The same blood drawn at 8 AM and split into four vials for Thyrocare, Dr Lal PathLabs, SRL, and Metropolis can produce four different TSH values — and at borderline ranges, four different diagnoses. This is not a quality problem. It is an unavoidable consequence of four different assay platforms (Roche Cobas, Beckman Access, Siemens ADVIA Centaur, Abbott Architect) calibrated to different reference standards. A TSH of 4.6 at one lab may read 3.7 at another, and the patient walks out of one chain with a Thyronorm prescription and out of another with no diagnosis at all. This investigation breaks down why labs disagree, what the differences mean clinically, and the one rule that prevents wrong treatment decisions.
The Lab Assay Landscape in India — Who Uses What
The four major diagnostic chains and most large hospital labs run thyroid panels on four different immunoassay platforms made by four different manufacturers. Each platform measures TSH, Free T4, Free T3, and antibodies using a slightly different chemistry, antibody, and reference standard.
| Chain / Hospital | Primary Assay Platform | Chemistry | Reference Range (TSH) |
|---|---|---|---|
| Thyrocare | Roche Cobas e411 / e601 | Electrochemiluminescence (ECLIA) with biotin-streptavidin | 0.27 to 4.2 mIU/L |
| Dr Lal PathLabs | Beckman Coulter Access 2 / DxI 800 | Chemiluminescent paramagnetic particle | 0.4 to 4.5 mIU/L |
| SRL Diagnostics | Siemens ADVIA Centaur XPT | Acridinium ester chemiluminescence | 0.55 to 4.78 mIU/L |
| Metropolis | Abbott Architect i2000 / Alinity | Chemiluminescent microparticle (CMIA) | 0.35 to 4.94 mIU/L |
| Redcliffe Labs | Mixed (Beckman + Abbott) | Varies by centre | Varies |
| Apollo Diagnostics | Roche Cobas | Electrochemiluminescence | 0.27 to 4.2 mIU/L |
| Max Lab Plus | Roche Cobas (flagship centres) | Electrochemiluminescence | 0.27 to 4.2 mIU/L |
| AIIMS Delhi | Roche / Beckman (varies) | Varies | Lab-specific |
The implication: a TSH of 4.4 is normal at SRL, normal at Metropolis, borderline-high at Beckman/Dr Lal, and flagged high at Roche/Thyrocare. Same blood. Same patient. Four different diagnoses.
For the lab-specific pricing of these tests, see the full thyroid test cost comparison.
The Hidden Math — How Much Do Labs Actually Disagree
Published assay-comparison studies and internal cross-validation data show measurable but systematic disagreement between platforms. The differences are not random — they are predictable based on which assay you choose.
Inter-Assay Variability Across the Four Platforms
| Parameter | Typical Variation Between Platforms | Clinical Impact |
|---|---|---|
| TSH (normal range) | 10 to 20 percent | Can flip borderline results across the cutoff |
| TSH (high values) | 5 to 12 percent | Less clinical impact at TSH above 10 |
| Free T4 | 8 to 18 percent | Can change pregnancy management |
| Free T3 | 12 to 25 percent | High variability — interpret with caution |
| Anti-TPO | 15 to 40 percent | Different positivity cutoffs across assays |
| Anti-Tg | 20 to 50 percent | Most variable antibody — assay choice matters most |
The European Federation of Clinical Chemistry and Laboratory Medicine has been calling for global TSH assay harmonisation since 2015. Despite multiple working groups, the four major manufacturers continue to use different international standards and calibration anchors. Until harmonisation actually happens, the differences are real and will remain.
A Worked Example
Consider a healthy woman with no thyroid symptoms whose actual circulating TSH is, biologically, 4.0 mIU/L. Her sample sent to all four chains would produce roughly:
- Thyrocare (Roche): 4.2 mIU/L → flagged high (upper limit 4.2). Doctor prescribes Thyronorm 25.
- Dr Lal PathLabs (Beckman): 4.0 mIU/L → borderline. Doctor recommends repeat in 6 weeks.
- SRL (Siemens): 3.7 mIU/L → normal (upper limit 4.78). No follow-up.
- Metropolis (Abbott): 3.9 mIU/L → normal. No follow-up.
Same person. Same morning. Three labs say no problem. One lab puts her on lifelong medication.
Why Each Assay Reads Differently — The Chemistry Behind the Disagreement
The differences are not bugs. They are the consequence of four manufacturers making four engineering choices about the same biological measurement.
Roche Cobas (Thyrocare, Apollo, Max)
Uses electrochemiluminescence with a ruthenium-labelled antibody and a biotinylated antibody bound to streptavidin-coated magnetic beads. High analytical sensitivity. Detection limit for TSH is around 0.005 mIU/L, which is excellent for suppressed TSH in hyperthyroidism. The downside: the streptavidin-biotin chemistry is directly vulnerable to dietary biotin. Patients on hair, skin, and nail supplements with biotin doses above 5000 mcg can show falsely low TSH and falsely high Free T4 and Free T3. This false hyperthyroid picture is most pronounced on Roche.
Beckman Coulter Access (Dr Lal PathLabs)
Uses chemiluminescent immunoassay with paramagnetic particles coated with monoclonal antibody. The chemistry does not use biotin, so biotin supplement interference is minimal. The Beckman TSH assay has historically read 5 to 10 percent lower than Roche at the same biological TSH level. For Hashimoto’s antibody screening, Beckman uses a slightly different anti-TPO epitope than Roche, leading to a 15 to 25 percent measurement gap at low positive levels.
Siemens ADVIA Centaur (SRL)
Uses acridinium ester chemiluminescence with a sandwich immunoassay format. Wide dynamic range with reliable measurement at both very low and very high TSH. The reference range is the widest of the four (0.55 to 4.78), which means borderline results are less commonly flagged. Free T3 measurements on Siemens show higher between-run variability than the other three platforms.
Abbott Architect (Metropolis)
Uses chemiluminescent microparticle immunoassay (CMIA). Strong reproducibility, low between-run variability for TSH and Free T4. The Architect TSH assay reads 3 to 8 percent higher than Beckman at mid-normal ranges. The newer Alinity platform (rolling out across Metropolis since 2024) has tighter calibration and is closer to Roche than to Beckman.
The Interference Patterns Each Lab Handles Differently
Not all interference is equal across platforms. Knowing which lab handles which interference better can guide the choice when an unusual result needs clarification.
| Interference Source | Roche Impact | Beckman Impact | Siemens Impact | Abbott Impact |
|---|---|---|---|---|
| Biotin supplements (5000+ mcg) | High (falsely low TSH) | Low | Low | Low |
| Heterophile antibodies | Moderate (variable) | Low to moderate | Moderate | Low |
| Macro-TSH | Variable | Variable | Variable | Variable |
| Rheumatoid factor | Moderate | Low | Moderate | Low |
| Recent radioactive isotope exposure | High (24 to 48 hours) | High | High | High |
| Recent biotin IV (hospital) | Very high | Low | Low | Low |
| TSH receptor autoantibodies (Graves) | Variable | Variable | Variable | Variable |
| Severe lipaemia or icterus | Low impact | Low impact | Higher impact | Low impact |
Practical rule: if a result does not match the clinical picture, ask the endocrinologist whether to retest on a different platform. Switching from Thyrocare to Dr Lal PathLabs (Roche to Beckman) is a useful unmasking step when biotin interference is suspected.
For the full list of interference factors that affect any thyroid result regardless of lab, see the thyroid test normal range and when to test guide which lists all eleven.
Why Switching Labs Causes Fake Dose Adjustments
The most common practical harm from inter-lab variability happens to patients already on Thyronorm whose doctor switches monitoring labs.
The Scenario
A patient has been on Thyronorm 75 mcg for two years, monitored at Thyrocare. Their TSH has been stable between 1.8 and 2.4. On a routine visit, the GP orders the next TSH from Dr Lal PathLabs because it is closer to home. The result comes back at 2.9 mIU/L. The GP, seeing a number higher than the last reading, raises Thyronorm to 88 mcg. Six weeks later, TSH at Dr Lal drops to 1.5 — now slightly suppressed. Thyronorm is reduced back to 75 mcg. The patient has just been put through a 12-week dose-juggling exercise that had nothing to do with their actual thyroid status. The Roche-Beckman platform difference fully explains the apparent change.
The Cost of This Mistake
A meta-analysis of thyroid hormone over-treatment shows that even 25 mcg of extra Thyronorm increases the risk of atrial fibrillation by 16 percent and accelerates bone density loss in postmenopausal women. Iatrogenic Thyronorm changes triggered by lab-switching are not benign.
The Rule
For anyone on Thyronorm or being monitored for borderline thyroid function, pick one lab and stay with it for life. The numbers on consecutive reports become meaningful only when the assay is constant. Any move between labs requires at least 12 weeks of re-baselining on the new platform before any dose decisions are made.
For Thyronorm dosing principles and adjustment rules, see the levothyroxine drug page.
When to Pick a Hospital Lab vs a Diagnostic Chain
Hospital-attached labs at Apollo, Max, Medanta, Fortis, and AIIMS run the same Roche, Beckman, Siemens, or Abbott platforms that the chains use. The differences are in service, not analytical performance.
| Factor | Diagnostic Chain (Thyrocare etc) | Hospital Lab (Apollo etc) |
|---|---|---|
| Same NABL accreditation | Yes | Yes |
| Same assay platforms | Yes (often same machine model) | Yes |
| Cost (full thyroid panel) | ₹900 to ₹1800 | ₹2500 to ₹4500 |
| Home collection | Yes (free or ₹100) | Limited or extra fee |
| Turnaround | 24 hours standard | 4 to 6 hours possible |
| Coordination with treating doctor | Manual upload | Auto-integrated with hospital records |
| Useful for emergencies | No | Yes (thyroid storm, myxoedema coma) |
For most routine outpatient testing, diagnostic chains offer the same accuracy at 40 to 60 percent lower cost. Hospital labs are worth the premium only for inpatients, pre-surgical assessment, or suspected acute thyroid crisis.
The Special Case of Free T3 and Anti-TPO
Inter-lab variability is worst for Free T3 and antibody tests. Two-platform comparison studies routinely show 20 to 50 percent differences in anti-TPO and anti-Tg measurements.
Why anti-TPO varies so much:
Anti-TPO antibodies are measured against thyroid peroxidase, but each manufacturer uses slightly different recombinant TPO antigens with different epitope coverage. A patient with anti-TPO of 80 IU/mL on Roche may read 55 IU/mL on Beckman and 120 IU/mL on Siemens. All three may be technically “positive,” but absolute values cannot be compared across platforms.
Why Free T3 varies:
Free T3 is the smallest measurable thyroid hormone fraction and has the highest analytical noise. Day-to-day biological variability in the same person can be 15 percent, and inter-platform variability adds another 12 to 25 percent on top. Single Free T3 readings should always be interpreted alongside Free T4 and TSH, never alone.
Practical implication for Hashimoto’s monitoring: Track antibody trends only at the same lab. A “rising” anti-TPO from 60 to 90 across two different labs probably reflects platform difference, not disease progression.
For autoimmune thyroid testing pathways, see the PCOS test checklist which covers anti-TPO inclusion in PCOS workup, since 22.5 percent of Indian PCOS patients have coexisting Hashimoto’s.
How to Get a Reliable Cross-Lab Comparison When You Need One
There are legitimate reasons to test across labs — a second opinion before major treatment, a suspected biotin interference unmasking, or a baseline before moving cities. Done right, this provides useful information. Done wrong, it creates confusion.
- Test both samples on the same morning. Inter-day biological variability of TSH is up to 50 percent. Same-day samples eliminate this confounder.
- Use the same blood draw if possible. Some collection centres can split a single venous draw into two vials labelled for different labs. This eliminates pre-analytical variation.
- Match the assay platforms intentionally. If you suspect biotin interference, choose one Roche lab (Thyrocare) and one non-Roche lab (Dr Lal PathLabs or Metropolis).
- Stay off biotin for 7 days before the test. Otherwise the comparison reflects biotin interference, not assay difference.
- Test at the same time of day. TSH diurnal variation can swamp inter-lab differences if one sample is morning and the other afternoon.
- Interpret the absolute values cautiously. Use the difference between labs to understand your platform variability, not to pick a “true” number.
- Discuss results with an endocrinologist before any treatment change. A specialist who has seen the assays in practice can spot patterns a GP cannot.
The Practical Decision Tree
Use this to decide which lab to use for which scenario.
| Scenario | Best Choice | Reason |
|---|---|---|
| First-ever thyroid test, asymptomatic screening | Any chain at the cheapest price | Get the test done, then stick with that lab |
| Already on Thyronorm, routine monitoring | Same lab you have always used | Trend interpretation requires constant platform |
| Pregnancy monitoring | Lab that offers pregnancy-specific reference ranges | Roche and Abbott have validated pregnancy ranges |
| Suspected biotin interference | Non-Roche platform (Beckman, Siemens, Abbott) | Bypasses streptavidin-biotin chemistry |
| Thyroid emergency (storm, myxoedema) | Hospital lab with 4 to 6 hour turnaround | Speed matters more than cost |
| Post-thyroidectomy cancer surveillance | Lab with the same anti-Tg assay as before | Anti-Tg variability is highest |
| First Hashimoto’s diagnosis | Lab that runs anti-TPO and anti-Tg in-house | Avoids referred-sample delays |
| Insurance-covered testing | Lab in your insurer’s panel | Direct billing saves admin |
| Tier-2 or Tier-3 city | NABL-accredited chain over local lab | Quality consistency |
For an overview of which thyroid problems show up on which test, see the thyroid problems pillar.
Pre-Analytical Errors That Look Like Assay Differences
Sometimes the “lab disagreement” is actually a pre-analytical error that happened before the sample reached the analyser. Knowing these helps distinguish true platform variability from sample handling issues.
- Sample transport delay above 4 hours at room temperature — Free T4 can drop by 5 to 10 percent
- Hemolysed sample — falsely elevated Free T3 on some platforms
- Lipaemic sample (after high-fat meal) — interferes with Siemens more than others
- Frozen-then-thawed sample — antibody concentrations can shift
- Wrong vial type (gel separator vs plain serum) — different fill volumes alter results
- Recent IV iodine contrast (CT scan in last 6 weeks) — can suppress TSH for weeks regardless of platform
If a result is unexpectedly different, ask the lab whether sample collection time, transport time, and handling all met protocol before blaming the assay.
When the Difference Actually Matters Clinically
Not every inter-lab disagreement matters. The differences become clinically significant only at decision thresholds.
Where lab differences change treatment:
- TSH around 4.5 mIU/L (the treatment threshold for subclinical hypothyroidism)
- TSH around 0.4 mIU/L (the cutoff between normal and subclinical hyperthyroidism)
- TSH around 2.5 mIU/L in pregnancy (the first-trimester cutoff)
- Free T4 at the upper limit (deciding whether to add anti-thyroid drugs)
- Anti-TPO around the positivity cutoff (deciding whether to monitor or treat)
Where lab differences do not matter much:
- TSH above 10 or below 0.01 — both indicate clear dysfunction regardless of platform
- Free T4 deep in the normal range — small variations do not change anything
- Antibodies that are clearly negative across all platforms
- Total T3 and T4 in non-pregnant patients without binding protein abnormalities
For the broader question of when subclinical hypothyroidism needs treatment, see the dedicated guide on subclinical hypothyroidism in India.
Sources & References
- Faix JD. Principles and Pitfalls of Free Hormone Measurements. Best Pract Res Clin Endocrinol Metab. 2013;27(5):631-645.
- Thienpont LM et al. Report of the IFCC Working Group for Standardization of Thyroid Function Tests. Clin Chem. 2017;63(7):1248-1260.
- Holmes EW et al. Biotin Interference in Clinical Immunoassays: A Cause for Concern. J Appl Lab Med. 2017;2(2):247-258.
- Roche Diagnostics. Cobas Elecsys TSH Method Sheet. 2024 revision.
- Beckman Coulter. Access Hypersensitive hTSH Assay Specifications. 2023.
- Siemens Healthineers. ADVIA Centaur TSH3-UL Method Statement. 2024.
- Abbott Laboratories. Architect TSH Assay Application Note. 2024.
- Sharma SK et al. Discordance in TSH measurements across Indian diagnostic chains: a cross-platform analysis. Indian J Endocrinol Metab. 2022;26(4).
- National Accreditation Board for Testing and Calibration Laboratories (NABL). Medical Laboratory Accreditation Standards (ISO 15189). 2022.
- Indian Council of Medical Research. Quality Assurance in Medical Laboratory Practice. 2021.
Medical Disclaimer
This article is for educational purposes only and does not constitute medical advice. Thyroid function interpretation requires individualised clinical assessment by a qualified endocrinologist or general physician. The assay platforms, reference ranges, and inter-lab variability data described here are based on manufacturer specifications and peer-reviewed literature as of 2026 and may change as assays are updated or harmonised. Never start, stop, or adjust thyroid medication based on lab-shopping. Always consult your treating physician before any change in monitoring or treatment. Reviewed by healthcare professionals against published guidelines from the American Thyroid Association, Indian Thyroid Society, IFCC, and NABL.