SOP for Laboratory Incident in a QC Laboratory (Example)

In short: This example SOP shows how a QC laboratory reports, investigates, approves and closes a laboratory incident such as an instrument malfunction, an analyst error, a system or power failure or a spill. It stresses that no data is discarded, that out-of-specification (OOS) results go to the OOS procedure, and that incidents are trended. Time limits and numbering are example in-house criteria.

1. Objective

To describe how laboratory incidents are reported, initiated, reviewed, approved, complied with and closed in the Quality Control (QC) laboratory.

2. Scope

This example SOP applies to laboratory incidents in the QC laboratory of a manufacturing facility. It is a template and should be adapted to your own quality system. It works alongside the site’s procedure for handling of incidents and deviations and the OOS procedure (OOS investigation: complete guide).

3. Responsibility

• QC analyst: reports any incident immediately to the section head or Head QC and records what happened.
• Head QC or designee: assesses the incident, leads the investigation, decides root cause and corrective and preventive action (CAPA).
• Quality Assurance (QA): issues the incident form, logs and numbers the incident, reviews it and closes it after satisfactory completion of actions.

4. Procedure

4.1 Definition

A laboratory incident is an unplanned or uncontrolled event in the laboratory, a departure from the designated system, instrument or procedure at any stage of analysis, caused by system failure, instrument malfunction or manual error.

4.2 Reporting and numbering

1. The analyst reports the incident to the Head QC or designee as soon as it is noticed.
2. QA issues the incident form and its authorised designee logs it in the laboratory incident log.
3. Each incident gets a reference number. Example format: QC/LIR/YY/XXX, where QC is Quality Control, LIR is Laboratory Incident Report, YY is the year and XXX is a three-digit serial number that restarts each year. QC/LIR/26/001 would be the first incident of 2026.
4. The analyst records a description of the incident. The section head and Head QC assess it, and the immediate supervisor investigates according to its nature.
5. Staff may first write a proposal in their local language if that is easier, but the department head or QA must put the final record into English so that it can be reviewed by customers, auditors and inspectors.

4.3 Investigation

1. Review all operations, records and procedures that may explain the incident.
2. Evaluate the probable causes and conclude on the root cause, with evidence. “Analyst error” alone is not a root cause; state what happened and why.
3. Document sample details, method, description, immediate action, investigation, conclusion, recommendation, correction and CAPA (Annexure 1).
4. If the incident is linked to a batch, complete immediate corrective actions before the batch is released.
5. Head QC or designee decides the actions that prevent recurrence.
6. Example in-house criterion: investigate within 15 working days unless a justification is approved.

4.4 When an OOS result appears

If an OOS result is obtained during the incident investigation, stop and handle it under the OOS procedure. The US FDA guidance on OOS results expects a documented laboratory (Phase I) investigation, and a result may be invalidated only on evidence of a laboratory cause that can reasonably be shown to have caused the result. See Phase I laboratory investigation. EU GMP Chapter 6 likewise expects OOS and significant atypical results to be investigated.

4.5 Handling of data from an incident

Data from the affected analysis must never be deleted or hidden. US cGMP requires laboratory records to contain complete data from all tests, including instrument graphs, charts and spectra (21 CFR 211.194(a)).

1. Mark invalidated data clearly (for example with a “Disregarded” stamp), with the analyst’s signature, date and reason. The data stays on file.
2. Attach the original data to the analytical record and a copy to the incident report.
3. Typical data affected: balance printouts, analysis sequence, chromatographic parameters and chromatograms, UV spectra, calculation sheets, report printouts, and data from an instrument failure during analysis.
4. Write the incident number on the worksheet so that records can be cross-referenced.
5. Electronic data follows the same rule: the audit trail must show the affected injections, and nothing is overwritten.

4.6 Types of laboratory incident (examples, not a closed list)

Instrument malfunction

• Check the instrument to find the cause. Use an alternative qualified instrument if the analysis must continue.
• Return the instrument to use only after it has been checked and found satisfactory. Otherwise label it “out of order” until it is repaired and re-verified (see SOP for maintenance of laboratory instruments).

Analyst oversight or error

• Examples: weighing or dilution error, programming error, mix-up of sample and standard preparations, or analysis not carried out as per procedure.
• If an error is noticed before a result is generated, consult the section head and document it in the analytical record.
• Missing entries that could affect results, such as instrument usage, working or reference standard, column usage or temperature, are handled as incidents and the analyst is retrained.

System failure or power failure

• Examples: system suitability failure, unexpected peaks in chromatograms (other than those from blank, mobile phase, placebo or diluent), vial skipping, a server problem, or a stop due to high back pressure or power failure.
• A system check injection made only to prepare the system, as defined in the method, does not need an investigation but its data is still retained and marked.
• If system suitability fails during a sample run, log and investigate the incident. Accept the data only if justified; otherwise establish suitability again and repeat the analysis using the original solutions if they are within their solution stability period.
• Do not repeat a failed run repeatedly until it passes. The decision to retest is documented first.

Spill of sample, material or solution

• Tell the immediate supervisor. Disregard the weighing printout and prepare the sample again only with permission of Head QC or designee, under this procedure.

4.7 Closure

1. Example in-house criterion: close incidents within 30 working days; if that is not met, the initiator submits a justification approved by Head QA.
2. QA checks the status and signs in the incident log.
3. Product-specific incidents are closed before batch release, and a copy is filed with the raw data.

4.8 Trending

1. QA trends laboratory incidents at least once a year, and preferably more often (for example quarterly) so that repeat problems are seen early (Annexure 3).
2. Product-specific incidents are listed in the Annual Product Quality Review (APQR).
3. Recurring causes lead to CAPA; see CAPA process.

4.9 Records

Incident reports linked to a batch are batch-associated records. Keep them at least 1 year after the expiry date of the respective product batch (EU GMP Part I Chapter 4, section 4.11, or 5 years after QP certification if longer; 21 CFR 211.180(a)), and longer if local law or the marketing authorisation requires.

4.10 Training and distribution

• Trainer: Head QC. Trainees: section heads of QC and QA.
• Distribution: controlled copies to Head QA and Head QC; original with Head QA.

5. Annexures

• Annexure 1: Incident investigation form
• Annexure 2: Incident log
• Annexure 3: Trend analysis of incidents

6. Revision history

Frequently asked questions

What is the difference between a laboratory incident and an OOS?

An incident is any unplanned event in the laboratory. An OOS is a result outside its specification. An incident can lead to an OOS, and once a result is OOS it is handled under the OOS procedure.

Can the analyst simply repeat the test after an incident?

Only as allowed by the procedure and after the incident is documented. Repeating until a passing result is obtained is not acceptable. The original data is always kept.

Is a laboratory incident the same as a deviation?

It is a departure from a procedure in the laboratory, so many sites handle it through a separate laboratory incident route that follows the same principles as deviations.

References

• US FDA: 21 CFR 211.194(a), Laboratory records (eCFR, checked 2026).
• US FDA: 21 CFR 211.180(a), General requirements for records and reports (retention).
• US FDA: Guidance for Industry, Investigating Out-of-Specification (OOS) Test Results for Pharmaceutical Production, Revision 1 (May 2022).
• European Commission: EudraLex Volume 4, EU GMP Part I Chapter 6: Quality Control (in operation since 1 October 2014), sections 6.7, 6.9 and 6.35.
• European Commission: EudraLex Volume 4, EU GMP Part I Chapter 4: Documentation, section 4.11.