Regulations and compliance · Practical guide

Reprocessing GMP data: which version supports the reported result?

Reintegrating a peak or correcting a calculation parameter can give a different result from the same raw data. The original measurement may still be valid. Link the raw data, processing parameters, results and decisions so a reviewer can establish why each version exists and which result was used.

Table of contents

At a glance

What needs to remain traceable when data are reprocessed?

When reprocessing GMP data, keep the raw data, the relevant processing parameters and each resulting version linked and traceable. Justify the change and assess whether it is methodologically appropriate. A later result that meets expectations does not automatically displace earlier relevant results. Decide which version is valid for further use on scientific grounds, independently of which file was saved last.

  • A new processing step is not equivalent to a new measurement.
  • Parameter status and result version are part of the testable traceability chain.
  • The professional judgment determines the applicable version, not the order of storage.

1. Separate measurement, processing, and decision-making

The example scope is human medicinal product GMP. The measurement generates source data; processing may derive integrated signals, concentrations, or statistical values from it. A further processing step does not necessarily alter the original measurement procedure, but it does affect the basis upon which a result decision is made.

EU GMP Chapter 6 requires recorded and critically reviewed results and calculations. Annex 11 supplements the framework for computerised processes. For the US Drug cGMP area, the FDA specifically states: For most laboratory analyses, reprocessing should not be routinely required. When chromatographic data are reprocessed, written procedures must exist and be followed; each result remains available for review (Question 14). The audit trail shall enable traceability of reprocessing, and the rationale for such actions must be documented (Question 1(c)).

Sources: [1] [2] [3]

2. Link data, processing, and decision-making together

A version number on a PDF is insufficient if the underlying parameters remain unknown. Conversely, a stored method alone does not prove that the exact version used for the results presented was actually executed. Establish the specific linkage.

The following internal matrix separates the required questions without requiring separate applications or individual files. The information can be integrated within a system or consolidated through controlled references.

Own assignment matrix for a reprocessing action
LevelTraceable contentTypical outstanding question
Source data Dataset identifier, sample, data capture context Were the same raw data actually used?
Processing Method, Parameters and Current Status Which setting led to which number?
Execution Time, responsible person or process Was is only one method stored or also executed?
Result Status, units, and calculation Is the old site still considered acceptable?
Decision Base, Testing and Further Use Why is this status professionally significant?

Sources: [4] [3] [5]

3. Describe the reason preceding the desired outcome

A reprocessing may be justified when, for example, a documented parameter misassignment is corrected or a methodologically planned integration is applied. The rationale should arise from a technical or scientific question and not be retroactively justified by a more favorable outcome.

An existing out-of-specification (OOS) result remains to be investigated. A subsequent result within specification does not justify the invalidity of the prior finding. Pre-established written guidelines and documented investigations distinguish a legitimate reprocessing from a selective search for a suitable result.

Sources: [6] [3] [1]

4. Worked example using the same raw data and a different blank value

For the considered procedure, the defined calculation is: Result = (Signal − Blank value) × Factor. The signal value is 0.842, the factor is 100. In the first processing stage, a blank value of 0.024 was entered. However, the relevant, separately obtained original record shows a value of 0.042. This assumed record discrepancy serves as the basis in the example for investigating a potential input error.

With 0.024 the result is 81.8. With 0.042 it is 80.0. The difference is 1.8 units. The lower result is neither better nor worse simply because of its magnitude. The original record and the method determine the assessment. The example specifies no actual laboratory method or specification limit.

Own target matrix of the two processing stages
Assessment pointVersion AVersion BExpected evidence
Original signal 0.842 0.842 Identical original data reference
Blank value 0.024 0.042 Evidence shows 0.042; deviation explained
Factor 100 100 The specified factor remains unchanged
Calculation (0.842 − 0.024) × 100 (0.842 − 0.042) × 100 81.8 or 80.0
Use For professional assessment, the following is provided. Intended for review and correction as a revised version Decision instead of overwriting
negative occurrence Only change the report count to 80.0 No documented new processing step Not to accept as a traceable correction

Sources: [1] [3]

5. Access to a clear change history is available

The history should show which relevant setting has been changed, by whom, when and for what reason. The scope of the audit trail evaluation depends on the GMP-relevant processes and risks. It is not about the general logging of each keystroke.

Retain previous processing results with their context. In the US Drug CGMP environment, the FDA expects each result to be available for review. A screenshot of the previous view cannot serve as adequate evidence. If a report displays only the released status, access to the remaining data and decisions must remain available; not every individual report needs to contain the full history.

Sources: [2] [4]

6. Test the actual change procedure

Do not verify only whether the application correctly performs subtraction and multiplication. Also assess whether a change is detectable against the previous state, whether the correct source file is used, and whether the intended access permissions are maintained. A result may mathematically be correct yet incorrectly associated with the wrong dataset.

In the example, a suitable negative case is the manual overwriting of the report number alone. The system or the proven control process must be able to detect or prevent such unproven correction. What technical and organizational combination is suitable depends on the system, process and risk.

Sources: [2] [4]

7. Focus the assessment on identifying relevant changes

The reviewer must have the rationale, the original baseline, the modified parameters, and the new result. Additionally, it must be assessed whether other results used the same erroneous parameter. This impact question must not disappear merely because a single report has been successfully corrected.

In small laboratories, a clear reference sequence provides guidance: dataset, processing A, finding, processing B, and decision. A complete automatic recalculation of all historical data cannot be inferred from this alone. The scope of additional tests follows from actual dependencies and investigative assessment.

Sources: [1] [7] [4]

8. Checklist prior to use of new result

A completed reprocessing procedure must enable a qualified person to understand the change and identify the relevant result version. A technically green job status alone does not address these questions.

Use the following custom list tailored to your method and available system functions. The specific format of the evidence generation is not a prescribed universal documentation package.

  • Rationale and technical question regarding the repeated processing documented.
  • Original data unequivocally assigned.
  • Current and updated relevant parameter values available.
  • Execution and results associated with the respective statuses.
  • Change origin, user, and timestamp are traceable.
  • Potential impact of additional results is assessed.
  • The technically significant version and further use have been decided.

Sources: [2] [1] [3]

Project experience and collaboration

Assess reprocessing against the purpose of the original analysis

With the laboratory manager, QA and system team, we trace a record through initial processing, a justified change and repeat review. We consider method version, parameters, previous results and the documented investigation together. A known processing error may need different handling from an unexplained OOS result. We define which information the reviewer needs and how it will remain accessible. A satisfactory final result alone does not justify removing the earlier result from the assessment.

Frequently asked questions about Reprocessing GMP data

Is reprocessing fundamentally unacceptable?

No. A methodologically appropriate, justified, and traceable handling process may be required. Concerning selective reprocessing until the desired outcome is achieved without a sound rationale. Review the justification, rules, original data, and the decision on the applicable stage jointly.

Sources: [3] [6]

Is reprocessing the same as a repeat measurement?

No. Reprocessing processes already existing initial data again. A repeat measurement generates new measurement data. Both processes can answer different scientific questions. Determine which path was actually chosen and why it is suitable for the concrete findings.

Sources: [3] [1]

Is the new method version number sufficient?

Not solely. It must also be verifiable that the correct dataset was processed with this status and that the resulting outcome is attributable to it. A stored method status lacking execution and result linkage does not close this traceability gap.

Sources: [4] [3]

Do we need to print each old result entry in the final report?

No. A focused report can illustrate the relevant status. However, the required original results and the rationale for the selection must remain accessible and understandable. The report must not imply a non-existent uniqueness of the data history.

Sources: [5] [4]

Can a manual integration be used?

This depends on the method, scientific validity, and the defined controls. Manual processing is not inherently invalid solely due to its manual nature. However, it must be justified, traceable, and professionally reviewed; merely producing appropriate results is insufficient.

Sources: [3] [1]

What should be considered in the case of an original out-of-specification (OOS) result?

The OOS investigation is not automatically invalidated by a new, appropriate processing step. Verify whether the proposed root cause is supported by evidence and assess how all relevant data contribute to the decision. A better outcome does not require an independent root cause analysis.

Sources: [6]

Which audit trail entries are important for inspection?

Review relevant parameter changes, processing steps and result decisions to the extent that they affect the particular GMP evidence. Set a risk-based scope. An unfiltered list of technical events is not an adequate review by itself.

Sources: [2] [4]

Must all previous results be recalculated upon a correction?

Not at all. Determine which results may be affected by the wrong parameter or the same cause. Appropriate existing evidence can bear a demarcation. However, unexplained dependencies should not simply be treated as non-affected.

Sources: [7] [4]

Can an export as an Excel file replace the history?

Only when it obtains reliable content and context necessary for the intended proof. A results list without raw data references, parameters, and change justifications does not substitute for a processing history. Verify the actual export scope, rather than inferring suitability from file format alone.

Sources: [4] [3]

Who determines which site is used?

Your quality system governs processes and responsibilities accordingly. Technical editing rights do not automatically imply business decision-making authority. The new system should only be used for its intended purpose after the planned assessments have been completed; the final data entry is not a substitute for that.

Sources: [2] [1]

Sources and review date

Source date: 7 October 2026. The technical examples relate to human medicinal product GMP. FDA statements are contextualized within the applicable U.S. medicinal product framework. Tables and numerical examples are intended as internal work aids.

  1. European Commission: EU GMP Chapter 6, Quality ControlRevision 2014; effective from 01.10.2014; §6.7 Documentation; §6.9 Out-of-Specification/Out-of-Trend; §§6.15–6.17 Test methods, calculations, and laboratory records
  2. European Commission: EU GMP Annex 11, Computerised SystemsRevision January 2011; from 30.06.2011; Principle; §§1,4,5,7,9–13,17; §12.4 Identity, Date and Time; §10 controlled changes
  3. FDA: Data Integrity and Compliance With Drug CGMP, Questions and AnswersFinal Guidance, December 2018; Section III Q1(b)–(d), Q2, Q4–10 and Q12–18; Reprocessing Q14, Scope of Investigation Q18
  4. PIC/S: PI 041-1, Data Management and IntegrityPIC/S PI 041-1, 1 July 2021. See section 3.7 for status, sections 7.5–7.8 and 8.1–8.8, sections 9.1.5 and 9.3 and 9.5–9.10, section 12.1 and the glossary.
  5. European Commission: EU GMP Chapter 4, DocumentationJanuary 2011; in effect since 30.06.2011; Principle; §§4.1–4.9, in particular timely recording 4.8 and visible correction 4.9; §§4.10–4.12 Storage
  6. FDA: Investigating Out-of-Specification Test Results for Pharmaceutical ProductionFinal guidance, Level 2 revision, May 2022. Section I defines OOS. Sections III–IV address investigations, IV.B retesting, IV.C averaging and outlier tests, and V the conclusion.
  7. European Commission: EU GMP Chapter 1, Pharmaceutical Quality SystemRevision 2013; effective from 31.01.2013; §1.4(xii)–(xiv): proactive change evaluation, post-evaluation, and root cause analysis/CAPA; §1.5 Management

Sources checked: .

Daniel Herrmann, Daniel Herrmann Consulting

Daniel Herrmann Consulting

Integrate processing and result decision-making

We support you with verifiable evaluation pathways, appropriate system controls, and clear rules for processing versions and their review.

  • Computer system validation →

    Bring your GxP systems into validated operation, from laboratory software, MES and eQMS to ERP. We define the scope, plan risk-based tests using GAMP 5 and hand the evidence to QA for release.

Send a project enquiry Customer testimonials →