical attention if:
Any result far outside the reference range on the new report, regardless of what the previous lab showed
A sharp rise in any tumour marker compared to your previous value, even if both values appear to fall within range
TSH below 0.1 or above 10 mIU/L, which requires urgent clinical review regardless of the platform used
Fasting glucose above 200 mg/dL on any certified platform
A creatinine value significantly higher than your previous reading, suggesting a possible acute change in kidney function
Any unexpected result accompanied by new or worsening symptoms: chest pain, breathlessness, unexplained weight loss, severe fatigue, or swelling
Do not adjust any medication, including thyroid, diabetes, or blood pressure medication, based on a result from a new laboratory without first discussing the change with your treating doctor or clinical biochemist.
Only if both labs are NGSP-certified or IFCC-aligned. The CAP proficiency data makes this concrete: even across certified labs, HbA1c can vary by ±0.4 percent on an identical sample. Ask your lab which certification standard their method holds before comparing values across reports.
My Vitamin D was 28 ng/mL at one lab and 19 ng/mL at another in the same month. Which is correct?
Both may be technically accurate on their respective platforms. Liquid chromatography tandem mass spectrometry (LC-MS/MS) is the reference method and gives lower values than immunoassay platforms, which are known to overestimate. Ask your lab which method they use, and standardise it to one lab for treatment monitoring.
Why did my blood glucose come back lower at the home collection than at the hospital lab?
This is a recognised pre-analytical issue. If your home collection sample was drawn into the wrong tube, or was not centrifuged and cooled promptly, continued glycolysis in the tube will have consumed glucose before testing. The lower result is an artefact. For any glucose test relevant to diabetes or gestational diabetes diagnosis, the sample should be collected in a sodium fluoride (grey cap) tube and reach the lab within two hours.
My TSH was normal in the morning and borderline at an afternoon test the next day. Is my thyroid problem worsening?
Almost certainly not. TSH follows a well-established diurnal rhythm, peaking between 2 am and 4 am and falling through the afternoon. A morning TSH of 3.5 mIU/L and an afternoon TSH of 4.8 mIU/L from the same person on consecutive days can both be within that person's normal biological range. This is a biological variation, not a pathological change. Standardise your thyroid tests to the same time of day, preferably 7 am to 9 am.
Is a NABL-accredited lab always better than a non-accredited one?
Accreditation guarantees documented procedures, calibrated equipment, maintained quality control records, and an audit trail. It does not guarantee any specific reagent brand or analyser generation. For complex, specialist, or treatment-monitoring tests, NABL accreditation is a meaningful quality indicator. For routine panels in an emergency, a well-run hospital lab may still be adequate. For any test where the result will influence a treatment decision, NABL accreditation should be a minimum requirement.
Important note: Patients with diabetes on medication, chronic kidney disease, thyroid disorders, cardiovascular disease, or a history of cancer should not adjust medications or draw clinical conclusions from a lab result difference without first consulting their treating physician. In these conditions, method variation can produce differences large enough to affect treatment decisions when misread in isolation. This article is intended as educational information for general readers and healthcare professionals. It does not constitute medical advice.
Clinical and Laboratory Standards Institute (CLSI). Defining, Establishing, and Verifying Reference Intervals in the Clinical Laboratory. EP28-A3c. CLSI; 2010. https://clsi.org/standards/products/method-evaluation/documents/ep28/
Farrell CL, Herrmann M. Determination of vitamin D and its metabolites. Best Pract Res Clin Endocrinol Metab. 2013;27(5):675-688. https://doi.org/10.1016/j.beem.2013.06.001
Little RR, Rohlfing CL. The long and winding road to optimal HbA1c measurement. Clin Chim Acta. 2013;418:63-71. https://pubmed.ncbi.nlm.nih.gov/23318564/
Sacks DB, et al. Guidelines and recommendations for laboratory analysis in the diagnosis and management of diabetes mellitus. Diabetes Care. 2023;46(10):e151-e199. https://doi.org/10.2337/dci23-0036
US FDA. Safety Communication: Update- The Importance of Discussing Biotin Supplements with Your Health Care Provider. FDA; 2019.
Tietz NW, ed. Tietz Textbook of Clinical Chemistry and Molecular Diagnostics. 6th ed. Elsevier; 2018.
Martin Bland J, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986;327(8476):307-310. https://doi.org/10.1016/S0140-6736(86)90837-8
