Reagent Stability: Shelf Life, In-Use and Shipping Evidence
Connect shelf-life, in-use and shipping claims to the reagent configuration, conditions and completed evidence that actually support them.

On this page · 9 sections
- 1What Reagent Stability means
- 2Separate the claims before designing the studies
- 3US law: storage instructions and assurance at time of use
- 4EU law: three distinct stability documentation categories
- 5Move from a proposed claim to an evidence-supported conclusion
- 6Fictional example: passing observations with a remaining gap
- 7Make every conclusion traceable to its product and observations
- 8Common mistakes and the trade-offs behind better choices
- 9Practical implementation checklist
How does Reagent Stability apply to your shop floor?
Pick your industry and scale — Ask V5 rewrites the definition in your context, gives a worked example, and shows what V5 does on day one.
01What Reagent Stability means
Reagent Stability is the ability of a diagnostic reagent to retain its specified performance under defined conditions over a defined period. Stability evidence demonstrates that ability for a particular product configuration and use situation. It is not simply an expiry date, a normal-looking solution or a plan to test the product later.
For an in vitro diagnostic product, the relevant performance characteristics depend on the reagent and its intended use. As a study-design practice, endpoints should be capable of detecting deterioration that matters to the assay: for example, changes in response, background, precision or performance near a decision boundary. Physical appearance may be relevant, but an unvalidated visual check cannot establish every aspect of analytical performance.
Stability is better understood as a set of bounded claims than as three numbers on a kit. Those bounds include duration, temperature, packaging, preparation, instrument exposure and handling. Shelf life, open-vial use, on-board use and shipping may interact, but evidence for one does not automatically establish the others. This article concerns reagent and device stability, not specimen stability.
02Separate the claims before designing the studies
A useful planning practice is to write each proposed claim with its starting event, conditions and endpoint. “Stable for 30 days” is incomplete unless it specifies whether the clock begins at opening, reconstitution or loading onto an analyser, and what happens between uses.
| Evidence question | Claim boundary to define | What cannot be assumed |
|---|---|---|
| Unopened shelf life | The dating reference, duration, storage range and final container configuration | Fresh-product performance alone does not establish performance at the proposed expiry. |
| Open-vial or reconstituted use | Opening or preparation event; handling, closure and storage between uses | An unopened study does not reproduce repeated access, preparation or exposure. |
| On-board use | Instrument environment, loading event, access cycles and permitted interruptions | A refrigerated open-vial result is not automatically an on-board result. |
| Shipping tolerance | Packaging and anticipated transport exposures, including relevant extremes | A transport result does not automatically establish subsequent shelf life or in-use duration. |
| Calibration stability | The claimed calibration interval and relevant instrument/reagent conditions | Reagent stability and calibration stability are related but distinct evidence questions. |
Interactions deserve explicit attention. A recently manufactured vial tested after opening may not represent a vial opened near the end of its shelf life. Similarly, a shipment challenge performed only on fresh product leaves uncertainty about later performance. Combining ageing, transport and use exposures can improve relevance, but increases study complexity. Separate studies are easier to interpret; any bridge between them needs a stated rationale rather than an assumption.
03US law: storage instructions and assurance at time of use
21 CFR 809.10(a) sets out IVD label information with applicability qualifications and exceptions. For reagents, paragraph (a)(5) requires appropriate storage instructions adequate to protect stability, including temperature, light, humidity and other pertinent factors when applicable. It also addresses storage instructions for reconstituted or mixed product that is to be stored in the original container. The basis for those instructions must be determined by reliable, meaningful and specific test methods.
Paragraph (a)(6) is not an expiration-date-only rule. It requires a means of assuring the user that the reagent meets appropriate standards of identity, strength, quality and purity at the time of use, both as provided and after applicable reconstitution or mixing. The label must include one or more of these alternatives:
- An expiration date based on the stated storage instructions.
- A statement of an observable indication that the product has altered beyond its appropriate standards, such as turbidity, colour change or precipitate.
- Instructions for a simple method by which the user can reasonably determine that the product meets its appropriate standards.
Accompanying labeling has its own applicability qualifications. Paragraphs (b)(5)(iv) and (vi) address appropriate storage instructions, including for reconstituted or mixed product, and physical, biological or chemical indications of instability or deterioration. These are legal labeling provisions—not a universal study schedule, sample count or statistical protocol. The alternatives in (a)(6) should be read within the section’s applicability provisions and product-class qualifications.
04EU law: three distinct stability documentation categories
IVDR Annex II, section 6.3, in the 10 January 2025 consolidated text, separates claimed shelf-life, in-use and shipping stability studies. These are EU technical-documentation requirements, not guidance and not a US protocol. The lot provisions differ by study category.
| Category | Lot and exposure provisions | Documentation specified |
|---|---|---|
| Claimed shelf life — 6.3.1 | At least three different lots manufactured under conditions essentially equivalent to routine production. They need not be consecutive. Accelerated studies or extrapolated real-time data are acceptable for initial claims, followed by real-time stability studies. | Study report including protocol, number of lots, acceptance criteria and testing intervals; accelerated-study method where performed in anticipation of real-time studies; conclusions and claimed shelf life. |
| In-use stability — 6.3.2 | Studies for one lot reflecting actual routine use, whether real or simulated. This may include open-vial and/or on-board stability. Supporting calibration data are included if calibration stability is claimed for automated instrumentation. | Study report including protocol, acceptance criteria and testing intervals; conclusions and claimed in-use stability. |
| Shipping stability — 6.3.3 | Studies for one lot evaluating anticipated shipping conditions. Real and/or simulated studies include variable conditions such as extreme heat and/or cold. | Study report including protocol and acceptance criteria; method used for simulated conditions; conclusion and recommended shipping conditions. |
The initial shelf-life allowance is therefore specific: accelerated or extrapolated evidence can support an initial EU claim, but real-time follow-up is required. It is not a general promise that any elevated-temperature experiment supports provisional dating in every jurisdiction. As an implementation practice, choose additional coverage according to product risks and unresolved variability; do not treat the stated lot provisions as a complete experimental design.
05Move from a proposed claim to an evidence-supported conclusion
As a study-design practice, define the product configuration, endpoints, acceptance criteria and analysis approach before testing. Choose time points and sample coverage to answer the actual claim, rather than adopting a universal calendar. Real-time observations directly measure performance during elapsed storage under the study conditions. Accelerated work can inform an initial claim or development decision, but the inference depends on whether the stress represents relevant deterioration mechanisms.
A hotter condition can change the degradation mechanism, damage packaging or produce effects that do not scale simply with time. Conversely, a convenient assay endpoint may miss deterioration elsewhere in the measurement system. An extrapolation should therefore state its assumptions, uncertainty and limits; a mathematical fit does not independently establish suitability.
- 1. Define the proposed claimSpecify the product, duration, conditions and starting event.
- Claim boundary
- Product configuration
- 2. Establish the study planDefine lots, exposures, endpoints, criteria and analysis.
- Protocol revision
- Planned observations
- 3. Execute and recordCapture actual exposures, measurements and deviations.
- Completed pulls
- Raw results
- 4. Review the evidenceEvaluate validity, trends, uncertainty and missing coverage.
- Report conclusions
- Unresolved gaps
- 5. Assess the claimSeparate supported conclusions from proposed durations and conditions.
- Claim assessment
- Applicable decision process
- 6. Maintain follow-upComplete outstanding work and assess changes or adverse findings.
- Follow-up obligations
- Change assessment
Keep planned, completed, reviewed and claim-supporting states distinct in records. A completed measurement may still have an unresolved deviation. A reviewed study may support only a narrower condition than originally proposed. Neither a passing release test nor an entry on a stability schedule fills a missing evidence interval.
06Fictional example: passing observations with a remaining gap
Suppose a manufacturer proposes 12 months of unopened refrigerated shelf life. One fictional routine-production-equivalent lot, L1, has completed observations through month 9. For this simplified example, the protocol defines one illustrative endpoint as signal retained relative to time zero, with an acceptance range of 90–110%. Assume the observations through month 9 have been reviewed without unresolved deviations.
| Time | Signal, arbitrary units | Retained signal: result ÷ 200 × 100 | Evidence state |
|---|---|---|---|
| Month 0 | 200 | 100% | Completed baseline |
| Month 3 | 198 | 99% | Completed; within illustrative range |
| Month 6 | 194 | 97% | Completed; within illustrative range |
| Month 9 | 188 | 94% | Completed; within illustrative range |
| Month 12 | Not measured | Not calculable | Planned; no result |
At month 9, retained signal is 188 ÷ 200 × 100 = 94%. The decrease from baseline is (200 − 188) ÷ 200 × 100 = 6%. The proposed 12-month claim extends 12 − 9 = 3 months beyond the completed real-time observations. Being four percentage points above the illustrative lower limit does not convert into a number of additional stable months.
The defensible conclusion is narrow: this lot met this illustrative endpoint at the measured times through month 9. That does not establish a complete nine-month shelf-life claim, much less the proposed 12 months. Other relevant performance endpoints remain outside this example. For the EU shelf-life documentation described above, at least two additional different routine-production-equivalent lots would also be missing. No in-use or shipping evidence has been supplied, and no justified accelerated or extrapolation analysis closes the duration gap.
07Make every conclusion traceable to its product and observations
A practical stability record should allow a reviewer to move from a claimed condition to the report, underlying observations, tested lots and product configuration. Reagent lot traceability matters because a lot identifier alone may not explain the formulation revision, container closure, component sources or manufacturing conditions represented by a study.
Candidate claim C1
References configuration P1 and assessment in report R1.
- Proposed duration and conditions
- Claim status
Configuration P1
Defines the product represented by lot L1.
- Formulation revision
- Container and relevant instrument configuration
Tested lot L1
Links to P1 and the corresponding production record.
- Lot identity
- Manufacturing context
Protocol S1
Defines the planned investigation of C1.
- Methods and criteria
- Planned exposures and pulls
Observations O1
Reference L1 and S1; distinguish actual testing from planned work.
- Measurements and exposure records
- Deviations and completion dates
Report R1
Assesses C1 using S1 and O1, including missing coverage.
- Reviewed conclusions
- Limits and unresolved gaps
For kit dating, a useful practice is to check component constraints and the evidence for the assembled configuration. Simply selecting the earliest component date does not, by itself, demonstrate assembled-kit performance. Keep the stability evidence distinguishable from the production history while linking both in the IVD manufacturing record dossier. A formulation or packaging change should prompt an evidence-applicability assessment, not automatic reuse of the old conclusion.
08Common mistakes and the trade-offs behind better choices
| Shortcut | Why it can mislead | Better practice and trade-off |
|---|---|---|
| Treat accelerated time as a fixed multiple of real time | The stress may alter the deterioration mechanism. | Justify the model and its limits; stronger understanding takes development effort. |
| Test only one convenient endpoint | An apparently stable response may conceal other relevant performance changes. | Use risk-relevant endpoints; broader coverage increases testing burden. |
| Study fresh opened product only | The result may not represent opening near the end of shelf life. | Address relevant age/use interactions; combined challenges can be harder to interpret. |
| Treat ordinary appearance as proof of stability | A casual visual inspection is not equivalent to a justified labeled indicator or analytical evidence. | Establish what an indicator detects and what it does not; retain other relevant evidence. |
| Treat a passing retest as erasing a failed pull | The original result and its cause remain relevant to the conclusion. | Preserve both results and investigate; extra testing does not automatically resolve validity or scope. |
As a quality practice, an unexpected or failed pull should trigger an assessment of test validity, storage history, handling, product change and potentially affected populations. It does not automatically establish that every lot is affected—or that the finding is confined to the tested lot. Escalate through the applicable quality process, using the evidence to assess scope and necessary action rather than assuming an automatic expiry extension, recall or notification outcome.
09Practical implementation checklist
Use this practice checklist when establishing or revising a stability programme. It is a planning aid, not a substitute for applicable legal requirements or a product-specific protocol.
- Write separate shelf-life, preparation, open-vial, on-board, shipping and calibration claims where relevant; define each clock and condition.
- Identify the formulation, packaging, instrument configuration and manufacturing conditions represented by the evidence.
- Map applicable jurisdictional requirements separately from internal study-design choices.
- Justify lot coverage, endpoints, criteria, time points and analysis without assuming a universal schedule.
- Record actual exposures and completed observations separately from planned pulls and proposed claims.
- Review deviations, trends, uncertainty and missing coverage before drawing conclusions.
- Link each conclusion to its protocol, observations, lot genealogy and report; identify which labeling condition it supports.
- Route changes and adverse findings through an applicability assessment, and keep required follow-up visible.
During medical-device design transfer, make evidence applicability an explicit handover question: does the tested configuration represent what routine production will make? An orderly dossier cannot compensate for missing data, but it makes the boundary between demonstrated performance and unfinished work easier to see.
Frequently asked questions
Q.Is reagent stability the same as shelf life?+
No. Shelf life concerns a defined storage period before use. Reagent stability also covers relevant preparation, open-vial, on-board and shipping conditions. Each conclusion is bounded by the conditions and performance characteristics studied.
Q.Does 21 CFR 809.10 require only an expiration date?+
No. Subject to its applicability provisions, paragraph (a)(6) specifies one or more alternatives: an expiration date, a statement of an observable alteration beyond appropriate standards, or instructions for a simple method of checking those standards. Storage-instruction provisions also apply.
Q.Can accelerated studies replace real-time studies?+
Not under the EU Annex II 6.3.1 provision discussed here: accelerated studies or extrapolated real-time data are acceptable for initial shelf-life claims, but must be followed by real-time stability studies. That allowance should not be generalized into a universal dating rule.
Q.Must the three EU shelf-life lots be consecutive?+
No. Annex II 6.3.1 specifies at least three different lots manufactured under conditions essentially equivalent to routine production; it expressly says they need not be consecutive.
Q.Are open-vial and on-board stability interchangeable?+
Not automatically. Instrument temperature, repeated access, evaporation and handling may differ from off-instrument storage. A bridge between conditions needs a product-specific justification.
Q.Can shipping exposure simply be subtracted from remaining shelf life?+
Not without a justified relationship between exposure and subsequent performance. A passing shipping study is not, by itself, a formula for calculating remaining shelf life or permission to extend an expiry date.
Primary sources
Further reading
- Reagent Lot TraceabilityConnect tested material to the manufacturing context represented by the evidence.
- IVD Manufacturing Record DossierPlace stability evidence alongside related manufacturing records without confusing their purposes.
- Medical Device Design TransferConsider whether development evidence represents the configuration entering routine production.
Want to see how Reagent Stability could fit into your own records and workflows? Explore the related V5 pages or talk to our team about what applies to your operation.
