MRBMaterial Review Board

A Material Review Board (MRB) is the formal, cross-functional decision body that determines what happens to any material — raw component, packaging, in-process intermediate, finished lot, returned product, or stranded inventory — that does not conform to its approved specification. The MRB is not a meeting; it is a controlled workflow that takes a documented nonconformance as input and produces a signed disposition as output. The disposition is one of a defined set of outcomes (use as-is, rework, regrade, return to vendor, scrap) and is bound by signed rationale to the spec, the test results, the impact assessment and the downstream batch record.

MRBs exist because the alternative — letting Production or Purchasing decide what to do with a failing lot — concentrates the conflict of interest in exactly the wrong place. Production has a schedule to make; Purchasing has a supplier relationship to preserve; only an independent body that includes Quality has the authority and the disinterest to render an objective disposition that survives a regulatory inspection. In every regulated industry the MRB is the named or de-facto venue where that independence lives.

MRB-style controls are codified across every major regulated-manufacturing framework, even where the literal phrase 'Material Review Board' does not appear. The common requirement is identification, segregation, evaluation by qualified personnel, documented disposition, and traceability of the decision through to the affected product.

The frameworks differ on detail but converge on six obligations: (1) segregate before deciding, (2) define the disposition outcomes in advance, (3) bring the right disciplines to the decision, (4) document the rationale and the evidence, (5) propagate the decision to inventory and the batch record, (6) feed the decision into a CAPA when patterns emerge.

Effective MRBs have a small, named, standing membership with a defined quorum and delegated authority for routine dispositions. A typical roster: a Quality chair (always required), a Manufacturing representative familiar with the process, an Engineering or Process Development representative, and as needed Regulatory Affairs (for any disposition that affects a filing or labelling), Supply Chain (for return-to-vendor) and the customer-facing function (for any finished-product disposition that could ship). For devices, AS9100 plants add a Customer Engineering authority for material disposition; for blood and tissue products, 21 CFR 1271 effectively adds a Medical Director.

• Chair: Quality, with delegated signing authority and an alternate.
• Manufacturing: a process owner who can speak to capability, not a generic operator.
• Engineering: someone with authority over the spec itself, not just the test method.
• Regulatory: for any disposition with a label, filing, or registered-spec impact.
• Supply Chain: for return-to-vendor and supplier-corrective-action workstreams.
• Customer / commercial: when the disposition could affect a contractually-bound spec.

Quorum should be defined in the SOP — typically chair plus two functions — and any disposition outside the standing authority (for example, a use-as-is on a registered drug spec, or a regrade across customer contracts) escalates to a named higher authority. This is the control that prevents 'MRB drift' where the same two people sign off on every disposition because the rest of the roster never shows up.

MRB outcomes should be a small, controlled list. Free-text dispositions create semantic drift in the audit trail and break the link between the MRB decision and the downstream inventory status. The canonical set is five, each with its own evidence rules.

1. Use as-is

The material is released for use despite the nonconformance, on the basis of documented impact assessment showing no risk to product quality, safety, efficacy or the regulatory filing. Use-as-is is the highest-scrutiny disposition: regulators read it as 'we did not meet our own spec but decided it was fine,' so the supporting evidence must be exceptional. Typical evidence: technical assessment by Engineering, historical data showing the parameter has no impact at the observed magnitude, customer concurrence where contractually required, and a CAPA opened for any recurring use-as-is on the same spec.

2. Rework

The material is reprocessed under a pre-approved rework instruction that brings it into conformance, with reassessment against the original spec after rework. Rework requires its own master instruction (not an ad-hoc procedure invented at the meeting), traceable rework batch record, and signed reassessment. ISO 13485 §8.3.2 and 21 CFR 820.90(b)(2) both require that the reworked product is re-evaluated; in practice this means the rework batch enters its own QA disposition step.

3. Regrade

The material is released against a different, lower-grade spec under which it does conform. Common in chemicals, plastics, ingredients and pharmaceutical excipients where the same lot can satisfy a USP-grade buyer or a technical-grade buyer at different price points. Regrade requires that the lower-grade spec is itself approved, that the lot meets every parameter of that spec (not just the failed one), and that downstream allocation prevents the regraded lot from being sold against the original grade.

4. Return to vendor (RTV)

The lot is returned to the supplier under a controlled RTV process with supplier corrective-action follow-up. RTV requires a printable manifest, a returned-goods authorisation (RGA) from the supplier, transport documentation, and posting against the supplier scorecard. The supplier corrective action runs in parallel and feeds the requalification interval for that supplier.

5. Scrap

The material is destroyed under a controlled scrap workflow. Scrap requires a witness (the second-person attestation that the material was actually destroyed, not diverted), a destruction record, and propagation to inventory and the genealogy of any work-in-process that consumed the same lot. For controlled substances, the destruction must follow DEA Form 41 procedures; for hazardous waste, EPA RCRA manifesting.

An MRB without an evidence package is a meeting where opinions outvote facts. The standard evidence package for any disposition should be assembled before the MRB convenes, attached to the disposition record, and reviewable as a single object by every member.

1. The nonconformance record itself: who observed it, when, on which lot, against which spec version.
2. The failing test result, including raw data and the analyst attribution.
3. The full COA or release dossier for the lot, including any out-of-trend observations on adjacent parameters.
4. Process history: what batch consumed or would consume this material, what equipment was involved.
5. Impact assessment: technical evaluation of effect on quality, safety, efficacy, regulatory filing and customer contract.
6. Historical context: have we seen this nonconformance before, on this supplier, on this spec.
7. Any open CAPA or change control on the affected material or supplier.
8. Proposed disposition with rationale, drafted by the requesting function and reviewed by Quality.

Across decades of FDA 483s, EU GMP non-compliance reports, ISO 13485 NCRs and AS9100 findings, the same MRB failure modes recur. Knowing them in advance lets you design the workflow to make them structurally impossible rather than relying on operator vigilance.

• Use-as-is with no impact assessment beyond a signature — the most common 483 pattern in the disposition space.
• Material released to production before the MRB record was signed (the workflow allowed it; in a paper system it usually did).
• Rework instructions invented at the meeting rather than pre-approved.
• Regrade decisions that did not retest the lot against every parameter of the lower-grade spec.
• Return-to-vendor with no supplier-corrective-action follow-up — the same nonconformance recurs on the next shipment.
• Scrap with no witness — the destruction record exists but the second-person attestation does not.
• Repeat MRBs on the same supplier or root cause that never escalated to a CAPA.
• Disposition that did not propagate to the batch record — the eBMR/eDHR says the material was released, the MRB record says it was reworked.
• MRB members who are not trained on the SOPs that govern the disposition.
• MRB log that cannot be filtered by supplier, by spec, by root cause, by disposition — so trend analysis is impossible.

The pattern across all of these is the same: the MRB became a paperwork exercise rather than a controlled workflow. Modern MES designs eliminate most of them structurally — the material cannot be moved out of quarantine without a signed disposition, the rework cannot start without an approved instruction, the disposition cannot be saved without the impact assessment populated. The MRB becomes a decision-making body again, not a signature-gathering body.

A single MRB disposition is a tactical decision. A pattern of MRB dispositions on the same root cause, the same supplier, the same spec or the same product family is a strategic signal. ICH Q10 and 21 CFR 820.100 both expect that a quality system uses the trend, not just the individual nonconformance, to drive corrective action.

Practical escalation triggers: three MRBs on the same supplier inside a rolling 12 months; two MRBs on the same spec parameter inside a rolling six months; any MRB disposition that required higher-than-routine authority (use-as-is on a registered drug spec, for example); any disposition where the rework or regrade affected the customer's contracted spec. Each trigger should automatically open a CAPA, not wait for someone to notice the pattern.

Pharmaceuticals and dietary supplements

21 CFR 211 and 21 CFR 111 require that any reprocessing or reworking of a finished or in-process lot is supported by an approved protocol and that the reworked lot is reassessed. The MRB cannot release a reworked lot without a signed reassessment against the original spec. Out-of-spec (OOS) results that drive an MRB must follow the OOS investigation procedure under FDA's 2006 OOS Guidance before disposition.

Medical devices and IVDs

21 CFR 820.90 requires that nonconforming product is evaluated and dispositioned, that the rework includes reassessment, and that the documentation includes the nature of the nonconformance, the disposition, the rationale and the personnel authorising it. EU MDR Article 10(9) and ISO 13485 §8.3 carry equivalent expectations. For implants and active devices the MRB record is part of the Design History File and Device Master Record traceability.

Food and beverage

21 CFR 117 and FSMA 204 add a food-safety impact dimension: any disposition that could release potentially unsafe product must include a food-safety assessment by the Preventive Controls Qualified Individual. ISO 22000 §8.9 codifies the same expectation internationally. For RTE meat under 9 CFR 430, an MRB on a Zone 1 Listeria event is a recall-prep workflow, not an inventory decision.

Aerospace and automotive

AS9100 expects engineering-authority concurrence on any material disposition that affects form, fit or function. IATF 16949 requires customer notification before shipping nonconforming product released under a deviation permit, and any disposition that affects the PPAP must trigger a PPAP submission.

Every field on this list has been the subject of an FDA 483 at some point. The record is the disposition; the meeting is the means by which the record was created.

In V5 the MRB is wired into the same workflow that quarantined the material. When a receiving inspection, in-process check, OOS investigation or QA review opens a nonconformance, V5 segregates the affected inventory (status flip in WMS, lock in MES), assembles the evidence package automatically (failing test, COA, batch history, supplier history, prior MRBs on the same spec or supplier) and routes the case to the standing MRB roster with the configured quorum.

• Evidence package auto-assembled — no manual collation from five systems.
• Quorum enforced — disposition cannot be signed without the configured roles present.
• Disposition outcomes are a controlled list — no free-text dispositions that drift in meaning.
• Rationale and impact-assessment fields are mandatory before signing.
• E-signatures meet 21 CFR Part 11 with role checks, identity verification and immutable audit trail.
• Downstream propagation is atomic — lot status, batch record, label print and CAPA queue all update together.
• Escalation patterns (repeat supplier, repeat spec, registered-product impact) auto-open a CAPA with the prior dispositions linked.
• Effectiveness checks for rework or supplier-corrective-action dispositions schedule themselves and chase their own closure.

Most MRB metric dashboards count meetings. The metrics that actually correlate with quality outcomes count something else: the disposition mix, the cycle time, the repeat-pattern rate, and the effectiveness of follow-up.

• Disposition mix by month — a rising use-as-is share is almost always a sign of degrading process capability or spec drift.
• Median and 90th-percentile cycle time from nonconformance open to disposition signed — long tails point to evidence-assembly bottlenecks.
• Repeat-supplier rate — what proportion of MRBs are on suppliers we have MRB'd in the past 12 months.
• Repeat-spec rate — what proportion are on parameters we have MRB'd in the past 6 months.
• Escalation-to-CAPA conversion rate — repeats should become CAPAs; if they do not, the escalation rule is broken.
• Effectiveness-check pass rate on rework dispositions — failing effectiveness checks are the leading indicator of CAPA effectiveness erosion.

Tracked over rolling 13-week windows, these metrics tell you whether the MRB is a healthy decision body or a process for documenting the same failures month after month.

When an FDA, EMA, MHRA, BRCGS or ISO auditor opens the MRB workstream, they ask a predictable sequence of questions. Being able to answer them in seconds, from a single record, is the difference between a clean inspection and a deep-dive.

1. Show me your MRB log for the last 12 months, filterable by disposition outcome.
2. Pick three use-as-is dispositions and walk me through the impact assessment.
3. Pick a rework disposition and show me the pre-approved rework instruction and the reassessment record.
4. Show me a supplier where you have had multiple MRBs and the corresponding CAPA.
5. Show me an MRB where the material was used before the disposition was signed (this is the trap question — the answer should be 'we cannot, the workflow prevents it').
6. Show me your training records for MRB members on the disposition SOP.
7. Show me a disposition where the regulatory impact required higher-than-routine authority and the approval path.

An MRB workflow built on a modern MES answers all seven in under a minute, from one screen, with the underlying records one click away. A paper MRB binder system answers them in three to five days, with the inspector watching.