Common OOS Investigation Mistakes and How to Prevent Them

Out-of-specification investigations are one of the most inspected processes in a pharmaceutical lab, and consistently one of the most cited. FDA’s warning letter data from 2025 and 2026 shows that laboratory control failures like testing into compliance, superficial root cause analysis, weak CAPA, and irregular audit trails continue to plague CGMP environments.

21 CFR 211 and the FDA’s 2022 revised guidance, Investigating Out-of-Specification (OOS) Test Results for Pharmaceutical Production governs OOS investigations with the two-phase model. During inspections, the FDA will likely review the scrutiny applied to the investigation records along with the evidence behind every OOS record decision.

This blog walks through the OOS mistakes that keep showing up in warning letters, the regulatory expectations that catch them, and how an eQMS like ACE can be configured so those mistakes are structurally difficult to make.

The Regulatory Backbone of an OOS Investigation

An OOS result is any test outcome that falls outside the acceptance criteria in an application, DMF, compendia, or the manufacturer’s own specifications — including in-process laboratory tests. Once identified, the following requirements apply:

• 21 CFR 211.192 requires that any unexplained discrepancy or batch/component failure be thoroughly investigated, whether the batch has been distributed, and requires the investigation to extend to other batches of the same product and other products that may be affected. A written record of the investigation shall be made and shall include the conclusions and follow-up.
• 21 CFR 211.194 requires complete laboratory records, including raw data, calculations, standards and reagents information, equipment calibration status, and analyst identity, to be preserved.
• 21 CFR 211.22 requires the Quality Unit to review and approve OOS investigations and conclusions before batch disposition.
• 21 CFR Part 11 governs the electronic records and signatures used throughout the investigation.
• FDA’s 2022 OOS guidance structures the investigation into Phase I (laboratory) and Phase II (full-scale) and is explicit that assignable laboratory error must be identified with objective evidence — vague “analyst error” claims are not enough to invalidate a result.

What Not to Do During an OOS Investigation

1. Do not discard test preparations, standards, or retained samples before the investigation is opened

Analysts should retain their prepared solutions until they have had the opportunity to check their data for compliance with test specifications. Once these solutions are gone, Phase I investigations become much harder to complete. Tidying up an analyst’s bench before flagging the result can end up invalidating investigations.

2. Do not retest until there is a passing result

Testing into compliance is the action of repeating a test with the goal of invalidating or average the OOS result away, and this is one of the most severe violations FDA cites. This is also being treated as a data integrity issue by regulators, not just a lab error.

3. Do not attribute the result to “analyst error” without documented evidence

Assignable laboratory errors must be confirmed with objective document evidence. A supervisor’s judgment call, on its own, cannot be used to establish analyst error as the cause of an OOS result.

4. Do not close the investigation with “re-training” as the corrective action

Many recent warning letters from the FDA specifically flag this. Reflexively assigning re-training as the fix for every deviation is no longer accepted as substantive root cause analysis. Concrete actions and/or procedural changes to prevent errors from reoccurring should be deployed following an investigation.

5. Do not treat the audit trail as optional

Another common feature in recent FDA warning letters is repeatedly citing facilities where audit trails were disabled or electronic records were altered. ALCOA+ is being strictly enforced by global regulators, and companies should expect their eQMS platforms and internal processes to support this requirement.

How ACE Can be Configured to Meet These Requirements

ACE handles OOS as a linked set of quality event records and digital workflows rather than a single form, with an audit trail that tracks every action in the records. The OOS workflows can also be configured to follow the two-phase model directly.

The OOS is initiated as a quality event record in ACE, capturing the sample, batch, method, instrument, analyst, calibration status, and specification as structured, required fields. The workflow can be set up so that the OOS record cannot advance without information in these fields, and with linked records, these fields can be populated with information directly from other ACE records.

Phase I investigations can be configured to mirror the operating site’s OOS procedures. Data entry fields such as system suitability, calculations, method version, training records, equipment calibration, standard and reagent expiry, and more can be added to the record as needed. Since Phase I investigations are done by laboratory personnel, record editing can be limited only to laboratory roles.

As Phase II investigations extend into the production suite, additional ACE modules like ACE EBR can be linked to the record. This is where batch record data, exception logs, and interval timers are available directly within the investigation. With EBR execution data tied into ACE Analytics and workspace search, Phase II reviewers won’t have to leave the platform to find production evidence.

Correcting Errors with ACE

Here is how each of the five mistakes above is addressed structurally by an ACE configuration.

Closing Thoughts

When reviewing OOS records, FDA inspectors expect to see documented, linked, and reviewable information in an easily accessible and reviewable format. The errors that result in observations are almost always structural, such skipping a step in the testing protocol, a record that is missing in the physical archive, or a link that is missing between the lab investigation and the batch record.

An eQMS configured around the two-phase model doesn’t completely prevent an OOS from happening, but it can help with thorough investigation of the OOS and prevent it from being found as a signification observation during an inspection. ACE does this in two-folds, with a comprehensive eQMS that features extensive digital workflows and records to document OOS occurrences, and a dedicated inspection module in ACE Inspection to bring all the information to the forefront during an inspection. Connect with us to see how ACE is used by the top life sciences companies to help them in operation and for inspections.