This is a ready-to-use protocol. Replace every <<FILL: ...>> placeholder with your own specifics, execute per your validation program, and route through your normal review and approval. A worked filled specimen follows the template. Verify each cited regulation against the current source before you rely on it.
Approval page
| Field | Entry |
|---|---|
| Protocol title | Swab and Rinse Recovery Study for <<FILL: residue name(s)>> on <<FILL: equipment / material of construction>> |
| Protocol number | <<FILL>> |
| Parent cleaning validation protocol | <<FILL: protocol number>> |
| Version | <<FILL>> |
| Effective date | <<FILL>> |
| Role | Name | Signature | Date |
|---|---|---|---|
| Author | <<FILL>> | ||
| QC analytical reviewer | <<FILL>> | ||
| QA approver | <<FILL>> |
1. Objective
To establish, with documented experimental evidence, the recovery factor for <<FILL: residue(s)>> from <<FILL: surface material(s)>> using the qualified sampling technique defined in <<FILL: sampling plan reference>>, so that field swab and rinse results can be corrected to reflect the true residue present on the equipment surface.
2. Scope
This study covers <<FILL: residue(s) and cleaning-agent(s) in scope>> on <<FILL: surface material(s) and finish, e.g. 316L stainless steel, 2B finish; glass; PTFE>>, for both swab and rinse methods where both are used in the parent cleaning validation protocol. A separate recovery study is required for each distinct combination of residue and surface material; do not extrapolate a factor established on one material to another.
3. Prerequisites
- The field sampling technique (swab material, solvent, stroke pattern, swabbed area template; rinse volume and collection method) is finalized in
<<FILL: sampling plan reference>>before this study starts, because the recovery study must mirror it exactly. - The analytical method is validated per ICH Q2(R2), with a limit of quantitation below the anticipated recovery-corrected acceptance limit.
- Coupons of the relevant surface material and finish are available, matching the actual equipment surface, not a generic substitute.
4. Roles
| Role | Responsibility |
|---|---|
| QC analytical | Prepares spiked coupons, performs swabbing/rinsing per the qualified technique, extracts and assays, calculates recovery |
| Validation | Designs the spike levels and replicate scheme, reviews the calculation, feeds the factor into the parent protocol |
| QA | Approves the protocol and the recovery factor before it is used to correct field results |
5. Method
5.1 Coupon preparation
- Obtain coupons of
<<FILL: material and finish>>matching the equipment surface at the sampled location. - Clean coupons to a verified residue-free baseline and confirm by blank assay before spiking.
5.2 Spike levels and replicates
Spike at levels bracketing the acceptance limit, in replicate:
| Level | Target (% of swab/rinse limit) | Amount spiked | Replicates |
|---|---|---|---|
| Low | <<FILL: e.g. 50%>> | <<FILL>> | <<FILL: e.g. n=3>> |
| Target | <<FILL: e.g. 100%>> | <<FILL>> | <<FILL: e.g. n=3>> |
| High | <<FILL: e.g. 150%>> | <<FILL>> | <<FILL: e.g. n=3>> |
5.3 Drying
Allow the spike to dry under conditions representative of real residue formation (<<FILL: time, temperature>>) before sampling. Do not sample a wet spike; wet recovery does not represent field conditions and will overstate the factor.
5.4 Sampling (swab)
Swab each coupon using the exact technique specified in <<FILL: sampling plan reference>>: swab type <<FILL>>, wetting solvent <<FILL>>, stroke pattern <<FILL>>, number of strokes <<FILL>>. Any departure from this technique invalidates the resulting factor for use against field samples taken by the qualified technique.
5.5 Sampling (rinse)
Rinse each coupon (or a representative surface) using the exact rinse volume, contact time, and collection method specified in <<FILL: sampling plan reference>>.
5.6 Extraction and assay
Extract each swab or rinse sample and assay per <<FILL: validated method reference>>. Include swab and vial blanks with every batch and subtract background per the documented convention.
5.7 Calculation
Recovery (%) = (amount recovered / amount spiked) x 100
Calculate recovery for each replicate, then the mean and %RSD per level. Round the applied recovery factor DOWN (conservative direction); never round up.
6. Acceptance criteria
| Parameter | Acceptance |
|---|---|
| Mean recovery | <<FILL: e.g. >= 70%, target >= 80%>>; recovery 50 to 70% usable only with documented scientific justification; recovery < 50% requires investigation before use |
| %RSD across replicates | <<FILL: e.g. <= 15 to 20%>> |
| Blanks | Below the method’s LOQ, or subtracted per the documented convention with the subtraction shown in the calculation |
7. Recovery-correction convention
State explicitly which convention this study’s factor feeds into the parent protocol: <<FILL: either "the swab limit is divided by the recovery factor to give a recovery-corrected acceptance limit" OR "the measured field result is divided by the recovery factor to give a recovery-corrected result, compared against the uncorrected limit">>. Use one convention consistently across the parent protocol and this study; mixing the two is a common source of wrong disposition.
8. Deviation handling
Any replicate outside the expected range, any blank above LOQ, or any %RSD outside acceptance is investigated per <<FILL: SOP-ID for deviations>> before the factor is finalized.
9. Summary and conclusion
<<FILL: state the established recovery factor(s), the material and residue they apply to, and confirm they are ready for use in the parent cleaning validation protocol>>.
References
EU GMP Annex 15 (Qualification and Validation), 2015 revision (recovery established for both swab and rinse methods). FDA Guide to Inspections Validation of Cleaning Processes (1993). ICH Q2(R2), Validation of Analytical Procedures.
Confirm the current version and clause numbers of each reference before issue.
Filled specimen
Objective (excerpt): Establish the swab recovery factor for Product A active on 316L stainless steel, 2B finish, feeding cleaning validation protocol CV-2026-014.
| Level | Target | Amount spiked | Replicates |
|---|---|---|---|
| Target (100%) | 100% of swab limit | 625 ug on a 25 cm2 coupon | n=3 |
| Replicate | Recovered (ug) | Recovery (%) |
|---|---|---|
| 1 | 455 | 72.8 |
| 2 | 442 | 70.7 |
| 3 | 468 | 74.9 |
| Mean | 72.8 | |
| %RSD | 2.9 |
Applied factor: 0.72 (rounded down from 0.728), against the acceptance band of >= 70%. %RSD of 2.9% is well within the 15% acceptance, confirming the technique is reproducible across replicates.
Recovery-correction convention (excerpt): This study’s factor is applied by dividing the swab limit by the recovery factor to give a recovery-corrected acceptance limit, per section 7. The parent protocol CV-2026-014 uses this same convention throughout.
Summary and conclusion (excerpt): The recovery factor for Product A on 316L stainless (2B finish) is established at 0.72, meeting acceptance criteria for mean recovery and %RSD. This factor is approved for use in protocol CV-2026-014 for all swab locations on this material of construction. A separate study is required before this factor may be applied to any other surface material.
Common inspection findings this protocol prevents
- A recovery factor applied that was established on a different surface finish than the actual equipment (bare stainless versus electropolished).
- Spike-and-swab performed wet, inflating recovery relative to dried real-world residue.
- One recovery factor applied across a multi-material equipment train with no material-specific study behind it.
- The recovery-correction convention undocumented, so the same result is calculated two different ways by two different reviewers.
How to adapt this protocol
- Run one copy of this protocol per distinct residue-and-surface-material combination in your cleaning validation program.
- Set the spike levels in section 5.2 to bracket your actual acceptance limit, not a generic default.
- Confirm the technique in sections 5.4 and 5.5 is copied verbatim from your finalized field sampling plan, not redrafted from memory.