This is a ready-to-use risk assessment for deciding whether a fleet of identical instruments can be qualified through a representative-unit (bracketing) strategy, and if so, exactly what is bracketed and what is not. It turns “these are all the same model, let’s not repeat the whole qualification five times” from an informal shortcut into a documented, QA-approved decision with a fixed floor that never moves. Replace every <<FILL: ...>> placeholder with your own specifics and route it through your validation and QA review procedures. A worked specimen follows. Verify each cited regulation against the current source before you rely on it. This content is educational reference, not regulatory advice.
Document control header
| Field | Entry |
|---|---|
| Assessment title | Instrument Fleet Bracketing Strategy, <<FILL: instrument make/model>> |
| Document number | <<FILL: RA-ID>> |
| Version | <<FILL: version>> |
| Effective date | <<FILL: date>> |
| Facilitator / author | <<FILL: metrology / instrument owner>> |
| Cross-functional reviewers | <<FILL: QA, metrology, affected site leads>> |
| Fleet covered (asset IDs and locations) | <<FILL: list or reference to inventory extract>> |
| Linked instrument group classification | <<FILL: form reference, per unit or per model>> |
1. Objective and scope
Decide, for the fleet of instruments named above, which units are qualified as representative units to the full extent, which units are bracketed to a reduced, unit-specific scope, and which qualification elements are never eligible for reduction regardless of how identical the fleet appears on paper. This assessment does not authorize skipping any unit’s qualification; it scopes how much of each unit’s qualification is unit-specific versus inherited from the representative unit or units. It does not alter the acceptance criteria, execution rules, or documentation requirements of <<FILL: cross-reference IQ/OQ/PQ protocol SOP-IDs>>, which still govern whatever scope is executed on each unit.
2. Eligibility gate: can this fleet be bracketed at all
Bracketing is only available if every question below is answered Yes. Any No stops the assessment; qualify every unit individually.
| # | Question | Answer | Evidence |
|---|---|---|---|
| 1 | Are all units the identical make and model? | <<FILL: Yes/No>> | <<FILL: asset register extract>> |
| 2 | Do all units run the identical firmware and, where applicable, control/data-system software version? | <<FILL: Yes/No>> | <<FILL: version log>> |
| 3 | Are all units configured identically (no unit has an added module, sensor, or software option the others lack)? | <<FILL: Yes/No>> | <<FILL: configuration record>> |
| 4 | Do all units support the same intended use and run the same method set, with no unit assigned a tighter-tolerance method the others were never tested against? | <<FILL: Yes/No>> | <<FILL: method assignment list>> |
| 5 | Is each unit’s group classification (USP <1058> Group A/B/C) the same across the fleet? | <<FILL: Yes/No>> | <<FILL: classification form references>> |
Gate result: <<FILL: eligible for bracketing / not eligible, qualify individually>>
3. What is never bracketed
Regardless of the gate result, the following are unit-specific in every case and are never inherited from a representative unit:
- Installation Qualification. Each unit’s own environment, utilities, and as-installed configuration are verified at its own physical location. An IQ performed at one site provides no evidence about a different site’s environment.
- Calibration. Each physical unit is calibrated in its own right against traceable standards on its own schedule. Two identical units drift differently once in service.
- Any environment class that differs from the representative unit’s. A unit sitting in a different temperature, humidity, or vibration environment carries its own environmental verification even while other elements stay bracketed.
4. Selecting representative units
| Field | Entry |
|---|---|
| Number of units in the fleet | <<FILL>> |
| Number selected as representative | <<FILL>> |
| Selection basis | <<FILL: e.g. one per distinct site/usage pattern, the highest-throughput unit per environment class>> |
| Units NOT eligible to represent others (why) | <<FILL: e.g. a unit with a documented history of repeated repairs, or the sole unit at an outlier environment>> |
A single representative unit for the whole fleet is defensible only when every remaining unit shares both the same environment class and a comparable usage pattern as the representative unit. Where sites differ meaningfully in environment or usage intensity, select one representative unit per distinct condition rather than one for the entire fleet.
5. Scoping table: qualification elements by unit role
| Qualification element | Representative unit(s) | Bracketed unit(s) |
|---|---|---|
| Design Qualification | Full, written once for the model | Referenced from the representative unit’s DQ, not repeated |
| Installation Qualification | Full, at own site | Full, at own site (never reduced) |
| Operational Qualification | Full, across the rated range | Confirmatory subset at the working range actually used (<<FILL: which specific tests>>) |
| Performance Qualification | Full, against the methods run | Full execution against the same methods, but acceptance basis and rationale inherited rather than rederived |
| Calibration | Full, own schedule | Full, own schedule (never reduced) |
| Computerized-system controls (if Group C) | Full, per unit’s own instance | Full, per unit’s own instance (audit trail and access control are never bracketed) |
6. Trigger check: what breaks the bracket
| Event | Effect |
|---|---|
| A firmware or software update applied to fewer than all units in the fleet | The updated unit(s) fall out of the bracket immediately; requalify against the new configuration before use |
| A unit’s environment changes (relocation, HVAC change, new adjacent equipment) | That unit’s environmental basis is reassessed; other units are unaffected |
| A unit is assigned a new method with tighter requirements than the bracket was tested against | That unit requires its own PQ demonstration against the new requirement; the bracket does not cover it |
| A representative unit is decommissioned or relocated | Select a new representative unit and reconfirm the basis for every bracketed unit that relied on it |
| Any unit shows an OOT, repeated repair, or performance drift the others do not | That unit is pulled from the bracket and investigated individually; do not assume the fleet-wide basis still holds |
7. Decision and approval
Decision: <<FILL: bracketed per the scoping table above / not eligible, individual qualification for all units>>
| Role | Name | Signature | Date |
|---|---|---|---|
| Metrology / instrument owner (facilitator) | <<FILL>> | ||
| QA | <<FILL>> | ||
| Site lead(s) for bracketed units | <<FILL>> |
QA approval of the bracketing rationale is mandatory before any unit’s qualification is executed under a reduced scope; the qualified-state claim for a bracketed unit rests on this record.
8. Acceptance criteria for this assessment
- The eligibility gate in section 2 was completed and answered before any scoping decision was made.
- Section 3’s never-bracketed elements are confirmed unaffected for every unit, with no exceptions.
- The representative-unit selection basis in section 4 reflects actual site and usage differences, not administrative convenience.
- The scoping table names the specific confirmatory OQ tests applied to bracketed units; a blanket statement such as “reduced OQ” without a named test list is not acceptable.
- The trigger table in section 6 is referenced by the site’s change-control procedure so a firmware update, relocation, or method change automatically reopens this assessment for the affected unit.
- QA approved the assessment before any bracketed-scope qualification was executed.
9. References
USP General Chapter <1058>, Analytical Instrument Qualification (proportionality and group classification principles). EU GMP Annex 15, Qualification and Validation (bracketing and worst-case qualification approaches). ICH Q9(R1), Quality Risk Management. ASTM E2500, Standard Guide for Specification, Design, and Verification of Pharmaceutical and Biopharmaceutical Manufacturing Systems and Equipment.
Confirm the current version and clause numbers of each reference before issue.
Filled specimen
The following shows the assessment completed for an illustrative fleet of five identical analytical balances, one make and model, across three QC sites. Asset IDs and numbers are invented.
Eligibility gate: All five units are the identical make and model (Q1 Yes), running identical firmware with no data-system attached (Q2 Yes), no unit has an added module or sensor (Q3 Yes), all five weigh samples for the same class of assay methods with no unit assigned a tighter-tolerance method (Q4 Yes), and all five are classified Group B (Q5 Yes). Gate result: eligible for bracketing.
Representative unit selection: Two units selected, one per site with a meaningfully different usage pattern (Site A, highest throughput; Site B, moderate throughput). Site C’s single unit is not eligible to represent the others because it is the only unit in a different environment class (a non-climate-controlled annex).
| Asset ID | Site | Role | Qualification scope |
|---|---|---|---|
| LAB-0501 | Site A | Representative | Full DQ (for the model), full IQ, full OQ across the rated range, full PQ |
| LAB-0502 | Site B | Representative | Full IQ, full OQ across the rated range, full PQ (DQ referenced from LAB-0501) |
| LAB-0503 | Site A | Bracketed | Full IQ (own site), calibration, confirmatory OQ (repeatability and linearity at the working range only), PQ against the same method |
| LAB-0504 | Site B | Bracketed | Full IQ (own site), calibration, confirmatory OQ (repeatability and linearity at the working range only), PQ against the same method |
| LAB-0505 | Site C | Bracketed, different environment | Full IQ including Site C-specific environmental verification, calibration, confirmatory OQ, PQ |
Decision: Bracketed per the table above. QA approved 2026-09-10. Trigger table referenced in change control CC-2026-0417 so any firmware update, relocation, or new method assignment reopens this assessment for the affected unit only.
Common inspection findings this assessment prevents
- Five “identical” units qualified independently with no documented rationale for why some got full testing and others did not, or worse, all five reduced uniformly with no basis at all.
- A firmware update rolled out to some units in a fleet with no reassessment of whether the bracket still holds for the updated units.
- Installation Qualification or calibration treated as inheritable across units, when both are always unit-specific.
- “Same model number” accepted as sufficient without checking firmware, software version, or configuration, when one unit in the fleet has quietly diverged.
- A bracketing decision made informally by whoever ran the qualification, with no QA approval on record before execution.
How to adapt this assessment
- Run the eligibility gate in section 2 first; if any answer is No, the rest of the assessment is unnecessary and every unit is qualified individually.
- Set your representative-unit selection basis from real differences in environment and usage across your fleet, not from which unit happens to be easiest to schedule.
- Name the specific confirmatory OQ tests applied to bracketed units in section 5; do not close this assessment with a vague “reduced scope.”
- Link the trigger table in section 6 to your change-control system so a firmware update or relocation automatically reopens the assessment for the affected unit.
- Get QA approval before any bracketed-scope qualification executes, and file the signed assessment alongside the qualification protocols it governs.