Using Standards for FDA IVD Compliance

IVDs are medical devices with diagnostic evidence needs

FDA regulates IVDs as medical devices under the FD&C Act and applicable 21 CFR provisions, including Part 809 and device classifications in Part 862 and related panels. Consensus standards can support review, but the submission must still demonstrate the performance needed for the intended use and pathway.

FDA clearance or approval and CLIA categorisation address different questions. A device's analytical and clinical evidence supports marketing authorisation; CLIA categorisation concerns the complexity and conditions of laboratory testing. Neither replaces the other.

Define the complete IVD and regulatory route

Identify whether the submission concerns a reagent, test kit, calibrator, control, instrument, software function, specimen receptacle or combined system. Confirm the classification regulation, product code, class, pathway, predicate strategy, special controls and device-specific guidance.

• Define intended use, indications, specimen types and patient population.
• State whether the output is qualitative, semi-quantitative or quantitative.
• Define professional, point-of-care, over-the-counter or home use.
• Identify instrument, reagent, calibrator, control, software and data-interface boundaries.
• List every performance claim and clinically important cut-off.
• Determine whether a Q-Submission is needed to agree study design or acceptance criteria.

Use MTL-302 — US FDA Medical-device Regulations: 21 CFR Overview for the wider regulatory framework.

Organise standards around the performance questions

IVD submissions commonly address precision, bias or method comparison, linearity and measuring interval, analytical sensitivity, limit of blank, detection and quantitation, analytical specificity and interference, cross-reactivity, carryover, specimen stability, reagent stability, reference intervals, cut-off performance and clinical performance.

MTL-120 — Statistical Methods and Measurement Assurance provides the connected statistical foundation.

Typical standards and guidance families for FDA IVDs

FDA recognises many standards developed by CLSI, ISO, IEC and other organisations, sometimes wholly and sometimes only in part. Common candidates include CLSI documents for precision, method comparison, linearity, interference, detection capability, reference intervals, qualitative performance, stability and risk-based quality control.

Numbers alone are not enough. Verify the exact edition, FDA recognition record, extent of recognition, transition period and Supplemental Information Sheet before planning or claiming conformity.

Use FDA recognition and CLSI methods precisely

A familiar CLSI method is not automatically the FDA-recognised method for every product. The recognition may be partial, may exclude certain claims or technologies, or may require additional information. The standard's study design may also need adaptation to the proposed intended use.

• Search the FDA database by designation, keyword, product code and regulation number.
• Record the recognition number, edition and extent of recognition.
• Read the Supplemental Information Sheet and transition details.
• Compare the standard's specimen, concentration range and statistical assumptions with the device.
• Identify deviations before the study begins.
• Use a declaration of conformity only where the recognition and actual conformance support it.

When a standard is not recognised or is used outside its recognised scope, it may still support the submission under general use, but the scientific rationale and study information need to be sufficient for FDA review.

Create a claim-to-standard-to-evidence matrix

For each intended-use statement and performance claim, identify the regulatory basis, risk, standard or guidance, study design, sample source, comparator, range, acceptance criteria, statistical method, software and instrument configuration, report and labelling consequence.

This matrix prevents an apparently successful analytical study from supporting a claim broader than the tested population, specimen type, range or user setting.

Example: a quantitative point-of-care assay

A quantitative point-of-care assay may require standards-based evidence for precision across users and sites, method comparison across the measuring range, linearity, detection capability, interference, specimen stability and reference intervals. The instrument may also need electrical-safety, EMC, software, usability and cybersecurity evidence.

The protocol must reflect point-of-care conditions, operator training, environmental variation, sample handling and clinically important decision points. Laboratory-quality precision data generated by expert users in a controlled environment may not support a claim for non-laboratory users.

Use MTL-110 — Usability and Human Factors and MTL-105 — Systems Engineering, Architecture and Interfaces to connect workflow and system evidence.

Make the submission explain, not merely cite

• Identify the standard, edition, recognition number and extent of conformity.
• State the claim or review question each standard addresses.
• Describe specimens, sites, users, comparators, instruments and software versions.
• Provide pre-specified acceptance criteria and appropriate statistical analyses.
• Explain deviations, exclusions, outliers, invalid results and missing data.
• Connect conclusions to risks, special controls and proposed labelling.
• Supply supplementary evidence where the standard is incomplete for the device.
• Keep summary tables consistent with protocols, reports, datasets and labels.

A declaration of conformity may streamline part of the review, but it does not cure a poorly matched study design or unsupported intended-use claim.

Maintain performance evidence after clearance or approval

Monitor changes to recognised standards, CLSI and ISO methods, product codes, guidance and scientific state of the art. Assess reagent, calibrator, supplier, software, instrument, specimen, manufacturing and labelling changes against established claims and prior submission evidence.

Complaints, quality-control trends, invalid-result rates, bias shifts, new interferents, lot variation and cybersecurity information can trigger risk review, corrective action, labelling changes, additional studies or a new submission assessment.

Common misconceptions

Practical checklist

• Define the complete IVD system and all proposed claims.
• Confirm product code, class, pathway, special controls and guidance.
• Identify each analytical, clinical, system and usability evidence question.
• Check current FDA recognition records and Supplemental Information Sheets.
• Map methods to claims, risks, specimens, statistics and labelling.
• Agree deviations and adaptations before starting studies.
• Control instruments, software, reagents, lots, comparators and datasets.
• Use declarations of conformity only within the actual recognised scope.
• Maintain the mapping through changes and post-market learning.