Biotin Interference: Link the Tested Concentration to the End-User Warning

Start with the tested concentration and its consequence

For a manufacturer preparing a biotin-technology immunoassay for a U.S. submission and an EU version, the immediate task is to connect the interference study to the information a laboratory will receive. FDA’s October 2020 final guidance recommends study designs similar to the most current CLSI EP07, evaluation up to 3500 ng/mL, and analyte levels near the device’s medical decision points. For susceptibility below that level, determine a concentration at which no interference is detected.

The guidance identifies percent difference or bias at each tested concentration and the consequence of interference, if observed, as relevant labeling information for qualitative and quantitative assays. The binding U.S. disclosure duties sit in 21 CFR 809.10. EU instructions for use and technical documentation have parallel interference-information duties under Regulation (EU) 2017/746 (IVDR). A shared study can support both files when it represents the relevant device configurations and intended uses; completing that study alone does not demonstrate conformity.

A practical claim map therefore needs more than a single tolerance number. It should show what was tested, what happened to the result, which claim the observation supports, and where that claim appears. The template below is an editorial synthesis for RA/QA teams to populate with their own evidence. It contains no simulated assay results or patient supplement-timing instructions.

Why the warning needs an assay-specific direction

In susceptible assays, free biotin in a specimen can compete with the assay’s biotinylated components for streptavidin binding. The January 2020 AACC guidance explains the usual consequences: sandwich assays can report falsely low results, while competitive assays can report falsely high results. These describe susceptible assay designs, rather than every assay that measures a particular analyte. Confirm the direction in the device’s own study.

That distinction changes the useful warning. A statement that biotin “may affect results” leaves the laboratory without the direction of error. A statement tied to measured conditions can explain whether the result may be falsely elevated or depressed and how that limits interpretation. Analyte name, reagent generation and platform matter when deciding whether published evidence is relevant to the claim.

Keep the concentration units visible throughout the map. Supplement intake is an amount per day, while interference testing concerns concentration in a specimen. A supplement dose is not an interchangeable substitute for a measured ng/mL value. Avoid turning an exposure example into an assurance that everyone taking that dose will have the same result.

Use the final FDA recommendation, with its limits

The October 2020 document finalizes the June 2019 draft and applies to IVDs using biotin technology, including relevant devices licensed under section 351 of the Public Health Service Act and used in donor screening. It expressly permits alternative approaches that satisfy applicable statutes and regulations. Describe its testing provisions as recommendations throughout the study rationale and labeling review.

• Design: use an interference study design similar to the most current CLSI EP07. Document the version and the study choices rather than claiming that the guidance supplies a complete assay-specific protocol.

• Biotin range: FDA recommends evaluation up to 3500 ng/mL, which it describes as three times the maximum expected clinical concentration. This is an evaluation level, not a universal tolerance claim or an upper limit on all possible exposure.

• Analyte levels: include samples near the device’s medical decision points. A result observed at one analyte concentration should not silently become a claim covering every decision point.

• Susceptibility: when interference occurs below 3500 ng/mL, determine a concentration at which it is not detected. Report that observation with its tested conditions and acceptance criterion.

The public CLSI EP07 catalog identifies the third edition, published April 2018 and reaffirmed October 2022. It describes screening, quantification and verification of interference, with recommended test-concentration information in EP37. The catalog also lists a practical EP07 Guide published in March 2026; that companion guide is not a new edition of the standard. The public catalog does not establish detailed mandatory protocol parameters.

Use exposure context to explain the study range

Exposure information helps explain why testing only around historical dietary levels can miss relevant interference. It does not supply a device-independent point at which results become unsafe. FDA reports serum or plasma levels below 1 ng/mL at a 30 µg daily intake, and observations above 1000 ng/mL with extremely high doses up to 300 mg/day. AACC distinguishes multivitamin intake from high-dose supplements. The following ladder keeps those observations separate from the study recommendation.

For study planning, keep the rationale for the selected biotin concentrations alongside the analyte levels. For labeling, keep the observed response alongside the concentration and specimen conditions. These two records answer different questions: why the experiment covered a range, and what a user can conclude from the results. Neither record should infer an individual patient’s biotin concentration from a supplement label.

Put U.S. disclosures in the appropriate insert sections

Section 809.10(b) specifies information and order for accompanying labeling, subject to its applicability provisions. The regulation’s general disclosure duties and the biotin guidance’s recommended detail should remain distinguishable. In particular, paragraph (b)(10) does not itself prescribe a numerical bias table or a universal biotin cutoff. The guidance’s footnote links its discussion to that paragraph’s duty to state known factors or substances affecting results.

A useful document review follows the same claim across these locations. If a performance summary says that a concentration was tested without detected interference, the limitations section should not imply that the device is unaffected at any concentration. If a known interferent is listed in specimen information, the insert should make its consequence understandable. Cross-references can help, provided the required information remains accessible and consistent.

Build the concentration-to-warning claim map

Create one row for each tested biotin concentration and relevant analyte/specimen condition, or link clearly to a result matrix containing those rows. Include the reagent generation and instrument configuration in the map’s scope. Separate observed effect from the conclusion under the protocol’s predefined acceptance criterion. The same percent change can have different significance near different medical decisions.

Every bracketed item below is a field to complete from the manufacturer’s study. The row roles illustrate how to organize the evidence; they do not prescribe intermediate concentrations, demonstrate interference, or provide ready-to-use labeling. Repeat rows for all additional concentrations and relevant conditions.

Add a traceable summary of the concentration at which no interference was detected when the assay is susceptible below 3500 ng/mL. Reference the supporting rows, specimen conditions, analyte levels and criterion. A discrete tested concentration is not an exact biological boundary. A passing result at the highest concentration also should not replace review of the other studied concentrations and decision points.

1. No detected interference within the studied conditions: limit the statement to those conditions and the evaluated range. Preserve the criterion used to reach that conclusion; avoid “biotin-proof” or an unqualified “no interference” claim.

2. Observed interference: communicate its measured size and direction and the supported limitation. Do not extrapolate one observation to every higher concentration or every assay on the platform.

3. Interference at clinically relevant analyte and biotin concentrations: FDA recommends contacting the appropriate CBER or CDRH review division. A completed warning does not by itself establish that the risk is adequately mitigated.

AACC notes that many manufacturers use ±10% deviation as an analytical-significance convention. It is not an FDA-mandated acceptance criterion. Explain the selected criterion’s relevance to the device’s decision points. For qualitative assays, describe the observed consequence for the reported result; a percentage alone can obscure whether a classification changes.

Check reagent-specific evidence before changing a claim

In the single-center 2025 study by Chiu and colleagues, measured tolerance differed by assay and reagent generation. Table 3 reported newer FT4, thyroglobulin and AFP reagent tolerances above 1500 ng/mL, compared with insert values at or below 1200 ng/mL, using a ±10% change definition. Those observations belong to the tested reagents and conditions, rather than all products from a manufacturer.

The newer anti-thyroglobulin row reports a measured category below 1250 ng/mL against an insert value at or below 1200 ng/mL. Those bounds do not alone prove failure below 1200 ng/mL. Likewise, improved tolerance in some assays does not establish that all existing warnings can be removed. Some experiments were performed only once because of limited sample volumes. This study provides a reason to verify a proposed claim against the relevant device configuration; it supplies neither a replacement tolerance for another assay nor a determination of the regulatory pathway for a labeling change.

Map the evidence separately to EU IVDR duties

For an EU version, review whether the U.S. study represents the EU device, intended purpose, specimen types and claims. A shared dataset can reduce duplicate work, but the legal mapping still needs its own assessment. The IVDR provisions below address performance, risk communication and documentation. They do not prescribe FDA’s 3500 ng/mL recommendation as an EU-wide numerical testing rule.

The useful cross-jurisdiction comparison is therefore a traceability exercise. The technical file holds the study detail; the IFU communicates performance and limitations needed by users. Review both against the actual EU configuration. A U.S. concentration rationale can inform that assessment, but it does not guarantee a conformity-assessment outcome. Document why the selected test range is adequate for the EU intended purpose and how remaining uncertainty is addressed.

Escalate the risk and keep the warning proportionate

FDA cautions that labeling alone may not sufficiently mitigate incorrect-result risk in every case. Its Federal Register notice mentions customer information letters and technical mitigations as options when the risk could significantly affect patient or public health, while declining to recommend specific mitigation strategies. Assess clinical relevance using the device’s analyte levels, biotin exposures and consequences of error. The guidance supplies no universal numerical escalation threshold.

FDA’s public troponin-interference list states that it identifies assays that have not addressed the risk. The live page, checked on October 22, 2026, reports continued receipt of adverse-event reports involving falsely low troponin results. It is not an incidence study, proof that every current reagent configuration malfunctions, or evidence of a particular commercial consequence. Use it as a risk communication and check device-specific labeling before drawing conclusions about a particular reagent generation.

Counter-evidence also needs its boundary. Vroemen and colleagues’ 2019 acute cardiac care evaluation found no detectable biotin-associated bias in Roche hs-cTnT measurements across 572 consecutive patients. Biotin was measured in a 100-patient sub-cohort, where the median was 331 ng/L, approximately 0.33 ng/mL. The investigators lacked supplement-intake information. This supports proportionate discussion of that cohort; it does not establish safety for high-dose users, other assays or later reagent generations.

Before finalizing the insert or IFU, compare every numerical statement with its source row. Confirm that the same reagent configuration, analyte level, specimen type and acceptance criterion support the claimed range. Resolve differences between the performance summary and limitation wording, and document the response when clinically relevant interference remains. The deliverable is a traceable concentration-to-warning argument that a reviewer and laboratory user can follow, with uncertainty retained wherever the study does not support a broader claim.