# Biotin × Falsely low TSH measurement bias

Verdict 3106 · Evidence Check

Stored-material review date: 2026-10-05  
Original evidence date: 2026-09-15  
Review-criteria version: NGM-v3106-ORAL-BIOTIN-TSH-MEASUREMENT-1.0  
Public page/material preparation date: 2026-11-03

Clinical papers were not read directly; existing review summaries were checked. No new clinical-literature search was performed. Unverified does not mean that research is absent or that there is no effect.

## Assessment scope

- Population: Detailed direct evidence concerns healthy adults; no automatic extension to patients, renal impairment or pregnancy
- Ingredient, formulation and route: Biotin · Study-specific biotin supplements; tablet/capsule, exact composition and batch unverified · Oral
- Dose: 5 mg/day or 10 mg/day; separated by study, platform and blood sampling time
- Comparator: The same participants before exposure and after washout, and comparator platforms on specimens from the same time point; not presented as a randomized placebo trial
- Measured outcome: TSH assay bias relative to pre-exposure/post-washout measurements or a biotin-independent method; clinical reference-limit crossing considered separately
- Duration: 7 or 8 days of exposure; dosing duration and the interval from the last dose to blood sampling kept separate
- Questions outside this scope: Thyroid function/treatment effects, diagnostic accuracy/actual misdiagnosis/patient harm rates, all analytes/reagents, a universal stopping interval, and extensions to safe human doses/special populations are separate questions.

## Checklist items

### N01 · Does this table address a measurement question and prevent the application of efficacy grades and scores?

Verification status: Met  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: The original data have measurement_only=true, question_type=measurement. Efficacy grade·score·axes are not_applicable, and efficacy badges, aggregate ratings and a universal stopping interval are prohibited. The original grade ?·score null are not used as a new efficacy assessment.
- Why unverified: This documentation condition links the original measurement classification and display prohibitions unchanged. It does not mean certification of clinical efficacy or absence of interference.
- Reason for missing information: No new efficacy grade or aggregate measurement score is created.
- Required materials and conditions: Measurement-only classification, fields marking efficacy as not applicable, and a display contract prohibiting efficacy badges/aggregate ratings/a universal stopping interval
- Materials to obtain first: Measurement-only classification, fields marking efficacy as not applicable, and a display contract prohibiting efficacy badges/aggregate ratings/a universal stopping interval
- Further research: Cross-check the display paths for measurements, reagent performance, document quality and patient outcomes separately to prevent conflated labeling.
- Conditions for verification: Preserve measurement_only, the measurement type, the non-applicability of efficacy grade/score/axes, and the original display_contract unchanged.
- Why this condition is needed: This is needed to prevent falsely low measurements from being recast as treatment effects or an aggregate grade.
- Application exceptions: The original grade ?·score null and document quality B are historical values. Blue in the condition status indicates verification of a narrow documentation condition, not an efficacy badge.
- Remaining uncertainty: No new efficacy grade or aggregate measurement score is created. This status reflects the documented scope of the specific narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/personal safety, the actual misdiagnosis rate, or absence of evidence worldwide.
- Referenced methods: C01: Question clarification principles in M03 §3.1, §3.2.1–3.2.4; the detailed separation rules are internally designed / C09: M04 §4.5, §4.6.3; limits on the current data transformation follow the user-defined scope and are internally designed / D01: Chamgap's own diagnostic condition; M03 question/population principles serve only as general references. Reason for application to the measurement family: This is needed to prevent falsely low measurements from being recast as treatment effects or an aggregate grade. This is not a formal implementation of diagnostic/QUADAS/STARD·RoB/GRADE/causality methods or clinical certification.
- Related sources: 

### N02 · Which exposure, specimen, comparison and measurement outcome are brought together in one table?

Verification status: Met  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: The single claim concerns falsely low readings in specific serum TSH assays after oral biotin. The actual comparisons are within the same people before exposure and after washout, or between platforms using specimens from the same time point; they do not concern biotin's thyroid treatment effect or actual misdiagnosis rates.
- Why unverified: The original claim and the boundaries of the measurement family were verified. Results from some assays were not extended to all TSH assays or clinical diseases.
- Reason for missing information: Other analytes, actual changes in thyroid function, misdiagnosis/patient harm and a universal stopping interval are separate questions.
- Required materials and conditions: The original claim and a fixed question specified by healthy adults, oral exposure, serum TSH, platform/reagent, dose and blood sampling time
- Materials to obtain first: The original claim and a fixed question specified by healthy adults, oral exposure, serum TSH, platform/reagent, dose and blood sampling time
- Further research: Construct a matrix of platform, reagent, exposure and time point, and cross-check measurement bias separately from physiological changes/clinical outcomes.
- Conditions for verification: Fix the oral exposure, serum TSH, assay-specific falsely low readings and actual within-person/same-specimen comparisons, and distinguish questions that extend beyond them.
- Why this condition is needed: This is needed to avoid conflating thyroid disease, other analytes and all test interference with this TSH measurement bias.
- Application exceptions: Even when some concepts from the diagnostic accuracy family are used as references, the question is not changed into one about sensitivity/specificity or a clinical treatment trial.
- Remaining uncertainty: Other analytes, actual changes in thyroid function, misdiagnosis/patient harm and a universal stopping interval are separate questions. This status reflects the documented scope of the specific narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/personal safety, the actual misdiagnosis rate, or absence of evidence worldwide.
- Referenced methods: C01: Question clarification principles in M03 §3.1, §3.2.1–3.2.4; the detailed separation rules are internally designed / C04: M03 §3.2.1; detailed population conditions are internally designed / C05: M03 §3.2.2–3.2.3; reverse-direction labeling and adaptations for testing are internally designed / C06: M03 §3.2.4, M15 §15.3.1–15.3.3; question-specific outcome adaptations are internally designed / D01: Chamgap's own diagnostic condition; M03 question/population principles serve only as general references / D09: Chamgap's own boundary for diagnostic extensions; formal diagnostic methodology evidence has not been reviewed / E09: M03 §3.2.4; internally designed boundary for extensions. Reason for application to the measurement family: This is needed to avoid conflating thyroid disease, other analytes and all test interference with this TSH measurement bias. This is not a formal implementation of diagnostic/QUADAS/STARD·RoB/GRADE/causality methods or clinical certification.
- Related sources: S01, S04, S05, S08

### N03 · Which TSH measure in which specimen does the observed outcome represent?

Verification status: Met  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: The selected outcome is the serum TSH assay value. Li's TSH is in mIU/L and the Roche document's criteria are in µIU/mL; these are distinguished from blood biotin in pg/mL and document values in ng/mL. The original units were not converted and combined.
- Why unverified: The condition distinguishing the retained original units and analytes was verified. Physiological thyroid status was not determined.
- Reason for missing information: Conversions between original documents or application of patient diagnostic thresholds are not performed in this review.
- Required materials and conditions: Serum TSH analyte, units, time-point/contrast definitions, and separate blood biotin units
- Materials to obtain first: Serum TSH analyte, units, time-point/contrast definitions, and separate blood biotin units
- Further research: Cross-check analytes, units and measures against each original table/document without generating new converted values or clinical status classifications.
- Conditions for verification: Assign assay-specific serum TSH time-point means, formal contrasts and reference-limit crossings to their respective original units/time points.
- Why this condition is needed: This is needed to avoid substituting blood biotin concentrations, TSH measurements or disease status for one another.
- Application exceptions: Keep blood biotin in pg/mL and reagent-document values in ng/mL, and TSH in mIU/L and document values in µIU/mL, each in its original units. No new conversions are performed.
- Remaining uncertainty: Conversions between original documents or application of patient diagnostic thresholds are not performed in this review. This status reflects the documented scope of the specific narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/personal safety, the actual misdiagnosis rate, or absence of evidence worldwide.
- Referenced methods: C06: M03 §3.2.4, M15 §15.3.1–15.3.3; question-specific outcome adaptations are internally designed / D03: Chamgap's own diagnostic condition; formal diagnostic methodology evidence has not been reviewed. Reason for application to the measurement family: This is needed to avoid substituting blood biotin concentrations, TSH measurements or disease status for one another. This is not a formal implementation of diagnostic/QUADAS/STARD·RoB/GRADE/causality methods or clinical certification.
- Related sources: S01, S03, S08

### N04 · Are the actual form, batch and total exposure to the biotin supplement known?

Verification status: Unverified  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: Oral biotin supplement exposure is retained in the record, but the manufacturing source, exact chemical form, tablet/capsule, composition, batch, purity and total dietary exposure remain unverified.
- Why unverified: The currently retained information does not identify product details or total actual exposure. The nominal dose was not used to fill these gaps.
- Reason for missing information: Product specifications, adherence/intake records and dietary exposure for each trial.
- Required materials and conditions: Exact chemical form, formulation, batch, purity and dosing records for the study product, and total dietary exposure data
- Materials to obtain first: Exact chemical form, formulation, batch, purity and dosing records for the study product, and total dietary exposure data
- Further research: Link study product specifications and actual intake records to measurements by platform/time point to compare differences between products.
- Conditions for verification: Distinguish each study's supplement exposure from unverified product/dietary details.
- Why this condition is needed: This is needed to avoid extending the name biotin and the nominal dose to the composition, purity and actual intake of all supplements.
- Application exceptions: Unverified manufacturing source, chemical form, tablet/capsule, batch and dietary exposure do not constitute a finding that a particular product is defective.
- Remaining uncertainty: Product specifications, adherence/intake records and dietary exposure for each trial. This status reflects the documented scope of the specific narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/personal safety, the actual misdiagnosis rate, or absence of evidence worldwide.
- Referenced methods: C05: M03 §3.2.2–3.2.3; reverse-direction labeling and adaptations for testing are internally designed / H05: Design considerations in M19 §19.2.2–19.2.4; detailed confounding records are internally designed. Reason for application to the measurement family: This is needed to avoid extending the name biotin and the nominal dose to the composition, purity and actual intake of all supplements. This is not a formal implementation of diagnostic/QUADAS/STARD·RoB/GRADE/causality methods or clinical certification.
- Related sources: S01, S04, S05

### N05 · Were dosing duration and the interval from the last dose to blood sampling distinguished?

Verification status: Partly verified  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: Li used 10 mg/day for 7 days, with sampling approximately 2 hours after the last dose on day 7, and after 7 days off on day 14. Zhang used the same 10 participants at 5/10 mg/day for 7 days each; Ylli used 10 mg/day for 8 days, with measurements on day 1/day 8 before dosing and at 2 hours and 5 hours.
- Why unverified: The reported timelines were linked, but not all records of individual actual intake/timing and total dietary exposure were verified.
- Reason for missing information: Actual individual dosing times, adherence, total exposure and original blood sampling records for each trial.
- Required materials and conditions: Nominal oral dose, dosing start/end, actual last dose and blood sampling times for each trial
- Materials to obtain first: Nominal oral dose, dosing start/end, actual last dose and blood sampling times for each trial
- Further research: Cross-check dose, dosing duration, last dose and blood sampling time against the original records as separate axes.
- Conditions for verification: Retain the oral dose, dosing duration, assessment day and post-dose blood sampling time for Li, Zhang and Ylli separately by study.
- Why this condition is needed: This is needed to avoid treating 7-day/8-day exposure and blood sampling after 2 hours/5 hours/8 hours as the same timeline.
- Application exceptions: The tested 5/10 mg doses are not interpreted as safe doses, optimal doses or personal dosing/stopping instructions.
- Remaining uncertainty: Actual individual dosing times, adherence, total exposure and original blood sampling records for each trial. This status reflects the documented scope of the specific narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/personal safety, the actual misdiagnosis rate, or absence of evidence worldwide.
- Referenced methods: C05: M03 §3.2.2–3.2.3; reverse-direction labeling and adaptations for testing are internally designed / C06: M03 §3.2.4, M15 §15.3.1–15.3.3; question-specific outcome adaptations are internally designed / H02: Design-specific harm questions in M19 §19.2; details of temporal relationships are internally designed / E08: M03 §3.2.1–3.2.4; details of dose, duration and effect modification are internally designed. Reason for application to the measurement family: This is needed to avoid treating 7-day/8-day exposure and blood sampling after 2 hours/5 hours/8 hours as the same timeline. This is not a formal implementation of diagnostic/QUADAS/STARD·RoB/GRADE/causality methods or clinical certification.
- Related sources: S01, S04, S05

### N06 · To which populations can the measurements in healthy adults be applied?

Verification status: Partly verified  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: The detailed direct evidence from Li and Zhang concerns healthy adults. Ylli has an abstract-level record of 13 adults, while the detailed analysis population remains unverified. These data do not establish the magnitude/duration in patients, people with renal impairment, pregnant people, children or high-dose treatment groups.
- Why unverified: The scope of the reported population is available, but detailed recruitment/baseline characteristics and general applicability are currently unverified.
- Reason for missing information: Selection, exclusion and baseline status for each population, and direct measurement information for each special population.
- Required materials and conditions: Recruitment/selection, baseline thyroid/renal function, age/sex composition and special-population data for each trial
- Materials to obtain first: Recruitment/selection, baseline thyroid/renal function, age/sex composition and special-population data for each trial
- Further research: Cross-check population composition and baseline status against exposure/measurement conditions, and assess applicability to other populations using direct data from those populations.
- Conditions for verification: Distinguish each study's actually reported population from unverified selection/baseline characteristics and limitations concerning special populations.
- Why this condition is needed: This is needed to avoid extending a small sample of healthy people to all patients, people with renal impairment, pregnant people, children and high-dose treatment groups.
- Application exceptions: The description of participants as healthy adults alone does not verify recruitment representativeness, sex/age composition and exclusion criteria in full.
- Remaining uncertainty: Selection, exclusion and baseline status for each population, and direct measurement information for each special population. This status reflects the documented scope of the specific narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/personal safety, the actual misdiagnosis rate, or absence of evidence worldwide.
- Referenced methods: C04: M03 §3.2.1; detailed population conditions are internally designed / D02: Chamgap's own diagnostic sampling condition; formal diagnostic methodology evidence has not been reviewed / H08: M19 §19.1.2–19.2.4; internal application. Reason for application to the measurement family: This is needed to avoid extending a small sample of healthy people to all patients, people with renal impairment, pregnant people, children and high-dose treatment groups. This is not a formal implementation of diagnostic/QUADAS/STARD·RoB/GRADE/causality methods or clinical certification.
- Related sources: S01, S04, S05

### N07 · What denominators do Li's 6 people and 18 time-point specimens represent?

Verification status: Met  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: Li compared 6 healthy adults before exposure, on day 7 and on day 14. The 18 serum time-point specimens were not treated as 18 participants or 18 independent samples.
- Why unverified: The retained distinction between participant and repeated-specimen denominators was verified. This did not newly establish every detailed analysis denominator.
- Reason for missing information: Detailed flow of analyzable specimens and exclusions/missingness by platform/time point.
- Required materials and conditions: Li's participant count, time-point specimen count, and measurement/analysis and exclusion flow by platform
- Materials to obtain first: Li's participant count, time-point specimen count, and measurement/analysis and exclusion flow by platform
- Further research: Cross-check the correspondence among participants, specimens, platforms and time points against the original flow/tables to verify the unit of analysis.
- Conditions for verification: Distinguish repeated measurements within the same 6 people from the 18 time-point specimens, and retain gaps in the actual analysis flow.
- Why this condition is needed: This is needed to avoid overstating sample size and precision by counting repeated specimens as independent participants.
- Application exceptions: Verification of 18 specimens does not mean that missingness and exclusion flows were verified for every assay/time point.
- Remaining uncertainty: Detailed flow of analyzable specimens and exclusions/missingness by platform/time point. This status reflects the documented scope of the specific narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/personal safety, the actual misdiagnosis rate, or absence of evidence worldwide.
- Referenced methods: C07: M08 Table 8.2.a, M15 §15.3, M19 §19.1.2–19.2.4; design-specific adaptations are internally designed / D06: Chamgap's own diagnostic condition; formal diagnostic methodology evidence has not been reviewed. Reason for application to the measurement family: This is needed to avoid overstating sample size and precision by counting repeated specimens as independent participants. This is not a formal implementation of diagnostic/QUADAS/STARD·RoB/GRADE/causality methods or clinical certification.
- Related sources: S01

### N08 · Do Zhang's two doses represent trials with different groups of 10 people each?

Verification status: Partly verified  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: In Zhang, the same 10 healthy adults used 5 mg/day and 10 mg/day for 7 days each. The exact numbers of analyzed specimens by platform and time point remain unverified.
- Why unverified: The boundary of the same 10 people is clear, but detailed analyzed-specimen denominators remain unresolved.
- Reason for missing information: Exposure order/intervals and actual numbers of analyzed specimens for each dose, platform and time point.
- Required materials and conditions: Identity of the same participants, exposure order and duration, and analysis flow by dose/platform/time point
- Materials to obtain first: Identity of the same participants, exposure order and duration, and analysis flow by dose/platform/time point
- Further research: Cross-check exposure order, washout intervals and period effects within the same people to verify the dependence of repeated measurements.
- Conditions for verification: Retain the same 10 people, the fixed sequence of 7 days each, and the unverified numbers of analyzed specimens by platform/time point.
- Why this condition is needed: This is needed to avoid counting fixed-sequence exposures in the same participants as 20 people or two independent trials.
- Application exceptions: The results at the two doses are not recast as a randomized dose comparison or an estimate of interaction between doses.
- Remaining uncertainty: Exposure order/intervals and actual numbers of analyzed specimens for each dose, platform and time point. This status reflects the documented scope of the specific narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/personal safety, the actual misdiagnosis rate, or absence of evidence worldwide.
- Referenced methods: C02: Distinction between studies and reports in M04 §4.6.1–4.6.2; cohort/outcome linkage is an internal application / C07: M08 Table 8.2.a, M15 §15.3, M19 §19.1.2–19.2.4; design-specific adaptations are internally designed / D06: Chamgap's own diagnostic condition; formal diagnostic methodology evidence has not been reviewed. Reason for application to the measurement family: This is needed to avoid counting fixed-sequence exposures in the same participants as 20 people or two independent trials. This is not a formal implementation of diagnostic/QUADAS/STARD·RoB/GRADE/causality methods or clinical certification.
- Related sources: S04

### N09 · Do Ylli's 13 people constitute the analysis denominator for every TSH value?

Verification status: Partly verified  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: Ylli is recorded as having enrolled 13 adults. Extraction is limited to the abstract and figure captions; detailed TSH analysis denominators and quantitative full-text tables remain unverified.
- Why unverified: The enrollment count and level of material were retained, and the actual analysis n was not arbitrarily fixed at 13.
- Reason for missing information: Completion and analysis counts and quantitative tables by time point/assay.
- Required materials and conditions: Ylli's enrollment/completion/TSH analysis denominators, and specimen flow for each original quantitative table and figure
- Materials to obtain first: Ylli's enrollment/completion/TSH analysis denominators, and specimen flow for each original quantitative table and figure
- Further research: Map enrollment counts and analysis populations/time points to the original tables to verify the scope of missingness and exclusions.
- Conditions for verification: Distinguish the enrollment of 13 adults and the scope of the abstract/figure captions from the unverified detailed TSH analysis denominators.
- Why this condition is needed: This is needed to avoid automatically inserting the enrollment count in the abstract as the analysis denominator for full-text tables.
- Application exceptions: Retain the direction reported in the figure captions and abstract, without reconstructing quantitative full-text tables that could not be accessed.
- Remaining uncertainty: Completion and analysis counts and quantitative tables by time point/assay. This status reflects the documented scope of the specific narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/personal safety, the actual misdiagnosis rate, or absence of evidence worldwide.
- Referenced methods: C07: M08 Table 8.2.a, M15 §15.3, M19 §19.1.2–19.2.4; design-specific adaptations are internally designed / D06: Chamgap's own diagnostic condition; formal diagnostic methodology evidence has not been reviewed / C03: Data collection/selection principles in M04 §4.5–4.6; the separation of access and reading and the hyperlink contract are internally designed. Reason for application to the measurement family: This is needed to avoid automatically inserting the enrollment count in the abstract as the analysis denominator for full-text tables. This is not a formal implementation of diagnostic/QUADAS/STARD·RoB/GRADE/causality methods or clinical certification.
- Related sources: S05

### N10 · What comparisons does the actual design provide, and what does it fail to control?

Verification status: Partly verified  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: Li is a nonrandomized comparison within the same people before exposure, after exposure and after washout, and between platforms using the same specimens. Zhang uses a fixed sequence, and Ylli uses prospective within-person exposure comparisons. These were not treated as treatment effects between randomized placebo groups.
- Why unverified: The actual design descriptions are verified, but detailed controls for order, concurrent changes and blinding remain unverified.
- Reason for missing information: Specific original records of order, background/concurrent changes and reading/measurement procedures.
- Required materials and conditions: Actual exposure order, concurrent changes, control of background factors, and measurement/reading conditions
- Materials to obtain first: Actual exposure order, concurrent changes, control of background factors, and measurement/reading conditions
- Further research: Distinguish comparisons across time points within the same people from comparisons of concurrent specimens to assess the effects of order, time and background changes.
- Conditions for verification: Retain Li's nonrandomized repeated measurements, Zhang's fixed sequence and Ylli's prospective within-person comparison, along with limitations concerning time/concurrent changes.
- Why this condition is needed: This is needed to avoid presenting within-person exposure comparisons as randomized placebo efficacy trials.
- Application exceptions: Direct measurement results are not removed merely because the design is nonrandomized. The existence of a clinical RCT is not an essential prerequisite for measurement bias.
- Remaining uncertainty: Specific original records of order, background/concurrent changes and reading/measurement procedures. This status reflects the documented scope of the specific narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/personal safety, the actual misdiagnosis rate, or absence of evidence worldwide.
- Referenced methods: E01: M08 §8.1–8.2.3; internal application / E04: M08 §8.2.2, M15 §15.3.1; internal application / D05: Chamgap's own diagnostic condition; formal diagnostic methodology evidence has not been reviewed / H05: Design considerations in M19 §19.2.2–19.2.4; detailed confounding records are internally designed. Reason for application to the measurement family: This is needed to avoid presenting within-person exposure comparisons as randomized placebo efficacy trials. This is not a formal implementation of diagnostic/QUADAS/STARD·RoB/GRADE/causality methods or clinical certification.
- Related sources: S01, S04, S05

### N11 · What comparison references do pre-exposure/post-washout measurements and other platforms provide?

Verification status: Partly verified  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: The actual comparisons involve the same people before exposure and after washout, and platforms using specimens from the same time point. Comparisons with biotin-independent methods are relevant to verifying measurement interference, but are not diagnostic tests that establish a patient's thyroid status.
- Why unverified: The content of the comparisons is linked; independent reference methods/true values and diagnostic accuracy have not been verified in these materials.
- Reason for missing information: Separate information on calibration/traceability of comparator methods, independent reference methods and physiological true values.
- Required materials and conditions: Definitions of time contrasts, and analytical principles and traceability/reference-method information for comparisons using the same specimens
- Materials to obtain first: Definitions of time contrasts, and analytical principles and traceability/reference-method information for comparisons using the same specimens
- Further research: Cross-check paired time contrasts and assay differences in specimens from the same time point separately, and distinguish them from determining a clinical true value.
- Conditions for verification: Link within-person time contrasts and comparator platforms using specimens from the same time point separately, and explain the limitations of the reference methods.
- Why this condition is needed: This is needed to avoid confusing differences between assays with established diagnostic criteria for patients' thyroid conditions.
- Application exceptions: Even a biotin-independent method does not automatically establish an independent physiological TSH true value or a gold standard for clinical disease.
- Remaining uncertainty: Separate information on calibration/traceability of comparator methods, independent reference methods and physiological true values. This status reflects the documented scope of the specific narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/personal safety, the actual misdiagnosis rate, or absence of evidence worldwide.
- Referenced methods: C05: M03 §3.2.2–3.2.3; reverse-direction labeling and adaptations for testing are internally designed / C06: M03 §3.2.4, M15 §15.3.1–15.3.3; question-specific outcome adaptations are internally designed / D04: Chamgap's own diagnostic condition; formal diagnostic methodology evidence has not been reviewed / E04: M08 §8.2.2, M15 §15.3.1; internal application. Reason for application to the measurement family: This is needed to avoid confusing differences between assays with established diagnostic criteria for patients' thyroid conditions. This is not a formal implementation of diagnostic/QUADAS/STARD·RoB/GRADE/causality methods or clinical certification.
- Related sources: S01, S02, S04

### N12 · Were the principles and instrument identities of the TSH assays distinguished?

Verification status: Partly verified  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: Distinguish Roche cobas e602, Ortho VITROS 5600, Siemens Dimension Vista 1500, Abbott ARCHITECT 2000 and Zhang's Beckman UniCel DxI 800. Also retain Vista's biotin dependence and the biotin independence of Zhang's Beckman/Abbott TSH assays with their respective results.
- Why unverified: Differences in retained instrument identities and principles are available, but historical lots and actual reagent installations in individual laboratories have not been fully verified.
- Reason for missing information: Original procedures, reagent lots/versions and installation records for each assay.
- Required materials and conditions: Instruments, reagents, principles, specimen procedures and interference tolerance thresholds for each analyte
- Materials to obtain first: Instruments, reagents, principles, specimen procedures and interference tolerance thresholds for each analyte
- Further research: Cross-check not only instrument names but also the TSH reagents' capture/detection methods against the results at the exposure tested.
- Conditions for verification: Link the TSH results from Roche, Ortho, Siemens, Abbott and Beckman separately to the level of verification of their principles/versions.
- Why this condition is needed: This is needed to avoid treating interference on all instruments as identical solely because methods are described as sandwich or biotin-dependent.
- Application exceptions: Vista uses a biotin-dependent TSH assay, yet there was no significant difference at the exposure tested. A positive finding is not assigned on the basis of the principle alone.
- Remaining uncertainty: Original procedures, reagent lots/versions and installation records for each assay. This status reflects the documented scope of the specific narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/personal safety, the actual misdiagnosis rate, or absence of evidence worldwide.
- Referenced methods: C05: M03 §3.2.2–3.2.3; reverse-direction labeling and adaptations for testing are internally designed / C06: M03 §3.2.4, M15 §15.3.1–15.3.3; question-specific outcome adaptations are internally designed / D03: Chamgap's own diagnostic condition; formal diagnostic methodology evidence has not been reviewed / E03: M08 Table 8.2.a, M03 §3.2.4; internal application. Reason for application to the measurement family: This is needed to avoid treating interference on all instruments as identical solely because methods are described as sandwich or biotin-dependent. This is not a formal implementation of diagnostic/QUADAS/STARD·RoB/GRADE/causality methods or clinical certification.
- Related sources: S01, S02, S04

### N13 · Are historical tests and current reagents with the same instrument name being treated as identical?

Verification status: Unverified  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: Historical reagent lots and interference tolerance thresholds remain unverified. Roche's 2024-10 V2.0 document provides a known version boundary, but does not verify reagent identity at the time of the trials or currently across all laboratories.
- Why unverified: The required identity of lots/installations cannot be established from instrument names alone.
- Reason for missing information: Direct information on historical/current reagent lots, versions and laboratory installations.
- Required materials and conditions: Reagent lots and method versions at the time of the trial, revision histories and actual laboratory installation records
- Materials to obtain first: Reagent lots and method versions at the time of the trial, revision histories and actual laboratory installation records
- Further research: Map instruments, reagents, versions and installation dates to verify identity/differences between historical trials and current documents.
- Conditions for verification: Separate the unverified historical reagents/lots from the known identity of the 2024-10 V2.0 document.
- Why this condition is needed: This is needed to avoid automatically transferring interference in historical trials to revised reagents or the versions currently installed in all laboratories.
- Application exceptions: The existence of a Roche 2024 document does not verify actual installations in individual laboratories or equivalence to historical lots.
- Remaining uncertainty: Direct information on historical/current reagent lots, versions and laboratory installations. This status reflects the documented scope of the specific narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/personal safety, the actual misdiagnosis rate, or absence of evidence worldwide.
- Referenced methods: C05: M03 §3.2.2–3.2.3; reverse-direction labeling and adaptations for testing are internally designed / D03: Chamgap's own diagnostic condition; formal diagnostic methodology evidence has not been reviewed / D08: Chamgap's own diagnostic applicability condition; M04 principles distinguishing reports/studies serve as general references / C09: M04 §4.5, §4.6.3; limits on the current data transformation follow the user-defined scope and are internally designed. Reason for application to the measurement family: This is needed to avoid automatically transferring interference in historical trials to revised reagents or the versions currently installed in all laboratories. This is not a formal implementation of diagnostic/QUADAS/STARD·RoB/GRADE/causality methods or clinical certification.
- Related sources: S01, S04, S08

### N14 · Which repeated-measures contrast does Li's formal estimate represent?

Verification status: Met  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: Li's formal estimate is day 7 − mean(day 0, day 14). It is a repeated-measures contrast within the same participants, not a simple change from baseline or an effect between two randomized groups. The reported 95% CI is the interval from the original model.
- Why unverified: The formal contrast definition and attribution of the CI were verified. No new estimation or variance calculation was performed.
- Reason for missing information: The original model's detailed specifications and checks of adjustment/multiplicity remain separate outstanding items.
- Required materials and conditions: Formal repeated-measures contrast formula and model, and original estimates/CI/P for each platform
- Materials to obtain first: Formal repeated-measures contrast formula and model, and original estimates/CI/P for each platform
- Further research: Cross-check the original contrast formula and time-point means to distinguish formal model results from historical simple arithmetic values.
- Conditions for verification: Link the within-person repeated-measures contrast day 7 − mean(day 0, day 14) to the estimates for each platform.
- Why this condition is needed: This is needed to avoid confusing the simple difference obtained by subtracting the day 7 mean from the baseline mean with the formal contrast in the original model.
- Application exceptions: The formal 95% CI is not converted to SD or SE, and no model is newly fitted without the original data.
- Remaining uncertainty: The original model's detailed specifications and checks of adjustment/multiplicity remain separate outstanding items. This status reflects the documented scope of the specific narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/personal safety, the actual misdiagnosis rate, or absence of evidence worldwide.
- Referenced methods: C06: M03 §3.2.4, M15 §15.3.1–15.3.3; question-specific outcome adaptations are internally designed / E04: M08 §8.2.2, M15 §15.3.1; internal application / E05: M15 §15.3.1–15.3.3; internal application. Reason for application to the measurement family: This is needed to avoid confusing the simple difference obtained by subtracting the day 7 mean from the baseline mean with the formal contrast in the original model. This is not a formal implementation of diagnostic/QUADAS/STARD·RoB/GRADE/causality methods or clinical certification.
- Related sources: S01

### N15 · Were Li Roche's formal bias and historical simple difference distinguished?

Verification status: Met  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: Li Roche cobas e602 TSH was 1.80→1.21→2.06 mIU/L, with a formal contrast of −0.72 mIU/L, 95% CI −1.13~−0.32, P = 0.006. −0.59 mIU/L is the historical simple before/after difference between the displayed means; it was not newly calculated.
- Why unverified: The retained original Roche values and the distinction between the two contrasts were verified. They were not extended to current reagents or patient treatment effects.
- Reason for missing information: Detailed verification of historical lots, actual laboratory installations and the original model.
- Required materials and conditions: Li Roche TSH time-point means, formal contrast, original CI/P and display of historical arithmetic
- Materials to obtain first: Li Roche TSH time-point means, formal contrast, original CI/P and display of historical arithmetic
- Further research: Map platform/time-point means and the formal contrast definition to the original table to verify the roles of the different values.
- Conditions for verification: Link 1.80→1.21→2.06 and the formal −0.72/CI/P only to Roche cobas e602 and the corresponding time points.
- Why this condition is needed: This is needed to avoid treating values from different contrast definitions as an error requiring correction or as the same estimate.
- Application exceptions: −0.59 is the historical simple difference in the manuscript, not a new calculation in this review. It is not combined with the formal CI.
- Remaining uncertainty: Detailed verification of historical lots, actual laboratory installations and the original model. This status concerns the scope of documentation for that narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/individual safety, actual misdiagnosis rates or worldwide absence of evidence.
- Referenced methods: C06: M03 §3.2.4, M15 §15.3.1–15.3.3; question-specific outcome adaptations are internally designed / C07: M08 Table 8.2.a, M15 §15.3, M19 §19.1.2–19.2.4; design-specific adaptations are internally designed / E04: M08 §8.2.2, M15 §15.3.1; internal application / E05: M15 §15.3.1–15.3.3; internal application / C08: M04 §4.6.3, M15 §15.3.1–15.3.3; status separation is internally designed. Reason for application to the measurement family: This is needed to avoid treating values from different contrast definitions as an error requiring correction or as the same estimate. This is not a formal implementation of diagnostic/QUADAS/STARD·RoB/GRADE/causality methods or clinical certification.
- Related sources: S01

### N16 · What is the scope of Li's VITROS values and observations relative to the lower reference limit?

Verification status: Met  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: Ortho VITROS 5600 had means of 1.64→0.10→1.89 mIU/L, a formal contrast of −1.67 mIU/L, 95% CI −2.08~−1.26, P < 0.001. After exposure, all 6 participants were below 0.15 mIU/L, and the assay's lower reference limit was 0.47 mIU/L.
- Why unverified: The retained platform values and lower limit were linked accurately. They were not converted into actual hyperthyroidism or a misdiagnosis rate.
- Reason for missing information: Original individual values, reagent versions and separate patient-outcome data.
- Required materials and conditions: VITROS time-point means, formal contrast/CI/P and the method's lower reference limit
- Materials to obtain first: VITROS time-point means, formal contrast/CI/P and the method's lower reference limit
- Further research: Distinguish the original formal contrast from individual reference-limit crossings and attribute them to the platform's measurement results.
- Conditions for verification: Preserve the means, formal contrast, CI/P and original lower reference limit for Ortho VITROS 5600.
- Why this condition is needed: This is needed to avoid turning platform-specific bias magnitudes and lower-limit crossings into findings for other assays or clinical disease incidence.
- Application exceptions: The fact that all 6 participants were below the lower limit is an observation in a small measurement sample, not a population frequency.
- Remaining uncertainty: Original individual values, reagent versions and separate patient-outcome data. This status concerns the scope of documentation for that narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/individual safety, actual misdiagnosis rates or worldwide absence of evidence.
- Referenced methods: C06: M03 §3.2.4, M15 §15.3.1–15.3.3; outcome adaptations for each question are internally designed / E04: M08 §8.2.2, M15 §15.3.1; internal application / E05: M15 §15.3.1–15.3.3; internal application / D03: Chamgap's internally designed diagnostic condition; formal diagnostic-method evidence not reviewed. Reason for application to the measurement family: This is needed to avoid turning platform-specific bias magnitudes and lower-limit crossings into findings for other assays or clinical disease incidence. This is not formal implementation of diagnostic/QUADAS/STARD, RoB/GRADE or causality assessments, or clinical certification.
- Related sources: S01, S02

### N17 · Was crossing the lower reference limit expressed as sensitivity, specificity or an actual misdiagnosis rate?

Verification status: Met  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: The VITROS observation that all 6 participants were below the lower limit after exposure is a measurement under these conditions. It was not converted into a population misdiagnosis rate, incidence of actual hyperthyroidism or diagnostic sensitivity/specificity.
- Why unverified: This condition preserves the source's stated boundary for extending measurements to clinical outcomes. It does not mean there is no patient harm.
- Reason for missing information: Denominators and follow-up records for actual patient clinical assessments, actions and harms are separately needed.
- Required materials and conditions: The method's reference limit, individual measurements and actual clinical assessments/follow-up outcome data
- Materials to obtain first: The method's reference limit, individual measurements and actual clinical assessments/follow-up outcome data
- Further research: Compare assay reference-limit crossings, independent clinical status and post-test actions separately; do not estimate patient rates without that connection.
- Conditions for verification: Separate lower-limit crossings under the test conditions from clinical disease, misdiagnosis and harm rates, and diagnostic accuracy.
- Why this condition is needed: This is needed to distinguish an observation of assay values crossing the lower limit from the frequency of diagnoses/harms in patients.
- Application exceptions: A measurement-bias question does not unconditionally require disease TP/FP/FN/TN tables or clinical outcome trials.
- Remaining uncertainty: Denominators and follow-up records for actual patient clinical assessments, actions and harms are separately needed. This status concerns the scope of documentation for that narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/individual safety, actual misdiagnosis rates or worldwide absence of evidence.
- Referenced methods: D03: Chamgap's internally designed diagnostic condition; formal diagnostic-method evidence not reviewed / D07: Chamgap's internally designed diagnostic-outcome condition; formal diagnostic-method evidence not reviewed / D09: Chamgap's internally designed boundary for diagnostic extensions; formal diagnostic-method evidence not reviewed / H04: M19 §19.1.2–19.2.4; internal application / H09: Chamgap's internally designed interpretation boundary. References M19 principles; not implementation of a formal causality tool. Reason for application to the measurement family: This is needed to distinguish an observation of assay values crossing the lower limit from the frequency of diagnoses/harms in patients. This is not formal implementation of diagnostic/QUADAS/STARD, RoB/GRADE or causality assessments, or clinical certification.
- Related sources: S01, S02, S09

### N18 · Was the nonsignificant result for biotin-dependent Vista also preserved?

Verification status: Met  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: In the same Li specimens, Siemens Dimension Vista 1500 had −0.11 mIU/L, 95% CI −0.49~0.26, P = 0.56. Vista TSH is biotin-dependent, but there was no significant difference at this exposure. This was not interpreted as bias of 0 or equivalence.
- Why unverified: The difference between the original negative finding/interval and the principle was preserved as recorded.
- Reason for missing information: This observation does not establish results at other doses, blood-draw times or reagent versions.
- Required materials and conditions: Vista's original assay principle, paired estimate, CI/P and the exposure conditions
- Materials to obtain first: Vista's original assay principle, paired estimate, CI/P and the exposure conditions
- Further research: Relate the negative observation and the range of bias still compatible with it to the original interval/exposure conditions, distinguishing them from other platforms.
- Conditions for verification: Preserve the Vista formal contrast and CI/P from the same Li specimens together with its biotin-dependent principle.
- Why this condition is needed: This is needed to avoid deleting a negative result that appears inconsistent with the principle or converting it into an effect of 0.
- Application exceptions: P = 0.56 does not verify equivalence or absence of bias.
- Remaining uncertainty: This observation does not establish results at other doses, blood-draw times or reagent versions. This status concerns the scope of documentation for that narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/individual safety, actual misdiagnosis rates or worldwide absence of evidence.
- Referenced methods: C08: M04 §4.6.3, M15 §15.3.1–15.3.3; separation of statuses is internally designed / E05: M15 §15.3.1–15.3.3; internal application / D03: Chamgap's internally designed diagnostic condition; formal diagnostic-method evidence not reviewed. Reason for application to the measurement family: This is needed to avoid deleting a negative result that appears inconsistent with the principle or converting it into an effect of 0. This is not formal implementation of diagnostic/QUADAS/STARD, RoB/GRADE or causality assessments, or clinical certification.
- Related sources: S01, S02

### N19 · Was Li's nonsignificant Abbott estimate kept within the scope of the comparator method?

Verification status: Met  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: Li's Abbott ARCHITECT 2000 formal contrast was −0.06 mIU/L, 95% CI −0.44~0.31, P = 0.74. Nonsignificance was not converted into absence of bias, equivalence or absence of interference at other exposures.
- Why unverified: The retained Abbott values from the same specimens were linked to the comparator method's limitations.
- Reason for missing information: Other versions/exposures and independent standards/traceability for the assays.
- Required materials and conditions: Abbott's original comparator method, exposure conditions, paired estimate and CI/P
- Materials to obtain first: Abbott's original comparator method, exposure conditions, paired estimate and CI/P
- Further research: Verify the original results by distinguishing platform comparisons on the same specimens from time contrasts with pre-exposure/post-washout measurements.
- Conditions for verification: Attribute the Abbott ARCHITECT 2000 formal contrast, interval and P to the same Li specimens.
- Why this condition is needed: This is needed to avoid extending a negative observation on the comparator platform into a guarantee of no interference under all conditions.
- Application exceptions: Do not count another assay on the same specimens as an additional human trial or independent participants.
- Remaining uncertainty: Other versions/exposures and independent standards/traceability for the assays. This status concerns the scope of documentation for that narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/individual safety, actual misdiagnosis rates or worldwide absence of evidence.
- Referenced methods: C08: M04 §4.6.3, M15 §15.3.1–15.3.3; separation of statuses is internally designed / D04: Chamgap's internally designed diagnostic condition; formal diagnostic-method evidence not reviewed / E05: M15 §15.3.1–15.3.3; internal application. Reason for application to the measurement family: This is needed to avoid extending a negative observation on the comparator platform into a guarantee of no interference under all conditions. This is not formal implementation of diagnostic/QUADAS/STARD, RoB/GRADE or causality assessments, or clinical certification.
- Related sources: S01, S02

### N20 · At which doses and blood-draw times was lower Roche TSH observed in Zhang?

Verification status: Partly verified  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: In the same 10 participants, Roche cobas e602 TSH was lower with 5 mg at 2 hours, P = 0.005, and with 10 mg at 2 hours and 8 hours, each P < 0.001. Exact mean differences and CIs are unverified.
- Why unverified: Direction, dose/time point and P are available, but direct quantitative bias and intervals remain to be verified.
- Reason for missing information: Mean differences, CIs, actual analysis n by dose/blood-draw time and direct estimates between doses.
- Required materials and conditions: Zhang's original Roche mean differences, variances, CIs, denominators and reported P by time point/dose
- Materials to obtain first: Zhang's original Roche mean differences, variances, CIs, denominators and reported P by time point/dose
- Further research: Attribute results versus baseline at each dose to their time point/assay, and check whether separate direct estimates exist for comparisons between doses.
- Conditions for verification: Preserve the reported P and decreases by dose/time point in the same 10 participants, leaving unverified estimates unverified.
- Why this condition is needed: This is needed to avoid turning the direction associated with a P value into a quantitative bias magnitude or a direct effect between doses.
- Application exceptions: Exact mean differences and CIs are unverified. Do not extract new values from graphs or search summaries.
- Remaining uncertainty: Mean differences, CIs, actual analysis n by dose/blood-draw time and direct estimates between doses. This status concerns the scope of documentation for that narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/individual safety, actual misdiagnosis rates or worldwide absence of evidence.
- Referenced methods: C06: M03 §3.2.4, M15 §15.3.1–15.3.3; outcome adaptations for each question are internally designed / E05: M15 §15.3.1–15.3.3; internal application / E08: M03 §3.2.1–3.2.4; dose, duration and effect-modification details are internally designed / H02: design-specific harm questions in M19 §19.2; temporal-relationship details are internally designed. Reason for application to the measurement family: This is needed to avoid turning the direction associated with a P value into a quantitative bias magnitude or a direct effect between doses. This is not formal implementation of diagnostic/QUADAS/STARD, RoB/GRADE or causality assessments, or clinical certification.
- Related sources: S04

### N21 · Were Zhang's negative TSH observations on Beckman and Abbott preserved by analyte?

Verification status: Partly verified  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: No interference was observed under the tested conditions for Zhang's Beckman UniCel DxI 800 TSH and Abbott ARCHITECT 2000 TSH. These TSH methods did not use biotin-streptavidin, and interference affecting other thyroid analytes on Beckman was not transferred to TSH.
- Why unverified: Direction/principle is available for these TSH methods, but exact mean differences, CIs and historical reagent details are unverified.
- Reason for missing information: Quantitative results, CIs and reagent lots/versions for each TSH method.
- Required materials and conditions: Each TSH method's reagents/principle, original results by dose/time and separate tables for other analytes
- Materials to obtain first: Each TSH method's reagents/principle, original results by dose/time and separate tables for other analytes
- Further research: Hold both instrument and analyte fixed and separately compare negative TSH observations with interference results for other analytes.
- Conditions for verification: Distinguish the negative observations and principles for Beckman UniCel DxI 800 and Abbott ARCHITECT 2000 TSH.
- Why this condition is needed: This is needed to avoid transferring interference in other thyroid analytes on the same instrument to TSH results.
- Application exceptions: Nonuse of biotin-streptavidin is retained information about those TSH methods, not a guarantee for all instruments/versions.
- Remaining uncertainty: Quantitative results, CIs and reagent lots/versions for each TSH method. This status concerns the scope of documentation for that narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/individual safety, actual misdiagnosis rates or worldwide absence of evidence.
- Referenced methods: C08: M04 §4.6.3, M15 §15.3.1–15.3.3; separation of statuses is internally designed / C06: M03 §3.2.4, M15 §15.3.1–15.3.3; outcome adaptations for each question are internally designed / D03: Chamgap's internally designed diagnostic condition; formal diagnostic-method evidence not reviewed / D08: Chamgap's internally designed diagnostic-applicability condition; general reference to M04 principles for distinguishing reports/studies. Reason for application to the measurement family: This is needed to avoid transferring interference in other thyroid analytes on the same instrument to TSH results. This is not formal implementation of diagnostic/QUADAS/STARD, RoB/GRADE or causality assessments, or clinical certification.
- Related sources: S04

### N22 · Was nonsignificance at 26 hours and 50 hours converted into a universal stopping interval?

Verification status: Partly verified  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: Zhang records that no significant difference versus baseline was reported at 26 hours and 50 hours for either dose. Exact mean differences and CIs are unverified, and this is not interpreted as a universal supplement-stopping interval.
- Why unverified: Results at the recorded later time points are available, but do not establish bias of 0 or a universal recovery time.
- Reason for missing information: Quantitative intervals and criteria for absence of interference at each time point, and duration data by population/assay.
- Required materials and conditions: Original results/intervals by blood-draw time, actual dosing/stopping times and duration information by assay/population
- Materials to obtain first: Original results/intervals by blood-draw time, actual dosing/stopping times and duration information by assay/population
- Further research: Distinguish observations at blood-draw times from criteria for absence of interference, and compare actual persistence/resolution with the original intervals and criteria for those conditions.
- Conditions for verification: Preserve reports of nonsignificance versus baseline at 26/50 hours for both doses only within the original conditions.
- Why this condition is needed: This is needed to avoid extending nonsignificance at tested time points into the time when interference ends for all people, doses and assays.
- Application exceptions: Time points after exposure stops are not recommendations for individuals to stop taking the supplement. Nonsignificance is not equivalence either.
- Remaining uncertainty: Quantitative intervals and criteria for absence of interference at each time point, and duration data by population/assay. This status concerns the scope of documentation for that narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/individual safety, actual misdiagnosis rates or worldwide absence of evidence.
- Referenced methods: C08: M04 §4.6.3, M15 §15.3.1–15.3.3; separation of statuses is internally designed / E08: M03 §3.2.1–3.2.4; dose, duration and effect-modification details are internally designed / H02: design-specific harm questions in M19 §19.2; temporal-relationship details are internally designed. Reason for application to the measurement family: This is needed to avoid extending nonsignificance at tested time points into the time when interference ends for all people, doses and assays. This is not formal implementation of diagnostic/QUADAS/STARD, RoB/GRADE or causality assessments, or clinical certification.
- Related sources: S04, S09

### N23 · Was the direction corroborated by Ylli distinguished from unverified quantitative tables?

Verification status: Partly verified  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: Ylli reported 13 adults, 10 mg/day for 8 days, and measurements on day 1/day 8 before dosing, at 2 hours and at 5 hours. The abstract supports Roche changes and the greatest interference at 2 hours, but full-text quantitative bias is unverified, and percentages from searches were not imported.
- Why unverified: The abstract's direction and timeline are available, but quantitative tables, exact analysis denominators/CIs remain to be verified.
- Reason for missing information: Quantitative bias, analysis n and CIs from the original full-text tables.
- Required materials and conditions: Ylli's original full-text quantitative tables, TSH analysis n/differences/CIs by time point and the abstract's role
- Materials to obtain first: Ylli's original full-text quantitative tables, TSH analysis n/differences/CIs by time point and the abstract's role
- Further research: Match the abstract/figure captions to the original tables to distinguish the level of directional verification from numerical verification.
- Conditions for verification: Record 10 mg/day, 8 days, day 1/day 8, predose/2/5 hours and the Roche direction within the scope of the abstract.
- Why this condition is needed: This is needed to avoid filling in the direction in the abstract/figure captions with percentages/bias magnitudes from full-text tables.
- Application exceptions: Do not import percentages from search summaries. The existence of a PMCID does not mean the full text was read.
- Remaining uncertainty: Quantitative bias, analysis n and CIs from the original full-text tables. This status concerns the scope of documentation for that narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/individual safety, actual misdiagnosis rates or worldwide absence of evidence.
- Referenced methods: C03: data collection/selection principles in M04 §4.5–4.6; separation of access from reading and the hyperlink contract are internally designed / C06: M03 §3.2.4, M15 §15.3.1–15.3.3; outcome adaptations for each question are internally designed / E05: M15 §15.3.1–15.3.3; internal application / D08: Chamgap's internally designed diagnostic-applicability condition; general reference to M04 principles for distinguishing reports/studies. Reason for application to the measurement family: This is needed to avoid filling in the direction in the abstract/figure captions with percentages/bias magnitudes from full-text tables. This is not formal implementation of diagnostic/QUADAS/STARD, RoB/GRADE or causality assessments, or clinical certification.
- Related sources: S05

### N24 · What are the exact version, product and platforms of the Roche method document?

Verification status: Met  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: The retained document is Roche Elecsys TSH US 2024-10 V2.0, document 08429324501V2.0, product ref 08429324160. It describes that reagent's performance on cobas e411/e601/e602 and does not verify current laboratory installations.
- Why unverified: The retained document's identity and scope of applicability were linked. Historical/current reagent equivalence or installation was not certified.
- Reason for missing information: Historical lots and current laboratory installation/revision histories.
- Required materials and conditions: Roche method document version, document/product identifiers, supported platforms and installation/revision data
- Materials to obtain first: Roche method document version, document/product identifiers, supported platforms and installation/revision data
- Further research: Link document identifiers to applicable platforms, and separately compare whether the versions at the time of testing/currently installed match.
- Conditions for verification: Preserve 2024-10 V2.0, 08429324501V2.0, 08429324160 and the e411/e601/e602 scope exactly as recorded.
- Why this condition is needed: This is needed to avoid using a specific version of the method document as evidence of identical performance for all historical Roche reagents.
- Application exceptions: The currently retained document version and the versions actually installed in all laboratories are different facts.
- Remaining uncertainty: Historical lots and current laboratory installation/revision histories. This status concerns the scope of documentation for that narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/individual safety, actual misdiagnosis rates or worldwide absence of evidence.
- Referenced methods: C05: M03 §3.2.2–3.2.3; reverse-direction labeling and adaptations for testing are internally designed / D03: Chamgap's internally designed diagnostic condition; formal diagnostic-method evidence not reviewed / D08: Chamgap's internally designed diagnostic-applicability condition; general reference to M04 principles for distinguishing reports/studies. Reason for application to the measurement family: This is needed to avoid using a specific version of the method document as evidence of identical performance for all historical Roche reagents. This is not formal implementation of diagnostic/QUADAS/STARD, RoB/GRADE or causality assessments, or clinical certification.
- Related sources: S08

### N25 · What do the document's biotin test range and allowable deviations mean?

Verification status: Met  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: The Roche document specifies a test range of biotin ≤1200 ng/mL. At TSH ≤0.2 µIU/mL, the allowable deviation is ≤0.02 µIU/mL; above that, it is ≤10%. This is an analytical performance document, not an oral safe dose or a guarantee of no interference in all laboratories.
- Why unverified: The original analytical acceptance values were accurately linked to their nonclinical scope.
- Reason for missing information: Direct comparison of actual laboratory installations/exposure conditions with human applicability.
- Required materials and conditions: The method document's test concentrations, allowable deviations, validation conditions and revision/installation information
- Materials to obtain first: The method document's test concentrations, allowable deviations, validation conditions and revision/installation information
- Further research: Compare the document's concentrations, TSH ranges and allowable deviations, and keep human oral-exposure/clinical-safety questions separate.
- Conditions for verification: Record ≤1200 ng/mL, an allowable deviation at TSH ≤0.2 µIU/mL of ≤0.02 µIU/mL, and ≤10% at higher TSH within the document's scope.
- Why this condition is needed: This is needed to avoid turning an analytical validation concentration into an oral safe dose or a guarantee of bias of 0.
- Application exceptions: Do not make new conversions between document concentrations and intake amounts. Allowable deviations differ from the absolute absence of interference.
- Remaining uncertainty: Direct comparison of actual laboratory installations/exposure conditions with human applicability. This status concerns the scope of documentation for that narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/individual safety, actual misdiagnosis rates or worldwide absence of evidence.
- Referenced methods: C06: M03 §3.2.4, M15 §15.3.1–15.3.3; outcome adaptations for each question are internally designed / D03: Chamgap's internally designed diagnostic condition; formal diagnostic-method evidence not reviewed / E05: M15 §15.3.1–15.3.3; internal application / H09: Chamgap's internally designed interpretation boundary. References M19 principles; not implementation of a formal causality tool. Reason for application to the measurement family: This is needed to avoid turning an analytical validation concentration into an oral safe dose or a guarantee of bias of 0. This is not formal implementation of diagnostic/QUADAS/STARD, RoB/GRADE or causality assessments, or clinical certification.
- Related sources: S08

### N26 · Was Li's published unit correction applied to the original units?

Verification status: Met  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: Under the 2017-11-21 correction, Li's serum-biotin unit was retained as pg/mL. TSH in mIU/L was unchanged. The S03 correction belongs to the same Li report family and is not an independent human trial.
- Why unverified: This is a documentation condition linking the retained unit correction to its applicability.
- Reason for missing information: No new search was conducted in this review for additional corrections/retractions after the reference date.
- Required materials and conditions: Identity and original wording of the published correction, units before/after correction and locations where it applies
- Materials to obtain first: Identity and original wording of the published correction, units before/after correction and locations where it applies
- Further research: Compare the correction with its linked numbers/units, and do not change values/units outside its applicability.
- Conditions for verification: Under the 2017-11-21 correction, distinguish biotin in pg/mL from TSH in mIU/L, which was unchanged.
- Why this condition is needed: This is needed to avoid repeating the serum-biotin unit error in citations or also changing TSH units.
- Application exceptions: Preserve the correction's original bibliographic information, including PMID/DOI. This is not a new external correction check or numerical conversion in this review.
- Remaining uncertainty: No new search was conducted in this review for additional corrections/retractions after the reference date. This status concerns the scope of documentation for that narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/individual safety, actual misdiagnosis rates or worldwide absence of evidence.
- Referenced methods: C08: M04 §4.6.3, M15 §15.3.1–15.3.3; separation of statuses is internally designed / C09: M04 §4.5, §4.6.3; limitations on the current data transformation follow the user's scope and internal design / C06: M03 §3.2.4, M15 §15.3.1–15.3.3; outcome adaptations for each question are internally designed. Reason for application to the measurement family: This is needed to avoid repeating the serum-biotin unit error in citations or also changing TSH units. This is not formal implementation of diagnostic/QUADAS/STARD, RoB/GRADE or causality assessments, or clinical certification.
- Related sources: S01, S03

### N27 · Is day 7 biotin of 3601 an actual precisely measured concentration?

Verification status: Met  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: The day 7 value of 3601 was assigned for analysis because all measurements exceeded 3600. It was not interpreted as an exact measured serum-biotin concentration or an individual dose-concentration relationship.
- Why unverified: The meaning of the retained upper-limit exceedance and assigned analysis value was verified.
- Reason for missing information: Exact concentrations above the upper limit and individual concentration-bias relationships are unverified.
- Required materials and conditions: Serum-biotin measurement range, handling of values above the upper limit and the actual analytical assignment rule
- Materials to obtain first: Serum-biotin measurement range, handling of values above the upper limit and the actual analytical assignment rule
- Further research: Compare the original detection range and assignment handling to distinguish actual observations, values above the upper limit and model inputs.
- Conditions for verification: Preserve the original statement that all day 7 measurements exceeded 3600 and were assigned 3601.
- Why this condition is needed: This is needed to avoid using an assigned analysis value above the upper detection limit as an exact serum concentration or dose-concentration relationship.
- Application exceptions: Do not convert 3601 and 3600 to other units or calculate new concentration/bias functions.
- Remaining uncertainty: Exact concentrations above the upper limit and individual concentration-bias relationships are unverified. This status concerns the scope of documentation for that narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/individual safety, actual misdiagnosis rates or worldwide absence of evidence.
- Referenced methods: C06: M03 §3.2.4, M15 §15.3.1–15.3.3; outcome adaptations for each question are internally designed / E05: M15 §15.3.1–15.3.3; internal application / C07: M08 Table 8.2.a, M15 §15.3, M19 §19.1.2–19.2.4; design-specific adaptations are internally designed. Reason for application to the measurement family: This is needed to avoid using an assigned analysis value above the upper detection limit as an exact serum concentration or dose-concentration relationship. This is not formal implementation of diagnostic/QUADAS/STARD, RoB/GRADE or causality assessments, or clinical certification.
- Related sources: S01, S03

### N28 · Are results for other thyroid analytes/nonhormone analytes applied to this TSH question?

Verification status: Not applicable  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: Interference in other thyroid analytes on Beckman was not transferred to TSH. Nonhormone assay results in the title of the Li paper and the FDA troponin safety document are also not combined as direct numerical evidence of this TSH bias.
- Why unverified: Direct assessment of other analytes is outside the selected TSH question. This status does not convert Not read into Not applicable.
- Reason for missing information: Effects, harms and accuracy for other analytes must be reviewed with separate boundaries and direct data.
- Required materials and conditions: Analyte-specific result tables within the paper/instrument and the original selected-endpoint boundary
- Materials to obtain first: Analyte-specific result tables within the paper/instrument and the original selected-endpoint boundary
- Further research: Separate results by analyte-platform-exposure-time point to prevent transferring other endpoints that share the same instrument name.
- Conditions for verification: Limit the currently selected outcome to TSH and leave extensions to other analytes/treatment/patient outcomes as separate questions.
- Why this condition is needed: This is needed to avoid combining interference in other analytes from the same paper or instrument into TSH bias.
- Application exceptions: Do not delete other results or judge them invalid. This exception means they are outside the scope of this standalone table.
- Remaining uncertainty: Effects, harms and accuracy for other analytes must be reviewed with separate boundaries and direct data. This status concerns the scope of documentation for that narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/individual safety, actual misdiagnosis rates or worldwide absence of evidence.
- Referenced methods: C01: question-clarification principles in M03 §3.1, §3.2.1–3.2.4; detailed separation rules are internally designed / C06: M03 §3.2.4, M15 §15.3.1–15.3.3; outcome adaptations for each question are internally designed / E09: M03 §3.2.4; internally designed boundary for extensions / D09: Chamgap's internally designed boundary for diagnostic extensions; formal diagnostic-method evidence not reviewed. Reason for application to the measurement family: This is needed to avoid combining interference in other analytes from the same paper or instrument into TSH bias. This is not formal implementation of diagnostic/QUADAS/STARD, RoB/GRADE or causality assessments, or clinical certification.
- Related sources: S01, S04, S07

### N29 · Can measurements/analyses and missingness be linked by platform and time point?

Verification status: Unverified  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: Li's 6 participants and 18 time-point specimens are retained. Zhang's exact analyzed-specimen counts by platform/time point and Ylli's detailed TSH analysis denominators are unverified. The current data were not used to create claims of no missingness/exclusions or a common analysis n.
- Why unverified: Original data are insufficient to fully link the analysis flow for each platform/time point.
- Reason for missing information: Original flow of successful measurements, analyses and reasons for missingness/exclusion by time point/platform.
- Required materials and conditions: Participant→blood draw→specimen→platform measurement→analysis flow and handling of indeterminate/excluded/missing results
- Materials to obtain first: Participant→blood draw→specimen→platform measurement→analysis flow and handling of indeterminate/excluded/missing results
- Further research: Compare the actual measurement/analysis flow by platform and time point, and do not arbitrarily fill missing results with normal/0.
- Conditions for verification: Separate currently known participants/specimens from the unverified platform-specific analysis flow.
- Why this condition is needed: This is needed to avoid substituting enrollment/participant counts or time-point specimen counts as the actual denominator for every analysis.
- Application exceptions: Do not present the absence of a detailed flow as 0 missing results or no exclusions.
- Remaining uncertainty: Original flow of successful measurements, analyses and reasons for missingness/exclusion by time point/platform. This status concerns the scope of documentation for that narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/individual safety, actual misdiagnosis rates or worldwide absence of evidence.
- Referenced methods: C07: M08 Table 8.2.a, M15 §15.3, M19 §19.1.2–19.2.4; design-specific adaptations are internally designed / D06: Chamgap's internally designed diagnostic condition; formal diagnostic-method evidence not reviewed / H06: M19 §19.1.2–19.2.4, M04 §4.6.3; internal application. Reason for application to the measurement family: This is needed to avoid substituting enrollment/participant counts or time-point specimen counts as the actual denominator for every analysis. This is not formal implementation of diagnostic/QUADAS/STARD, RoB/GRADE or causality assessments, or clinical certification.
- Related sources: S01, S04, S05

### N30 · To what level were measurement procedures, calibration/QC and specimen handling verified?

Verification status: Unverified  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: Retained S01/S02/S04 contain historical locations where methods/assay principles were verified. The KO/structure currently read does not embed detailed values for calibration/QC, storage or measurement order/reading, so those procedures were not established.
- Why unverified: The existence of a verification location differs from the detailed values embedded in the retained summary read in this review.
- Reason for missing information: Actual comparison of original methodological procedures, calibration/QC, specimen handling and measurement/reading records.
- Required materials and conditions: Assay-specific calibration/QC, specimen collection/storage, measurement order, reading/blinding and original methodological procedures
- Materials to obtain first: Assay-specific calibration/QC, specimen collection/storage, measurement order, reading/blinding and original methodological procedures
- Further research: Compare handling/calibration/measurement conditions for the same specimens to verify procedures that could explain differences between time points/platforms.
- Conditions for verification: Distinguish the original verification locations from the information currently embedded, leaving gaps in comparison of calibration/QC, storage and measurement order/reading.
- Why this condition is needed: This is needed to avoid claiming that all analytical procedures and alternative explanations were verified solely from platform names and full-text reading labels.
- Application exceptions: This is limited to the absence of detailed procedural values in the KO/structure currently read. It does not mean that the entire original full text failed to report procedures.
- Remaining uncertainty: Actual comparison of original methodological procedures, calibration/QC, specimen handling and measurement/reading records. This status concerns the scope of documentation for that narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/individual safety, actual misdiagnosis rates or worldwide absence of evidence.
- Referenced methods: D03: Chamgap's internally designed diagnostic condition; formal diagnostic-method evidence not reviewed / D05: Chamgap's internally designed diagnostic condition; formal diagnostic-method evidence not reviewed / E03: M08 Table 8.2.a, M03 §3.2.4; internal application. Reason for application to the measurement family: This is needed to avoid claiming that all analytical procedures and alternative explanations were verified solely from platform names and full-text reading labels. This is not formal implementation of diagnostic/QUADAS/STARD, RoB/GRADE or causality assessments, or clinical certification.
- Related sources: S01, S02, S04

### N31 · Were registration, protocol and SAP compared with actual platforms/time points/contrasts?

Verification status: Unverified  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: Li's NCT03034707 and Zhang's ChiCTR1800020472 identifiers are retained. Some registration histories, protocols and SAPs, as well as numerical supplements/full tables that were not accessible in the historical record, are currently unverified. Agreement with the original plan or selective reporting was not established.
- Why unverified: Identifiers and analysis locations alone do not complete the comparison with advance plans/multiplicity.
- Reason for missing information: Actual original-source comparison of registration histories, plan versions and the complete reports/supplements.
- Required materials and conditions: NCT03034707 and ChiCTR1800020472 histories, protocol/SAP, plans for platforms/endpoints/time points and numerical supplements
- Materials to obtain first: NCT03034707 and ChiCTR1800020472 histories, protocol/SAP, plans for platforms/endpoints/time points and numerical supplements
- Further research: Compare registration/plan versions and dates with the contrasts/time points actually reported, distinguishing reasons for changes and handling of multiplicity.
- Conditions for verification: Link registration identifiers to original analysis locations and record that plans/multiplicity and the full numerical supplements have not been compared.
- Why this condition is needed: This is needed to avoid arbitrarily establishing that favorable reported platforms/time points match the original advance plan.
- Application exceptions: Do not convert unverified registration/planning into confirmed nonregistration or selective reporting.
- Remaining uncertainty: Actual original-source comparison of registration histories, plan versions and the complete reports/supplements. This status concerns the scope of documentation for that narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/individual safety, actual misdiagnosis rates or worldwide absence of evidence.
- Referenced methods: E04: M08 §8.2.2, M15 §15.3.1; internal application / E06: M08 Table 8.2.a, M04 §4.6.3; internal application / C09: M04 §4.5, §4.6.3; limitations on the current data transformation follow the user's scope and internal design / D05: Chamgap's internally designed diagnostic condition; formal diagnostic-method evidence not reviewed. Reason for application to the measurement family: This is needed to avoid arbitrarily establishing that favorable reported platforms/time points match the original advance plan. This is not formal implementation of diagnostic/QUADAS/STARD, RoB/GRADE or causality assessments, or clinical certification.
- Related sources: S01, S02, S04, S05

### N32 · Were reports, platforms, corrections and method documents counted as independent human replications?

Verification status: Met  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: F01's Li main article S01, supplement S02 and correction S03 belong to the same family. F02's Zhang S04, F03's Ylli S05 and A01's manufacturer document S08 have distinct roles. Multiple platforms, 12 bibliographic entries and secondary citations in official documents were not counted as 12 human replications.
- Why unverified: The retained family identities and report roles have been linked. Replication was not calculated from the number of platforms or bibliographic entries.
- Reason for missing information: Do not assume that evidence levels or population/reagent conditions are the same, or add up unverified unique participants.
- Required materials and conditions: Report-family-assay/outcome relationships and links to registrations, corrections, method documents and repeat citations
- Materials to obtain first: Report-family-assay/outcome relationships and links to registrations, corrections, method documents and repeat citations
- Further research: Map unique participant families to report, platform and document types, and assess independence and evidence level separately.
- Conditions for verification: Link the identities and roles of F01's S01/S02/S03, F02's S04, F03's S05 and A01's S08.
- Why this condition is needed: This is needed to avoid adding together the main article/supplement/correction from the same study, multiple platforms using the same specimens, and manufacturer documents as independent human replications.
- Application exceptions: The direct data from F01/F02 and the abstract-level corroboration from F03 differ in evidence level. A01 is not a human study.
- Remaining uncertainty: Do not assume that evidence levels or population/reagent conditions are the same, or add up unverified unique participants. This status describes the scope of the record for the specified narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/individual safety, the actual misdiagnosis rate, or a worldwide absence of data.
- Referenced methods: C02: Distinction between studies and reports in M04 §4.6.1–4.6.2; linking cohorts/outcomes is an internal application / E07: M04 §4.6.1–4.6.2, M15 §15.3.1; internal application / H07: M04 §4.6.1–4.6.2; adaptation for harm data is internally designed / D08: Chamgap's own diagnostic-application condition; general reference to M04's principle of distinguishing reports from studies. Reason for application to the measurement family: This is needed to avoid adding together the main article/supplement/correction from the same study, multiple platforms using the same specimens, and manufacturer documents as independent human replications. This is not formal implementation of diagnostic/QUADAS/STARD, RoB/GRADE or causality methods, or clinical certification.
- Related sources: S01, S02, S03, S04, S05, S08

### N33 · To which report do funding, product provision and author relationships belong?

Verification status: Partly verified  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: Li records NIH, university and participating-laboratory support and author-specific commercial relationships as disclosed in the original source. Zhang declares no funding/conflicts of interest; this is not independent verification. Ylli's detailed funding and product roles are unverified, and the Roche method document is a manufacturer document.
- Why unverified: The verification levels of the reports are distinguished, but full comparison against the original detailed declarations and product roles remains outstanding.
- Reason for missing information: Original detailed declarations of funding, product supply/roles and author relationships for each report.
- Required materials and conditions: Study-specific funding, product roles, author relationships, declarations in manufacturer performance documents, and their original locations
- Materials to obtain first: Study-specific funding, product roles, author relationships, declarations in manufacturer performance documents, and their original locations
- Further research: Compare the funding/supply/declarations for the relevant report without transferring them to the independence of another study or the validity of its results.
- Conditions for verification: Retain Li's support/relationships, Zhang's declaration of none, Ylli's unverified details and the Roche manufacturer document separately.
- Why this condition is needed: This is needed to avoid treating manufacturer documents or unverified funding as independent performance validation.
- Application exceptions: Support or competing interests do not automatically invalidate a result, and a declaration of none does not establish independent verification.
- Remaining uncertainty: Original detailed declarations of funding, product supply/roles and author relationships for each report. This status describes the scope of the record for the specified narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/individual safety, the actual misdiagnosis rate, or a worldwide absence of data.
- Referenced methods: C10: Chamgap's own independence and attribution condition. The official methods sections used in this demonstration do not warrant all of these detailed items / C02: Distinction between studies and reports in M04 §4.6.1–4.6.2; linking cohorts/outcomes is an internal application. Reason for application to the measurement family: This is needed to avoid treating manufacturer documents or unverified funding as independent performance validation. This is not formal implementation of diagnostic/QUADAS/STARD, RoB/GRADE or causality methods, or clinical certification.
- Related sources: S01, S04, S05, S08

### N34 · Has current reading of local summaries been distinguished from historical external access and access dates?

Verification status: Met  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: The current stored manuscript's 17 KO fields, 4 structured evidence entries and 12 bibliographic entries, plus measurement/display/correction and historical metadata, were read. Historical full text/PDF access, official extracts, abstracts and access failures are historical records. All external materials remain unread in this review; individual access dates are null, and the original review date, 2026-09-15, is separate.
- Why unverified: This is a record condition that separates materials actually read in the current review from historical access and dates.
- Reason for missing information: The information currently available does not establish report-specific actual access dates or preservation/SHA of historical original binaries.
- Required materials and conditions: Physical material types, URLs, historical access_level/location, report-specific access-date records and actions in this review
- Materials to obtain first: Physical material types, URLs, historical access_level/location, report-specific access-date records and actions in this review
- Further research: Compare each link's actual material type with its historical access level, and record dates/current access events separately.
- Conditions for verification: Separate current reading of the retained KO/structure and bibliography from unattempted access, non-acquisition and non-reading for each external material type, and from historical access levels.
- Why this condition is needed: This is needed to avoid elevating publisher_full_text or other historical access_level values, or links, to original-source reading in this review.
- Application exceptions: The original verified_date is a global review date. If a report-specific access date is absent, it remains null; historical access failures are not displayed as current failures.
- Remaining uncertainty: The information currently available does not establish report-specific actual access dates or preservation/SHA of historical original binaries. This status describes the scope of the record for the specified narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/individual safety, the actual misdiagnosis rate, or a worldwide absence of data.
- Referenced methods: C03: Data collection/selection principles in M04 §4.5–4.6; separation of access from reading and the hyperlink contract are internally designed / C09: M04 §4.5, §4.6.3; limits on the current data transition follow the user-defined scope and internal design. Reason for application to the measurement family: This is needed to avoid elevating publisher_full_text or other historical access_level values, or links, to original-source reading in this review. This is not formal implementation of diagnostic/QUADAS/STARD, RoB/GRADE or causality methods, or clinical certification.
- Related sources: S01, S02, S03, S04, S05, S06, S07, S08, S09, S10, S11, S12

### N35 · Were discovery leads for S10/S11/S12 elevated to direct performance figures?

Verification status: Unverified  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: S10 ETA2026 is a historical record of metadata discovery only and a failed attempt to access the publisher's full text. The original concentration-response tables for S11 LOCI and the original extraction for S12 routine immunoassays were also unverified at that time. These were not elevated to new access or direct quantitative evidence in this review.
- Why unverified: The currently retained summaries contain no direct performance or concentration-response information.
- Reason for missing information: Actual comparison against the three candidates' original full texts and tables, their currency and the reagents to which they apply.
- Required materials and conditions: Original full texts/concentration-response tables and identity/access records for the S10 guideline, S11 LOCI and S12 clinical-measurement candidates
- Materials to obtain first: Original full texts/concentration-response tables and identity/access records for the S10 guideline, S11 LOCI and S12 clinical-measurement candidates
- Further research: Link discovered bibliographic records to full-text results without extracting or estimating concentrations/bias figures from tables that were not accessed.
- Conditions for verification: Retain the actual identities of the three candidates, the scope of historical discovery/failure, and the unverified original tables.
- Why this condition is needed: This is needed to avoid inventing original quantitative tables from the titles of guideline, LOCI and routine immunoassay papers whose full texts were not accessed.
- Application exceptions: Separate unattempted access in the current review from historical full-text access failures. A link/PMCID is not evidence that the full text was read.
- Remaining uncertainty: Actual comparison against the three candidates' original full texts and tables, their currency and the reagents to which they apply. This status describes the scope of the record for the specified narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/individual safety, the actual misdiagnosis rate, or a worldwide absence of data.
- Referenced methods: C03: Data collection/selection principles in M04 §4.5–4.6; separation of access from reading and the hyperlink contract are internally designed / C09: M04 §4.5, §4.6.3; limits on the current data transition follow the user-defined scope and internal design / C08: M04 §4.6.3, M15 §15.3.1–15.3.3; status separation is internally designed. Reason for application to the measurement family: This is needed to avoid inventing original quantitative tables from the titles of guideline, LOCI and routine immunoassay papers whose full texts were not accessed. This is not formal implementation of diagnostic/QUADAS/STARD, RoB/GRADE or causality methods, or clinical certification.
- Related sources: S10, S11, S12

### N36 · Were the historical NIH/FDA extracts counted as independent TSH studies?

Verification status: Met  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: NIH S06 and FDA S07 inherit the original sources, extracts and access records from earlier work, with no new search. They are general safety materials and were not counted as independent TSH studies or evidence of the same bias across all analytes.
- Why unverified: The distinction between official background, historical inheritance and independent replication was verified. The official documents have not been newly read in the current review.
- Reason for missing information: No current comparison against the previous separate extraction/access logs and original official documents was performed.
- Required materials and conditions: Historical NIH/FDA extracts, original access histories, and the documents' actual scope of analytes/secondary citations
- Materials to obtain first: Historical NIH/FDA extracts, original access histories, and the documents' actual scope of analytes/secondary citations
- Further research: Link safety background/secondary citations in official documents to unique studies and distinguish them from independent results.
- Conditions for verification: Retain S06/S07's inherited_official_extract status, general safety role and boundary of no new search.
- Why this condition is needed: This is needed to avoid double-counting official general safety documents and secondary citations as independent human measurement results.
- Application exceptions: Do not transfer the analyte scope of the FDA troponin document to direct TSH figures.
- Remaining uncertainty: No current comparison against the previous separate extraction/access logs and original official documents was performed. This status describes the scope of the record for the specified narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/individual safety, the actual misdiagnosis rate, or a worldwide absence of data.
- Referenced methods: C02: Distinction between studies and reports in M04 §4.6.1–4.6.2; linking cohorts/outcomes is an internal application / C03: Data collection/selection principles in M04 §4.5–4.6; separation of access from reading and the hyperlink contract are internally designed / H09: Chamgap's own interpretation boundary. Reference to M19 principles, not implementation of a formal causality tool. Reason for application to the measurement family: This is needed to avoid double-counting official general safety documents and secondary citations as independent human measurement results. This is not formal implementation of diagnostic/QUADAS/STARD, RoB/GRADE or causality methods, or clinical certification.
- Related sources: S06, S07

### N37 · Is AACC guidance distinguished from individual stopping intervals and testing/treatment schedules?

Verification status: Partly verified  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: S09 AACC2020 guidance is a historical original-source record of mechanisms, assay specificity, renal elimination and recommendations. The original table does not prescribe an individual's supplement stopping interval, testing schedule or treatment change, and prohibits a universal stopping interval.
- Why unverified: The role of the guidance and scope of action have been linked, but the original recommendation context that is not currently embedded, and individual conditions, have not been verified.
- Reason for missing information: The original recommendation context for the relevant conditions and information specific to actual exposure/installations and populations.
- Required materials and conditions: The actual context of guidance recommendations by population, renal function/exposure and assay, and individual testing conditions
- Materials to obtain first: The actual context of guidance recommendations by population, renal function/exposure and assay, and individual testing conditions
- Further research: Compare assay/population-specific guidance conditions with actual installations and exposure, without generalizing to a universal interval or individual action.
- Conditions for verification: Link AACC's historical scope to the prohibition on a universal stopping interval or individual treatment changes.
- Why this condition is needed: This is needed to avoid converting general guidance or study time points into a safe stopping interval for a particular person.
- Application exceptions: The original numerical recommendations in the guidance are not embedded in the current JSON, so no new values are presented.
- Remaining uncertainty: The original recommendation context for the relevant conditions and information specific to actual exposure/installations and populations. This status describes the scope of the record for the specified narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/individual safety, the actual misdiagnosis rate, or a worldwide absence of data.
- Referenced methods: D09: Chamgap's own boundary for diagnostic extensions; official diagnostic-method evidence not reviewed / E08: M03 §3.2.1–3.2.4; dose, duration and effect-modification details are internally designed / H02: Design-specific harm questions in M19 §19.2; timing-relationship details are internally designed / H08: M19 §19.1.2–19.2.4; internal application. Reason for application to the measurement family: This is needed to avoid converting general guidance or study time points into a safe stopping interval for a particular person. This is not formal implementation of diagnostic/QUADAS/STARD, RoB/GRADE or causality methods, or clinical certification.
- Related sources: S09

### N38 · What possibility and what actual observations does the Warning represent?

Verification status: Partly verified  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: The original safety status is Warning. This reflects the possibility that falsely low TSH could lead to interpretation errors; the direct observations are assay-specific bias and reference-limit crossing. The incidence of actual patient harm is not an established result.
- Why unverified: The original reason for the Warning and the level of direct observation have been retained. There are no data establishing the frequency/causation of actual harm.
- Reason for missing information: Actual definitions and collection of patient events, denominators/follow-up, post-test actions and alternative causes.
- Required materials and conditions: Reference-limit crossing of measured values, actual clinical judgments/actions after testing, and harm definitions/collection/denominators/follow-up
- Materials to obtain first: Reference-limit crossing of measured values, actual clinical judgments/actions after testing, and harm definitions/collection/denominators/follow-up
- Further research: Compare analytical bias separately with the timing, alternative explanations and denominators for post-test decisions and patient events to verify the scope of direct harm evidence.
- Conditions for verification: Retain the original safety Warning, measured values and reference-limit crossing, while distinguishing the absence of data on actual misdiagnosis/harm.
- Why this condition is needed: This is needed to avoid converting the possibility of interpretation errors from falsely low measurements into an actual patient harm rate or established causation of an event.
- Application exceptions: Do not reuse the presence of measurement bias as a judgment of individual safety/efficacy.
- Remaining uncertainty: Actual definitions and collection of patient events, denominators/follow-up, post-test actions and alternative causes. This status describes the scope of the record for the specified narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/individual safety, the actual misdiagnosis rate, or a worldwide absence of data.
- Referenced methods: H01: M19 §19.1.2–19.2.4; detailed distinctions are internally designed / H04: M19 §19.1.2–19.2.4; internal application / H09: Chamgap's own interpretation boundary. Reference to M19 principles, not implementation of a formal causality tool / D09: Chamgap's own boundary for diagnostic extensions; official diagnostic-method evidence not reviewed. Reason for application to the measurement family: This is needed to avoid converting the possibility of interpretation errors from falsely low measurements into an actual patient harm rate or established causation of an event. This is not formal implementation of diagnostic/QUADAS/STARD, RoB/GRADE or causality methods, or clinical certification.
- Related sources: S01, S04, S09

### N39 · Have bias magnitude and duration outside the healthy-adult conditions been verified?

Verification status: Unverified  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: These healthy-adult data did not establish bias magnitude or duration in populations with renal impairment, pregnancy, childhood or high-dose treatment. Do not generalize to individual stopping intervals or long-term safety.
- Why unverified: The currently retained materials contain no direct measurement information beyond the selected healthy-adult conditions.
- Reason for missing information: Direct data on quantitative bias, intervals and duration by population, exposure and reagent.
- Required materials and conditions: Actual exposure, reagents and blood-draw-time-specific bias/duration information for populations defined by renal function, pregnancy, childhood or high-dose treatment
- Materials to obtain first: Actual exposure, reagents and blood-draw-time-specific bias/duration information for populations defined by renal function, pregnancy, childhood or high-dose treatment
- Further research: Compare exposure/elimination, underlying conditions and assay/time-point data for special populations, and distinguish them from healthy-adult results.
- Conditions for verification: Record the direct study conditions separately from unverified information on special populations and long-term/rare situations.
- Why this condition is needed: This is needed to avoid extending the results of short-term exposure in healthy adults to all populations with renal impairment, pregnancy, childhood or high-dose treatment.
- Application exceptions: Studies in every vulnerable population are not mandatory to establish this narrow measurement question, but direct information is required for the corresponding extension claims.
- Remaining uncertainty: Direct data on quantitative bias, intervals and duration by population, exposure and reagent. This status describes the scope of the record for the specified narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/individual safety, the actual misdiagnosis rate, or a worldwide absence of data.
- Referenced methods: C04: M03 §3.2.1; detailed population conditions are internally designed / H08: M19 §19.1.2–19.2.4; internal application / E08: M03 §3.2.1–3.2.4; dose, duration and effect-modification details are internally designed. Reason for application to the measurement family: This is needed to avoid extending the results of short-term exposure in healthy adults to all populations with renal impairment, pregnancy, childhood or high-dose treatment. This is not formal implementation of diagnostic/QUADAS/STARD, RoB/GRADE or causality methods, or clinical certification.
- Related sources: S01, S04, S05, S09

### N40 · Which retained reference date underlies the statements, and what changes trigger a new review?

Verification status: Partly verified  
Reading scope: Summary read: The specified pointers and original values in the current stored manuscript were actually read. Access to these external material types was not attempted in this review; they were neither acquired nor read.

- Current findings: The evidence-review date is 2026-09-15, and this retained-material transition is dated 2026-10-05. The original revision_history records the initial measurement-only current value and retains the unit correction. No new search, original-source verification or verification of reagent installations/the latest guidance was performed.
- Why unverified: The recorded reference date and reasons for updates have been linked, but up-to-date completeness and actual current installations have not been verified.
- Reason for missing information: Authorized actual comparison of reagent/publication changes, corrections, retractions and numerical/classification changes after the reference date.
- Required materials and conditions: Reference date, current review date, search/access records, reagent revisions, original corrections/retractions, and new numerical/classification data
- Materials to obtain first: Reference date, current review date, search/access records, reagent revisions, original corrections/retractions, and new numerical/classification data
- Further research: Retain before/after original values, versions, dates and access levels under the same ID, and compare the basis for changes and revisions of conditions as separate events.
- Conditions for verification: Link the evidence-review date of 2026-09-15 and this review date to unverified information, corrections/revisions and the update scope.
- Why this condition is needed: This is needed to avoid overstating the retained current value as verification of all the latest literature or current installations, or as an exhaustive search.
- Application exceptions: Verification of new reagent documents, corrections, retractions and numerical/classification errors is not a request for automatic execution now.
- Remaining uncertainty: Authorized actual comparison of reagent/publication changes, corrections, retractions and numerical/classification changes after the reference date. This status describes the scope of the record for the specified narrow measurement condition. It does not establish absence of interference, equivalence, efficacy/individual safety, the actual misdiagnosis rate, or a worldwide absence of data.
- Referenced methods: C09: M04 §4.5, §4.6.3; limits on the current data transition follow the user-defined scope and internal design / C08: M04 §4.6.3, M15 §15.3.1–15.3.3; status separation is internally designed / C03: Data collection/selection principles in M04 §4.5–4.6; separation of access from reading and the hyperlink contract are internally designed. Reason for application to the measurement family: This is needed to avoid overstating the retained current value as verification of all the latest literature or current installations, or as an exhaustive search. This is not formal implementation of diagnostic/QUADAS/STARD, RoB/GRADE or causality methods, or clinical certification.
- Related sources: S03, S08, S10

## Evidence and sources

### S01

Li et al. 2017. Association of Biotin Ingestion With Performance of Hormone and Nonhormone Assays in Healthy Adults

[Source](https://jamanetwork.com/journals/jama/fullarticle/2654822)

Materials actually read: The relevant bibliographic entry, summary and structure in the stored manuscript were read. The linked external original materials were not acquired or read in this review.  
Findings recorded in the summary: Record at the time of retention: 6 healthy adults, 18 time-point specimens, 10 mg/day for 7 days. day 7 was approximately 2 hours after the last dose, and day 14 was 7 days after stopping. The formal contrast day 7 − mean(day 0, day 14) is retained separately for Roche −0.72/CI −1.13~−0.32/P = 0.006, Ortho −1.67/CI −2.08~−1.26/P < 0.001, Vista −0.11/CI −0.49~0.26/P = 0.56 and Abbott −0.06/CI −0.44~0.31/P = 0.74 mIU/L. The historical simple difference between means, −0.59, is separate from the model contrast. Serum-biotin 3601 is an assigned value after the upper limit of 3600 was exceeded. NIH, university and participating-laboratory support/author relationships are records for this study.  
Record level: This is a transition of the KO/structured evidence, bibliography and historical access locations retained as of 2026-09-15. It is not external original-source reading, new numerical calculation or reassessment in this review.  
Independent replication: False

| Material | Public availability | Access status | Actually read | Notes |
|---|---|---|---|---|
| Publisher's full HTML text of the original article | Unverified (no external lookup in this review) | Access not attempted | Not read | The input actually read in this review is the source JSON containing the retained bibliographic entry and Korean summary/structure. No access to this external material type was attempted in this review; it was not acquired or read. This link matches the actual material type of the URL in the currently retained bibliographic entry. Historical access level: publisher_full_text. Methods; Statistical Analysis; Table 1 TSH rows; Table 4; Discussion; Article Information. Report-specific historical access dates are unverified. The stored manuscript's evidence-review date, 2026-09-15, is not the date on which the original material was actually accessed. No access to external material was attempted and no original source was read in this review. |
| Registration version history | Unverified (no external lookup in this review) | Access not attempted | Not read | The input actually read in this review is the source JSON containing the retained bibliographic entry and Korean summary/structure. No access to this external material type was attempted in this review; it was not acquired or read. A separate actual URL for this material type is currently unverified. The original bibliographic link is an identifying reference and is not presented as the original source for this material type. Historical access level: publisher_full_text. Methods; Statistical Analysis; Table 1 TSH rows; Table 4; Discussion; Article Information. Report-specific historical access dates are unverified. The stored manuscript's evidence-review date, 2026-09-15, is not the date on which the original material was actually accessed. No access to external material was attempted and no original source was read in this review. |
| Protocol | Unverified (no external lookup in this review) | Access not attempted | Not read | The input actually read in this review is the source JSON containing the retained bibliographic entry and Korean summary/structure. No access to this external material type was attempted in this review; it was not acquired or read. A separate actual URL for this material type is currently unverified. The original bibliographic link is an identifying reference and is not presented as the original source for this material type. Historical access level: publisher_full_text. Methods; Statistical Analysis; Table 1 TSH rows; Table 4; Discussion; Article Information. Report-specific historical access dates are unverified. The stored manuscript's evidence-review date, 2026-09-15, is not the date on which the original material was actually accessed. No access to external material was attempted and no original source was read in this review. |
| Analysis plan (SAP) | Unverified (no external lookup in this review) | Access not attempted | Not read | The input actually read in this review is the source JSON containing the retained bibliographic entry and Korean summary/structure. No access to this external material type was attempted in this review; it was not acquired or read. A separate actual URL for this material type is currently unverified. The original bibliographic link is an identifying reference and is not presented as the original source for this material type. Historical access level: publisher_full_text. Methods; Statistical Analysis; Table 1 TSH rows; Table 4; Discussion; Article Information. Report-specific historical access dates are unverified. The stored manuscript's evidence-review date, 2026-09-15, is not the date on which the original material was actually accessed. No access to external material was attempted and no original source was read in this review. |

### S02

Li et al. 2017 supplementary methods and tables

[Source](https://cdn.jamanetwork.com/ama/content_public/journal/jama/936497/joi170104supp1_prod.pdf)

Materials actually read: The relevant bibliographic entry, summary and structure in the stored manuscript were read. The linked external original materials were not acquired or read in this review.  
Findings recorded in the summary: Record at the time of retention: eFigure 1 and eTable 1, PDF p6, from the same Li study are historical locations where TSH assay principles and reference intervals were checked. These locations relate to the VITROS lower limit of 0.47 mIU/L and all 6 participants' post-exposure readings below 0.15 mIU/L, nonsignificance at this exposure despite Vista's biotin dependence, and interpretation boundaries for same-specimen comparisons. No independent participants/studies are added.  
Record level: This is a transition of the KO/structured evidence, bibliography and historical access locations retained as of 2026-09-15. It is not external original-source reading, new numerical calculation or reassessment in this review.  
Independent replication: False

| Material | Public availability | Access status | Actually read | Notes |
|---|---|---|---|---|
| Methods and numerical supplement PDF | Unverified (no external lookup in this review) | Access not attempted | Not read | The input actually read in this review is the source JSON containing the retained bibliographic entry and Korean summary/structure. No access to this external material type was attempted in this review; it was not acquired or read. This link matches the actual material type of the URL in the currently retained bibliographic entry. Historical access level: publisher_full_pdf. eFigure 1; eTable 1, PDF p6: TSH assay principles and reference intervals. Report-specific historical access dates are unverified. The stored manuscript's evidence-review date, 2026-09-15, is not the date on which the original material was actually accessed. No access to external material was attempted and no original source was read in this review. |

### S03

Incorrect Unit of Measure. JAMA 2017;318(19):1937

[Source](https://jamanetwork.com/journals/jama/fullarticle/2664017)

Materials actually read: The relevant bibliographic entry, summary and structure in the stored manuscript were read. The linked external original materials were not acquired or read in this review.  
Findings recorded in the summary: Record at the time of retention: The published correction dated 2017-11-21 retains the serum-biotin unit as pg/mL without changing TSH mIU/L. The day 7 value of 3601 is an assignment after the upper limit of 3600 was exceeded, not an actual precisely measured concentration. The correction belongs to the same Li report family and is not a new human replication.  
Record level: This is a transition of the KO/structured evidence, bibliography and historical access locations retained as of 2026-09-15. It is not external original-source reading, new numerical calculation or reassessment in this review.  
Independent replication: False

| Material | Public availability | Access status | Actually read | Notes |
|---|---|---|---|---|
| Full HTML text of the published unit correction | Unverified (no external lookup in this review) | Access not attempted | Not read | The input actually read in this review is the source JSON containing the retained bibliographic entry and Korean summary/structure. No access to this external material type was attempted in this review; it was not acquired or read. This link matches the actual material type of the URL in the currently retained bibliographic entry. Historical access level: publisher_full_text. Correction paragraph published 2017-11-21. Report-specific historical access dates are unverified. The stored manuscript's evidence-review date, 2026-09-15, is not the date on which the original material was actually accessed. No access to external material was attempted and no original source was read in this review. |

### S04

Zhang et al. 2020. Assessment of biotin interference in thyroid function tests

[Source](https://www.researchgate.net/publication/339621681_Assessment_of_biotin_interference_in_thyroid_function_tests)

Materials actually read: The relevant bibliographic entry, summary and structure in the stored manuscript were read. The linked external original materials were not acquired or read in this review.  
Findings recorded in the summary: Record at the time of retention: The same 10 healthy adults used 5/10 mg/day for 7 days each in a fixed sequence. Roche TSH was lower at 5 mg, 2 hours, P = 0.005, and at 10 mg, 2/8 hours, each P < 0.001; at 26/50 hours, both doses had nonsignificant comparisons with baseline. Exact mean differences/CI and analyzed specimen counts by platform/time point are unverified. The specified TSH methods on Beckman UniCel DxI 800 and Abbott ARCHITECT 2000 did not use biotin-streptavidin, and interference was not observed under the tested conditions; results for other analytes are not transferred. No funding/conflicts of interest is an author declaration.  
Record level: This is a transition of the KO/structured evidence, bibliography and historical access locations retained as of 2026-09-15. It is not external original-source reading, new numerical calculation or reassessment in this review.  
Independent replication: False

| Material | Public availability | Access status | Actually read | Notes |
|---|---|---|---|---|
| Full text of the original article in the ResearchGate repository | Unverified (no external lookup in this review) | Access not attempted | Not read | The input actually read in this review is the source JSON containing the retained bibliographic entry and Korean summary/structure. No access to this external material type was attempted in this review; it was not acquired or read. This link matches the actual material type of the URL in the currently retained bibliographic entry. Historical access level: primary_article_full_text_on_repository. Medicine 99(9):e19232; sections 2.1–2.3, 3.1–3.3, Table 1 and discussion; Figure 1 timing. Report-specific historical access dates are unverified. The stored manuscript's evidence-review date, 2026-09-15, is not the date on which the original material was actually accessed. No access to external material was attempted and no original source was read in this review. |
| Registration version history | Unverified (no external lookup in this review) | Access not attempted | Not read | The input actually read in this review is the source JSON containing the retained bibliographic entry and Korean summary/structure. No access to this external material type was attempted in this review; it was not acquired or read. A separate actual URL for this material type is currently unverified. The original bibliographic link is an identifying reference and is not presented as the original source for this material type. Historical access level: primary_article_full_text_on_repository. Medicine 99(9):e19232; sections 2.1–2.3, 3.1–3.3, Table 1 and discussion; Figure 1 timing. Report-specific historical access dates are unverified. The stored manuscript's evidence-review date, 2026-09-15, is not the date on which the original material was actually accessed. No access to external material was attempted and no original source was read in this review. |
| Protocol | Unverified (no external lookup in this review) | Access not attempted | Not read | The input actually read in this review is the source JSON containing the retained bibliographic entry and Korean summary/structure. No access to this external material type was attempted in this review; it was not acquired or read. A separate actual URL for this material type is currently unverified. The original bibliographic link is an identifying reference and is not presented as the original source for this material type. Historical access level: primary_article_full_text_on_repository. Medicine 99(9):e19232; sections 2.1–2.3, 3.1–3.3, Table 1 and discussion; Figure 1 timing. Report-specific historical access dates are unverified. The stored manuscript's evidence-review date, 2026-09-15, is not the date on which the original material was actually accessed. No access to external material was attempted and no original source was read in this review. |
| Analysis plan (SAP) | Unverified (no external lookup in this review) | Access not attempted | Not read | The input actually read in this review is the source JSON containing the retained bibliographic entry and Korean summary/structure. No access to this external material type was attempted in this review; it was not acquired or read. A separate actual URL for this material type is currently unverified. The original bibliographic link is an identifying reference and is not presented as the original source for this material type. Historical access level: primary_article_full_text_on_repository. Medicine 99(9):e19232; sections 2.1–2.3, 3.1–3.3, Table 1 and discussion; Figure 1 timing. Report-specific historical access dates are unverified. The stored manuscript's evidence-review date, 2026-09-15, is not the date on which the original material was actually accessed. No access to external material was attempted and no original source was read in this review. |

### S05

Ylli et al. 2021. Biotin Interference in Assays for Thyroid Hormones, Thyrotropin and Thyroglobulin

[Source](https://pubmed.ncbi.nlm.nih.gov/34042535/)

Materials actually read: The relevant bibliographic entry, summary and structure in the stored manuscript were read. The linked external original materials were not acquired or read in this review.  
Findings recorded in the summary: Record at the time of retention: 13 adults enrolled, 10 mg/day for 8 days, with measurements on day 1/day 8 before dosing and at 2 hours and 5 hours. The abstract supports Roche TSH changes and the greatest interference at 2 hours. Detailed TSH analysis n, full-text quantitative tables, exact bias/CI and funding/product roles are unverified; percentages from search summaries were not imported.  
Record level: This is a transition of the KO/structured evidence, bibliography and historical access locations retained as of 2026-09-15. It is not external original-source reading, new numerical calculation or reassessment in this review.  
Independent replication: False

| Material | Public availability | Access status | Actually read | Notes |
|---|---|---|---|---|
| PubMed abstract and figure captions | Unverified (no external lookup in this review) | Access not attempted | Not read | The input actually read in this review is the source JSON containing the retained bibliographic entry and Korean summary/structure. No access to this external material type was attempted in this review; it was not acquired or read. This link matches the actual material type of the URL in the currently retained bibliographic entry. Historical access level: pubmed_abstract_and_figure_captions. Abstract Methods/Results; TSH figure caption; conflicts statement. Report-specific historical access dates are unverified. The stored manuscript's evidence-review date, 2026-09-15, is not the date on which the original material was actually accessed. No access to external material was attempted and no original source was read in this review. |
| Quantitative tables in the original article's full text | Unverified (no external lookup in this review) | Access not attempted | Not read | The input actually read in this review is the source JSON containing the retained bibliographic entry and Korean summary/structure. No access to this external material type was attempted in this review; it was not acquired or read. A separate actual URL for this material type is currently unverified. The original bibliographic link is an identifying reference and is not presented as the original source for this material type. Historical access level: pubmed_abstract_and_figure_captions. Abstract Methods/Results; TSH figure caption; conflicts statement. Report-specific historical access dates are unverified. The stored manuscript's evidence-review date, 2026-09-15, is not the date on which the original material was actually accessed. No access to external material was attempted and no original source was read in this review. |
| Original numerical supplement | Unverified (no external lookup in this review) | Access not attempted | Not read | The input actually read in this review is the source JSON containing the retained bibliographic entry and Korean summary/structure. No access to this external material type was attempted in this review; it was not acquired or read. A separate actual URL for this material type is currently unverified. The original bibliographic link is an identifying reference and is not presented as the original source for this material type. Historical access level: pubmed_abstract_and_figure_captions. Abstract Methods/Results; TSH figure caption; conflicts statement. Report-specific historical access dates are unverified. The stored manuscript's evidence-review date, 2026-09-15, is not the date on which the original material was actually accessed. No access to external material was attempted and no original source was read in this review. |

### S06

NIH Office of Dietary Supplements: Biotin—Health Professional Fact Sheet

[Source](https://ods.od.nih.gov/factsheets/Biotin-HealthProfessional/)

Materials actually read: The relevant bibliographic entry, summary and structure in the stored manuscript were read. The linked external original materials were not acquired or read in this review.  
Findings recorded in the summary: General background record at the time of retention: The scope is inherited original sources/extracts/access records for NIH ODS Table1, risks from excessive biotin and interactions with laboratory tests, refs9/38. No new search was conducted, and this was not counted as an independent TSH study or the same bias across all analytes.  
Record level: This is a transition of the KO/structured evidence, bibliography and historical access locations retained as of 2026-09-15. It is not external original-source reading, new numerical calculation or reassessment in this review.  
Independent replication: False

| Material | Public availability | Access status | Actually read | Notes |
|---|---|---|---|---|
| Official NIH ODS web document | Unverified (no external lookup in this review) | Access not attempted | Not read | The input actually read in this review is the source JSON containing the retained bibliographic entry and Korean summary/structure. No access to this external material type was attempted in this review; it was not acquired or read. This link matches the actual material type of the URL in the currently retained bibliographic entry. Historical access level: inherited_official_extract. Table1; Health Risks from Excessive Biotin; Interactions with Laboratory Tests, refs9/38. Report-specific historical access dates are unverified. The stored manuscript's evidence-review date, 2026-09-15, is not the date on which the original material was actually accessed. No access to external material was attempted and no original source was read in this review. |

### S07

FDA: Biotin Interference with Troponin Lab Tests—Assays Subject to Biotin Interference

[Source](https://www.fda.gov/medical-devices/in-vitro-diagnostics/biotin-interference-troponin-lab-tests-assays-subject-biotin-interference)

Materials actually read: The relevant bibliographic entry, summary and structure in the stored manuscript were read. The linked external original materials were not acquired or read in this review.  
Findings recorded in the summary: General background record at the time of retention: The opening safety statement and paragraphs on residual assay interference in the FDA troponin-testing document are inherited from an earlier extraction. This is not elevated to a new search or original-source reading in this review, or to independent TSH figures/studies.  
Record level: This is a transition of the KO/structured evidence, bibliography and historical access locations retained as of 2026-09-15. It is not external original-source reading, new numerical calculation or reassessment in this review.  
Independent replication: False

| Material | Public availability | Access status | Actually read | Notes |
|---|---|---|---|---|
| Official FDA web document on troponin tests | Unverified (no external lookup in this review) | Access not attempted | Not read | The input actually read in this review is the source JSON containing the retained bibliographic entry and Korean summary/structure. No access to this external material type was attempted in this review; it was not acquired or read. This link matches the actual material type of the URL in the currently retained bibliographic entry. Historical access level: inherited_official_extract. Opening safety statement and paragraphs on residual assay interference. Report-specific historical access dates are unverified. The stored manuscript's evidence-review date, 2026-09-15, is not the date on which the original material was actually accessed. No access to external material was attempted and no original source was read in this review. |

### S08

Roche Elecsys TSH method sheet, US, 2024-10 V2.0

[Source](https://elabdoc-prod.roche.com/eLD/api/downloads/4f54ba6c-732d-ef11-2591-005056a71a5d?countryIsoCode=us)

Materials actually read: The relevant bibliographic entry, summary and structure in the stored manuscript were read. The linked external original materials were not acquired or read in this review.  
Findings recorded in the summary: Record at the time of retention: US 2024-10 V2.0, 08429324501V2.0, product ref 08429324160 is a method document for cobas e411/e601/e602. For biotin ≤1200 ng/mL, at TSH ≤0.2 µIU/mL the deviation criterion is ≤0.02 µIU/mL; at higher TSH it is ≤10%. This is a manufacturer analytical document. It does not guarantee a safe oral dose in humans or bias 0, verify every laboratory's installed version, or establish identity with historical reagents.  
Record level: This is a transition of the KO/structured evidence, bibliography and historical access locations retained as of 2026-09-15. It is not external original-source reading, new numerical calculation or reassessment in this review.  
Independent replication: False

| Material | Public availability | Access status | Actually read | Notes |
|---|---|---|---|---|
| Official Roche reagent method PDF | Unverified (no external lookup in this review) | Access not attempted | Not read | The input actually read in this review is the source JSON containing the retained bibliographic entry and Korean summary/structure. No access to this external material type was attempted in this review; it was not acquired or read. This link matches the actual material type of the URL in the currently retained bibliographic entry. Historical access level: official_full_pdf. PDF p1 product/platform/test principle; pp2–3 Limitations–interference; p3 measurement range. Report-specific historical access dates are unverified. The stored manuscript's evidence-review date, 2026-09-15, is not the date on which the original material was actually accessed. No access to external material was attempted and no original source was read in this review. |

### S09

AACC Academy guidance: Biotin Interference in Laboratory Tests

[Source](https://myadlm.org/science-and-research/academy-guidance/biotin-interference-in-laboratory-tests)

Materials actually read: The relevant bibliographic entry, summary and structure in the stored manuscript were read. The linked external original materials were not acquired or read in this review.  
Findings recorded in the summary: General background record at the time of retention: This is a historical record of checking the locations of mechanisms, assay specificity, renal elimination and recommendations in the 2020-01-13 AACC guidance. Numerical recommendations not embedded in the current JSON are not newly presented or used as individual stopping intervals, testing schedules or treatment changes.  
Record level: This is a transition of the KO/structured evidence, bibliography and historical access locations retained as of 2026-09-15. It is not external original-source reading, new numerical calculation or reassessment in this review.  
Independent replication: False

| Material | Public availability | Access status | Actually read | Notes |
|---|---|---|---|---|
| Full HTML text of official AACC guidance | Unverified (no external lookup in this review) | Access not attempted | Not read | The input actually read in this review is the source JSON containing the retained bibliographic entry and Korean summary/structure. No access to this external material type was attempted in this review; it was not acquired or read. This link matches the actual material type of the URL in the currently retained bibliographic entry. Historical access level: official_full_text. 2020-01-13 guidance; mechanisms, assay specificity, renal elimination and recommendations. Report-specific historical access dates are unverified. The stored manuscript's evidence-review date, 2026-09-15, is not the date on which the original material was actually accessed. No access to external material was attempted and no original source was read in this review. |

### S10

2026 ETA guideline on interference in immunoassay measurements used in assessment of thyroid function

[Source](https://pubmed.ncbi.nlm.nih.gov/42466561/)

Materials actually read: The relevant bibliographic entry, summary and structure in the stored manuscript were read. The linked external original materials were not acquired or read in this review.  
Findings recorded in the summary: Record at the time of retention: Guideline metadata for DOI 10.1530/ETJ-26-0011, PMID 42466561 were discovered, but the publisher full-text request failed at that time. No current access was attempted, and no detailed performance figures/new recommendations were generated from the guideline.  
Record level: This is a transition of the KO/structured evidence, bibliography and historical access locations retained as of 2026-09-15. It is not external original-source reading, new numerical calculation or reassessment in this review.  
Independent replication: False

| Material | Public availability | Access status | Actually read | Notes |
|---|---|---|---|---|
| PubMed bibliographic record for the guideline | Unverified (no external lookup in this review) | Access not attempted | Not read | The input actually read in this review is the source JSON containing the retained bibliographic entry and Korean summary/structure. No access to this external material type was attempted in this review; it was not acquired or read. This link matches the actual material type of the URL in the currently retained bibliographic entry. Historical access level: metadata_discovered_full_text_inaccessible. Search metadata only; publisher full text request failed. Report-specific historical access dates are unverified. The stored manuscript's evidence-review date, 2026-09-15, is not the date on which the original material was actually accessed. No access to external material was attempted and no original source was read in this review. |
| Publisher's full text of the guideline | Unverified (no external lookup in this review) | Access not attempted | Not read | The input actually read in this review is the source JSON containing the retained bibliographic entry and Korean summary/structure. No access to this external material type was attempted in this review; it was not acquired or read. A separate actual URL for this material type is currently unverified. The original bibliographic link is an identifying reference and is not presented as the original source for this material type. Historical access level: metadata_discovered_full_text_inaccessible. Search metadata only; publisher full text request failed. Report-specific historical access dates are unverified. The stored manuscript's evidence-review date, 2026-09-15, is not the date on which the original material was actually accessed. No access to external material was attempted and no original source was read in this review. |

### S11

Biotin interference in TSH, FT4, and FT3 assays based on the LOCI technology: identifying interference by dilution

[Source](https://pmc.ncbi.nlm.nih.gov/articles/PMC6818550/)

Materials actually read: The relevant bibliographic entry, summary and structure in the stored manuscript were read. The linked external original materials were not acquired or read in this review.  
Findings recorded in the summary: Record at the time of retention: The candidate identified by DOI 10.1002/jcla.22667, PMCID PMC6818550 concerns LOCI TSH/FT4/FT3 and is limited to discovery metadata. Historical full-text access limitations left the original concentration-response tables unverified, and the PMC original text was not opened in this review.  
Record level: This is a transition of the KO/structured evidence, bibliography and historical access locations retained as of 2026-09-15. It is not external original-source reading, new numerical calculation or reassessment in this review.  
Independent replication: False

| Material | Public availability | Access status | Actually read | Notes |
|---|---|---|---|---|
| Full HTML text of the original article on PMC (not read in the current review) | Unverified (no external lookup in this review) | Access not attempted | Not read | The input actually read in this review is the source JSON containing the retained bibliographic entry and Korean summary/structure. No access to this external material type was attempted in this review; it was not acquired or read. This link matches the actual material type of the URL in the currently retained bibliographic entry. Historical access level: search_metadata_only_full_text_inaccessible. Discovery metadata; no primary concentration-response table verified. Report-specific historical access dates are unverified. The stored manuscript's evidence-review date, 2026-09-15, is not the date on which the original material was actually accessed. No access to external material was attempted and no original source was read in this review. |

### S12

Biotin interference in routine clinical immunoassays

[Source](https://pmc.ncbi.nlm.nih.gov/articles/PMC12099460/)

Materials actually read: The relevant bibliographic entry, summary and structure in the stored manuscript were read. The linked external original materials were not acquired or read in this review.  
Findings recorded in the summary: Record at the time of retention: The candidate with DOI 10.1016/j.plabm.2025.e00472, PMCID PMC12099460 is limited to discovery metadata. The record states that a PMC challenge blocked original extraction at that time; this is not an access failure in this review. No quantitative bias, denominators or independent study identities were newly generated.  
Record level: This is a transition of the KO/structured evidence, bibliography and historical access locations retained as of 2026-09-15. It is not external original-source reading, new numerical calculation or reassessment in this review.  
Independent replication: False

| Material | Public availability | Access status | Actually read | Notes |
|---|---|---|---|---|
| Full HTML text of the original article on PMC (not read in the current review) | Unverified (no external lookup in this review) | Access not attempted | Not read | The input actually read in this review is the source JSON containing the retained bibliographic entry and Korean summary/structure. No access to this external material type was attempted in this review; it was not acquired or read. This link matches the actual material type of the URL in the currently retained bibliographic entry. Historical access level: search_metadata_only_full_text_inaccessible. Discovery metadata; PMC challenge prevented primary extraction. Report-specific historical access dates are unverified. The stored manuscript's evidence-review date, 2026-09-15, is not the date on which the original material was actually accessed. No access to external material was attempted and no original source was read in this review. |

## Review criteria and methods

These are Chamgap's own application conditions, informed by the relevant sections of official methodological guidance. They do not constitute official assessment-tool certification or a new efficacy score.

- M03 [Cochrane Handbook 6.5 (2024), chapter updated 2023-08](https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-03) · §3.1, §3.2.1–3.2.3, MECIR C5/C7/C13 · Clearly record the question, eligibility criteria and reasons for changes, and distinguish variations in the intervention. · This is a comparison of actual copies/SHA for case 1585's 5 original methodology HTML files, retaining their original bytes, and historical reuse of the principles from that time. It is not new reading of full methods texts, implementation of formal assessment tools or clinical certification.
- M04 [Cochrane Handbook 6.5.1 (2025), chapter updated 2025-03](https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-04) · §4.5, §4.6.1–4.6.3, MECIR C42 · Distinguish and link studies and reports, and do not omit a study solely because its outcomes are not reported. · This is a comparison of actual copies/SHA for case 1585's 5 original methodology HTML files, retaining their original bytes, and historical reuse of the principles from that time. It is not new reading of full methods texts, implementation of formal assessment tools or clinical certification.
- M08 [Cochrane Handbook 6.5 (2024), chapter updated 2019-10](https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-08) · §8.1–8.2.3, Table 8.2.a · Hold the outcome and effect of interest fixed, then assess randomization, deviations, missing data, measurement and selective reporting. · This is a comparison of actual copies/SHA for case 1585's 5 original methodology HTML files, retaining their original bytes, and historical reuse of the principles from that time. It is not new reading of full methods texts, implementation of formal assessment tools or clinical certification.
- M15 [Cochrane Handbook 6.5 (2024), chapter updated 2023-08](https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-15) · §15.2.4, §15.3.1–15.3.3 · Distinguish effect estimates, intervals, importance and applicability. Sample size alone does not determine precision. · This is a comparison of actual copies/SHA for case 1585's 5 original methodology HTML files, retaining their original bytes, and historical reuse of the principles from that time. It is not new reading of full methods texts, implementation of formal assessment tools or clinical certification.
- M19 [Cochrane Handbook 6.5 (2024), chapter updated 2019-10](https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-19) · §19.1.2, §19.2.1–19.2.4 · Questions about harms may require designs, populations and follow-up different from efficacy questions. Case reports do not establish incidence. · This is a comparison of actual copies/SHA for case 1585's 5 original methodology HTML files, retaining their original bytes, and historical reuse of the principles from that time. It is not new reading of full methods texts, implementation of formal assessment tools or clinical certification.

## What the statuses mean

- Met: The record meets the specified narrow conditions for measurement identity, attribution/units, version/reading and display. This does not establish efficacy, absence of interference or individual safety, or verify that the external full text has been checked in the current review.
- Partly verified: Relevant retained observations exist, but some necessary methods, reagents/versions, denominators, quantitative intervals or comparisons against original sources remain unverified.
- Unverified: The value could not be established from the currently retained materials and the scope verified at that time. This does not mean bias 0, normality, absence/failure of data or absence of interference.
- Not reported: Used only within the scope of non-reporting in sections actually read or in the historical summaries currently specified. It is not extended to entire external full texts that remain unread in this review.
- Conflicting information: Specify retained discrepancies or possible errors concerning the same fact in each row. Different platforms, time points or formal contrasts/simple differences do not in themselves constitute a conflict.
- Inaccessible: Used only for an actual access failure in this review. No external access was attempted in this review, so historical failures are not carried over as current status.
- Review incomplete: Limited to the scope in which comparison against current external materials has not been performed. It does not mean that the entire review of the retained JSON and model, which was actually completed, is incomplete.
- Not applicable: Limited to other analytes outside the required outcomes of the narrow TSH question. This is not a substitute for unread/unverified status, or certification of no interference or safety.

## Change history

2026-11-03 · Public page/materials 1.0. The 40 review items for the existing measurement question have been transferred to the public-facing display and data. Original verification statuses, figures, units, versions and unverified information are retained, without a new clinical search, scoring or reassessment.

## Measurement-bias brief

- Measurement question: After oral biotin exposure, what direction and magnitude of measurement bias were observed in specific serum TSH assays in healthy adults, compared with the same people before exposure/after washout or comparator platforms using specimens from the same time point? Distinguish findings by platform, reagent version, dose, dosing duration and time from the last dose to blood draw.
- Outside this scope: Thyroid function/treatment effects, diagnostic accuracy/actual misdiagnosis/patient harm rates, all analytes/reagents, a universal stopping interval, and extensions to safe human doses/special populations are separate questions.
- Meaning of verification status: The preceding verification statuses describe the verification scope of each narrow record condition; they are not badges of efficacy, absence of interference or individual safety.
- Safety record: The original safety status is Warning. This reflects the possibility that falsely low TSH could lead to interpretation errors; the direct observations are assay-specific bias and reference-limit crossing. The incidence of actual patient harm is not an established result.
- Document-quality record: The retained document-quality record is B, separate from the magnitude of the measurement effect or an individual safety grade.