OOS Investigation (US FDA): Phase I Laboratory Investigation

In short: An out of specification (OOS) result is any test result outside the limits set in the application, the DMF, a pharmacopoeia or the manufacturer’s own specification. US FDA expects a Phase I laboratory investigation first, to find out whether an obvious laboratory error caused it. This post is for QC analysts, supervisors and QA reviewers, and follows FDA’s May 2022 guidance.

See also: Phase II: full-scale OOS investigation (US FDA) and the UK view in OOS investigation Phase II (MHRA).

What counts as an OOS result

An OOS result is any test result that falls outside the specification or acceptance criteria set in a drug application, a drug master file (DMF), an official compendium, or by the manufacturer. In-process laboratory tests that fall outside their specifications are also covered.

The guidance applies to chemistry-based laboratory testing of drugs regulated by CDER. This includes APIs, excipients and other components, in-process materials and finished products, to the extent the CGMP regulations (21 CFR Parts 210 and 211) apply. It also applies to in-house testing of purchased components, and contract laboratories can use it too. It is not written for biological assays such as in vivo tests or immunoassays, and it does not cover in-process tests done only to adjust equipment in real time.

What changed from the 2006 guidance

The current text is Revision 1 (May 2022). It keeps the 2006 structure, but readers of older material should note these points:

• “Quality control unit (QCU)” is now “quality unit (QU)”.
• A full-scale (Phase II) investigation is triggered when no causative laboratory error is found. The 2006 text said “no meaningful errors”.
• OOS results must not be blamed on analytical error unless the investigation clearly establishes a laboratory root cause.
• The guidance now refers to both corrective and preventive action (CAPA) and links this to ICH Q10.
• A footnote says much of the guidance is also useful for out-of-trend (OOT) results.

Why the investigation is required

Under 21 CFR 211.192, any OOS result must be investigated and the work recorded in writing, including conclusions and follow-up. The aim is to find the cause. The source is either a problem in the measurement (laboratory) or a problem in the manufacturing process.

Rejecting the batch does not remove the duty to investigate. The result may also affect other batches or other products. The investigation should be thorough, timely, unbiased, well documented and scientifically sound.

Phase I is the first part. It is an initial check of how accurate the laboratory data are. Where possible it should be done before test preparations are discarded, including the composite or homogeneous source of the aliquot, so that ideas about laboratory error or instrument faults can be tested on the same preparations.

If this check finds no causative error in the analytical work, a full-scale investigation follows (see Phase II). When a contract laboratory gets the result, it passes its data, findings and supporting records to the manufacturer’s QU. The QU then starts the Phase II investigation if no clearly causative laboratory error was found.

Responsibility of the analyst

The first responsibility for a correct result lies with the analyst doing the test. Key expectations:

• Use only instruments that meet their performance specifications and are properly calibrated (21 CFR 211.160(b)(4)).
• Do not use a system that fails its system suitability requirements. In chromatography, reference standard solutions may be injected during the run to monitor drift, noise and repeatability. If these show the system is not working properly, identify the data from the suspect period and do not use it. Find the cause, and correct it if possible, before deciding about earlier data.
• Check results against the specification before discarding test or standard preparations.
• If a result is unexpected and has no obvious explanation, keep the preparations (if stable) and tell the supervisor straight away so the assessment can begin at once.
• If an obvious error occurs, such as a spilled sample solution or an incomplete transfer, document it immediately.
• Do not carry on with an analysis you already expect to invalidate for a known cause, just to see what result comes out.

Responsibilities of the laboratory supervisor

The supervisor’s assessment must be objective and prompt, with no assumption about the cause. The data should be reviewed quickly to judge whether the result points to laboratory error or to a manufacturing problem. Re-examining the actual solutions, test units and glassware used may add weight to a laboratory-error hypothesis.

The supervisor’s assessment should cover these steps:

1. Discuss the test method with the analyst and confirm the analyst knows and followed the correct procedure.
2. Examine the raw data, including chromatograms and spectra, for anything unusual or suspect.
3. Check that the calculations from raw data to final result are scientifically sound and correct. Also check whether anyone made unauthorised or unvalidated changes to automated calculations.
4. Confirm that the instruments performed correctly.
5. Confirm that the right reference standards, solvents, reagents and other solutions were used and met their quality specifications.
6. Judge whether the method is performing as expected from its validation and historical data.
7. Document the whole assessment fully and keep the records.

Examining retained preparations

Finding a cause is much easier if retained preparations are examined promptly, and each hypothesis (for example a dilution error or an instrument fault) should be tested. FDA gives three examples:

• Re-injection of solutions when a transient equipment fault is suspected. This is hard to prove, but it can give strong evidence that the instrument, not the sample or its preparation, was the problem.
• Examining the original dosage unit in release-rate testing of specialised dosage forms that are not destroyed by the test. Damage during laboratory handling would support invalidating the result and a retest.
• Further extraction of a dosage unit, where possible, to see whether the first extraction was complete. Incomplete extraction could invalidate the result and raises questions about method validation.

Concluding Phase I

Every step must be documented. Laboratory management should look at more than the single value. It should consider what the OOS results mean for the laboratory’s quality assurance programme and watch for trends. Laboratory error should be rare. Frequent errors suggest poor training, badly maintained or uncalibrated equipment, or careless work. When laboratory error is found, identify its source, take corrective action to prevent recurrence and document that action.

The outcome rules are simple:

• If there is clear evidence of laboratory error, the laboratory result is invalidated.
• If the evidence is unclear, the manufacturer must carry out a full-scale OOS investigation to find the cause.
• A result must not be put down to analytical error without an investigation that clearly establishes a laboratory root cause. Both the initial assessment and the later investigation must be fully documented.

Frequently asked questions

Can I retest straight away if a result is OOS?

No. First assess the original data and test preparations under Phase I. Any retest must follow a predefined SOP and plan, as explained in Phase II.

Is “analyst error” an acceptable root cause?

Only if the investigation clearly shows what the error was, with documented evidence. A general statement of analyst error, without evidence, is not acceptable.

Why keep the test preparations?

They allow the laboratory to test hypotheses such as dilution error or instrument malfunction on the same material, which makes the investigation more credible.

Does this guidance cover microbiology tests?

FDA states that it applies to chemistry-based laboratory testing. It is not meant to address biological assays. For microbiological excursions, follow your own SOPs and the relevant pharmacopoeial and regulatory expectations.

Is the guidance legally binding?

It is a nonbinding recommendation. The binding requirements are in the regulations, for example 21 CFR 211.192. Indian firms exporting to the US are inspected against these expectations in practice.

References

• US FDA, CDER. Guidance for Industry: Investigating Out-of-Specification (OOS) Test Results for Pharmaceutical Production, Revision 1, May 2022 (supersedes the October 2006 version). Available on fda.gov.
• US FDA. 21 CFR Part 211, Subparts I and J (including 211.160, 211.165, 211.180, 211.192, 211.194).
• ICH Q10, Pharmaceutical Quality System (April 2009), as cited in the FDA guidance.

This post is a study summary in original wording and not a substitute for the guidance itself. Read the full FDA text before writing or auditing an SOP.