# OOS Investigation Checklist for GMP QC Laboratories > Source: https://netray.co/tools/oos-investigation-checklist > Publisher: Netray (netray.co), AI and integration engineering for ERP and manufacturing systems. > Section: Free Tools ## Short answer A step-by-step out of specification checklist covering immediate lab assessment, full-scale investigation, retesting, resampling and batch disposition. ## Category quality ## Industry Pharma and Life Sciences ## Audience - QC analyst - Quality assurance - Laboratory supervisor ## Tool **Kind:** checklist **Groups:** #### Immediate response (analyst and supervisor) **Items:** **Text:** Stop and report the suspect result to the supervisor without discarding anything **Critical:** true **Text:** Retain all solutions, standards, sample preparations, vials and glassware as found **Critical:** true **Text:** Record the result and time in the raw data; do not delete, overwrite or re-inject to see a better value **Critical:** true Open an OOS number in the register and notify QA the same day Confirm the product, batch, test, specification and method version are correctly identified Secure the raw data, chromatograms, balance printouts and instrument logs #### Phase I: laboratory assessment **Items:** **Text:** Check calculations, dilutions, sample weights and units against raw data **Critical:** true **Text:** Verify the instrument was calibrated, qualified and had passed system suitability **Critical:** true Check standard and reagent identity, grade, potency, preparation date and expiry Check the analyst training record for this method **Text:** Review the audit trail for the run for deleted, reprocessed or manually integrated data **Critical:** true Check sample storage, handling, container and time between sampling and testing Interview the analyst without assigning blame and record observations Check the method was followed exactly, including column, mobile phase, temperature and timing **Text:** Document whether an assignable laboratory error was found and its evidence **Critical:** true #### Retesting and resampling rules **Items:** **Text:** Retest only after an assignable cause is documented or per the approved retest protocol **Critical:** true Define in advance the number of retests and who performs them, not the same analyst alone **Text:** Do not average a failing result with passing results to hide the failure **Critical:** true **Text:** Do not test into compliance by repeating until a passing result appears **Critical:** true Use the original homogeneous sample or retained sample, and document the amount used Resample only when sampling error is documented and justified Treat all valid results, including the original, as reportable unless invalidated with evidence #### Phase II: full-scale investigation (production and QC) **Items:** **Text:** Extend the review to the manufacturing batch record, process parameters and deviations **Critical:** true Review raw materials, equipment, cleaning, environment and operator actions for the batch **Text:** Check other batches and products made on the same equipment or with the same lot of material **Critical:** true Review historical OOS and trend data for the same test and product Run additional testing under a written protocol approved by QA before it starts Use a structured tool such as a fishbone or 5 Whys and record the root cause or probable cause **Text:** Assess impact on stability studies, released batches and marketed product **Critical:** true Consider a field alert or regulatory notification if distributed product is affected #### Conclusion, disposition and closure **Items:** **Text:** State a clear conclusion: invalidated with assignable cause, or confirmed OOS **Critical:** true **Text:** Disposition of the batch (release, rework, reject) documented and approved by QA **Critical:** true Corrective and preventive actions defined with owners and dates Link the OOS to any deviation, CAPA or change control numbers Update the trending record and the OOS log Complete the investigation within the time set by your SOP (commonly 30 working days) QA head signs the final report **Scoring Note:** Every critical item must be ticked before the OOS report is closed. Phase I does not end with a retest unless a lab error is proven. Based on the structure of FDA guidance on investigating out of specification results (2006) and EU GMP Chapter 6, but your own SOP governs. ## Content **Intro:** An out of specification result is any test result that falls outside the acceptance criteria in a specification or approved procedure. Regulators treat the handling of OOS results as a test of a QC lab's integrity, so the sequence matters. This checklist keeps the investigation in order. **Sections:** #### How to use the checklist **Body:** Work through the groups in order. Start the first group the moment a suspect result is seen, before any retest. Tick items as you complete them, and keep a note of where the evidence is, such as the notebook page or the audit trail ID. Export the checklist to PDF to attach to the investigation report. The critical items are the ones inspectors most often cite as failures, so none should remain open at closure. **Bullets:** - Do not retest before Phase I is done - Never discard the original preparations - Involve QA from the first day - Keep a single OOS number from start to end #### Phase I and Phase II **Body:** Phase I is the laboratory assessment, led by the analyst and supervisor, looking for obvious errors in calculation, instrument, standard or method execution. If a clear assignable cause is found and documented, the original result may be invalidated and a retest performed. If not, the investigation moves to Phase II, which is a full-scale investigation involving production, QA and possibly other batches. The FDA 2006 guidance sets out this structure. #### Retesting without testing into compliance **Body:** The retest plan must be written before testing, with a fixed number of retests and rules for evaluating them. Repeating a test until it passes and ignoring the failing results is considered testing into compliance and is a data integrity violation. Averaging is acceptable only when averaging is the method's defined reporting rule. All results, valid and invalid, stay in the records with the reason for any invalidation. #### Common mistakes and systems notes **Body:** Findings often cite informal retesting, late QA notification, missing audit trail review, investigation that ignores other batches and root cause stated only as analyst error. In LIMS and ERP quality modules, place the batch on quality hold when the OOS is registered. OOT (out of trend) results, which are inside the specification but atypical, need their own procedure but benefit from the same discipline. **Cta:** Netray can help build a private assistant that checks LIMS and ERP data for related batches when an OOS is opened. The checklist works fully on its own. **Faq:** #### What is the difference between OOS and OOT? **Answer:** An OOS result is outside the registered or approved specification. An OOT result is within specification but outside the expected trend, for example a stability result dropping faster than earlier lots. OOT is an early warning and should trigger a review. #### Can I retest an OOS sample immediately? **Answer:** Not before documenting why. A retest is justified when a laboratory error is confirmed or when the approved SOP defines a retest plan. Immediate unplanned repeating is a classic data integrity finding. #### How long should an OOS investigation take? **Answer:** Many SOPs set 30 working days for closure, with the FDA expecting investigations to be timely and documented. Extensions need written justification, and the batch stays on hold while the investigation is open. #### Can an OOS be invalidated? **Answer:** Yes, only when a documented, verified laboratory error is found as the assignable cause. Suspicion alone, or a later passing result, does not invalidate a failing result. ## Search terms OOS investigation checklist, out of specification investigation procedure, OOS investigation SOP, FDA OOS guidance checklist, phase 1 and phase 2 OOS investigation, OOS retest rules, laboratory investigation checklist, OOT and OOS difference --- Talk to an engineer: https://netray.co/contact?intent=erp_ai | Machine catalog: https://netray.co/agent/catalog.json