Table of contents
At a glance
What should a traceable audit trail review in Empower cover?
Link relevant Empower events to the data, method versions and results actually reviewed. Assess changes and earlier results in context, record open questions and document the scope, reviewer and decision. Define a separate review for system-wide events. A sign-off or review indicator alone does not demonstrate that all relevant data were selected or that the assessment was adequate.
- Data review, data-related audit trails, and system audit trails have different responsibilities.
- Manual handling and reintegration must be evaluated based on the specific procedure.
- Result set, result history, and employed method versions determine the testable data basis.
- Review and sign-off require appropriate rights, configuration, and a documented inspection process.
1. Clearly define the review assignment and empower version
An Empower Audit Trail Review should begin with a clear question: Is the origin of a specific analytical result or the controlled operation of the system being evaluated over a defined period? While these tasks may be interrelated, they involve different data volumes, triggers, and responsibilities. The general Empower system guide addresses validation scope; here, the focus is on the ongoing assessment of specific audit trail records.
The regulatory reference use is the quality control of human medicinal products. The GMP guideline Annex 11 specifies periodic audit trail reviews within a risk-based GMP context. In the U.S. pharmaceutical sector, additional requirements for laboratory recordkeeping must also be observed. Part 11 addresses electronic records and signatures within its scope of applicability and does not constitute a general approval for a chromatography data system.
Maintain records of your installed Empower version, project settings, audit trail configurations, and user permissions. The following functional references pertain primarily to Empower 3.8.0 and 3.9.0. These are documented reference states, not assertions regarding the latest available version or any specific customer installation. A work instruction must align with your actual current state.
2. Separate data entry and change tracking
The data recording describes the investigation and its result. The audit trail provides the context of relevant actions and changes. It does not replace chromatograms, method parameters, calculation or professional evaluation. Conversely, a plausible chromatogram alone does not show whether other results have been generated or decisive settings have been changed.
For the review, we recommend the following internal work matrix. It helps assign existing evidence to specific processes and identify gaps. A row does not imply that all information must originate from a single Empower report. What is essential is a clear and traceable linkage to the process under consideration.
| Level | Process question | Appropriate test reference |
|---|---|---|
| Analysis data and results | Has the investigation been technically correct and fully evaluated? | Chromatograms, sample assignment, calibration, calculations, and acceptance criteria |
| Project Audit Trail | What relevant actions took place in the project? | Project-related processing, result-related and method-related events |
| Method Audit Trail | What changed between the method versions used? | Version assignment, parameter comparison, and reason for change |
| Result Audit Viewer | How are results, prior results, and associated histories related? | Result Set, Result History, Results and Method Comparison |
| System Audit Trail | Which system-wide actions affect trustworthiness? | User, projects, systems, and relevant login/logout events |
| Review and sign-off | Who assessed which site with what result? | Unambiguous scope, findings, release decision, and assignment to signature |
3. From the result set to the complete inspection population
Begin with the samples and injections expected following the inspection task. Assign the selected result set to this expectation. Also assess whether discontinued, repeated, or otherwise modified procedures are relevant for evaluation. Relying exclusively on the provided final outcome prevents the ability to infer the completeness of the selection from that outcome alone.
The Result Audit Viewer aggregates information from project, method, sample, and sample set histories. Waters displays a comparison of previous, replaced results and multiple method versions within this view. In the documented example, a date filter limits the initially displayed history to three months; therefore, an empty section does not serve as evidence of a period without events.
Document the filter criteria, time period, and selection settings such that a second reviewer can reproduce the scope. Compare relevant processes with the Project Audit Trail when the inquiry extends beyond individual result-level validation. In particular, for older datasets or subsequent reprocessing, the examined time period must align with the historical context of data creation. A saved view is helpful, but its suitability should remain subject to verification.
4. Assess professional reintegration, not merely count
A manual reintegration may be scientifically justified. For its evaluation, you require the rationale, the current integration rule, the intervention, and its impact. Compare the initial result with the subsequent result and verify that the same justification was consistently applied to comparable peaks. A comment such as "corrected" typically does not provide sufficient analytical justification.
Don't confuse an event name with a finished result. Waters explains for Empower 3 that Created Manual Result can appear in Review when processing individual injections. Therefore, the entry proves neither a manually moved peak nor an inadmissible result change in itself. The evaluation must be carried out on the specific dataset.
In the FDA pharmaceutical context, the Data Integrity Guidance states that for reprocessed chromatography, not only the most recent result should be retained. The required result history must be maintained over the relevant retention period, and the rationale for the use or non-use of results must be documented. A subsequent reintegration shall not serve as a substitute for the required investigation of an original out-of-specification (OOS) result.
5. Correctly assign method stands and time specifications
A method name often gives too little information for a review. Establish which version was used for processing and which changes from the preceding version matter. Waters describes method audit trails that record version history and support comparison of method parameters. The comparison identifies a difference. Use your method procedure to assess whether the change was appropriate and approved.
Consider, in addition to the Processing Method, other relevant methods and settings pertinent to the process. A changed output presentation may obscure information, even if the underlying calculation remains unchanged. Similarly, a new calculation may generate different results without altering the raw data. Ensure such differences are clearly distinguished in the findings description.
Timestamps also require their technical context. Waters documents a situation for individually reviewed Empower-3 results where the result timestamp predates the later stored method status. This can be reasonably explained by the generation and concurrent storage of data. Therefore, assess event significance and workflow sequence before concluding a manipulation allegation from an atypical sequence.
6. Plan the System Audit Trail as a separate control area
The System Audit Trail captures activities across the system, such as login and logout events, as well as the creation, modification, or deletion of users, projects, and system objects. Waters designates this functionality under the permission "View System Audit Trails." A laboratory inspector with access to project data does not automatically possess the full scope of the task or the required privileges for this control area.
Define who evaluates which events and what information the laboratory receives in the event of a potential data reference. For example, an unexpected change to a user or a restored project may trigger a follow-up inquiry regarding the affected analysis period. The system owner investigates the technical procedure; the subject matter reviewer assesses the potential implications for the specific records.
Select a justified time period and event filter. A monthly or annual rhythm is not a universal default for all events. Data-driven reviews form part of the respective record review; system-wide anomalies may additionally trigger immediate assessment. Document routine scope, event-driven escalation procedures, and representation responsibilities. Periodic evaluation of the entire system constitutes a separate process.
7. Review example: A retrospective peak evaluation
The following example serves as a reference guide. It does not describe a completed test. In a results set containing standards and samples, a result was repeatedly stored after the first processing step due to an adjustment of the peak threshold. The reviewer must determine whether the change is justifiable and which results are relevant for the evaluation.
The review protocol references the project, sample set, result set, affected injection, as well as the previous and new result assignments. Actual identifiers are entered during operation. For the example, no measurement results or binding acceptance criteria are provided; these are derived from the respective analytical method.
| Step | Question to test | Expected demonstration or consequence |
|---|---|---|
| Population balancing | Are samples, standards, and relevant duplicates fully considered? | List of expected and actually reviewed scope; gaps will be addressed |
| Open history | Does the time period also include the first processing? | Matching filter and traceable assignment of previous results |
| Change comparison | Which peak limit and which result value differ? | Comparison with source data and documented impact on the process |
| Assess the method | Does the current integration rule cover the intervention? | Concrete method standard and clear rationale |
| Evaluate adverse effects | Are additional samples or standards affected by the change? | Targeted comparison or expansion of the review scope |
| Evaluate findings | Is the change explained or does a relevant uncertainty remain? | Justified closure or linked investigation with responsible party |
| Assign closure | Which result level was confirmed with which review scope? | Unique review traceability and corresponding sign-off |
8. Do review indicators and sign-off actually convey meaning
Empower 3.8.0 introduced a system policy for verification of reviewed audit trail records at Sign Off 2. According to the Release Notes, this feature is initially disabled and applies to standard projects with Full Audit Trail. In addition to the required rights, the signer must also have the specified permissions for Project Audit Trail and Sign Off Level 2. Therefore, a documented function name does not yet indicate active control in your installation.
The results display indicates at least one instance of confirmation. It does not itself verify whether the employee selected the correct data, understood relevant events, or addressed outstanding questions. Define where the review scope and findings are documented and how they contribute to the confirmed result status. Specifically assess the significance of the display during subsequent processing or a further sign-off.
Waters describes result sign-off via Report Publisher and its conditions. Even the channel lock possible after Sign Off 2 depends on settings and selection. Therefore, do not derive any universal immutability of all project data from a signature. For FDA-relevant electronic signatures, identity, meaning and connection to the recording state remain to be considered independently.
9. Results, responsibilities, and resource allocation are actionable and measurable
Describe an anomaly initially neutrally: affected dataset, observed action, timestamp, stated reason, and remaining question. An unusual change is neither automatically fraudulent nor out-of-specification (OOS). OOS refers to results falling outside predefined specifications or acceptance criteria. A permission error, lack of justification, and an OOS result may be interrelated but require a logically appropriate investigation.
The reviewer requires analytical expertise, access to relevant records, and knowledge of the employed Empower functions. The administrator provides support regarding technical issues and permissions. QA, in collaboration with the laboratory, defines the required control and escalation procedures. Under the scope of 21 CFR 211.194, the review of original records by a second individual is explicitly required; this does not imply a general equivalence of that individual to the system administrator.
The effort depends on data volume, reprocessing, method diversity, filter quality, and open questions. Therefore, test your intended workflow using a representative data set. Empower 3.9.0 alters the visibility of certain quantitative information for users with restricted rights, depending on whether a result has been saved. Specifically evaluate version differences during role changes, training sessions, and software updates, rather than assuming unchanged usability.
10. Checklist for a traceable review closure
The following internal checklist assists in setting up and validating the procedure. It complements your analytical and quality-related requirements; it is not a complete SOP for every laboratory.
- Project, sample size, result sets, and time period under consideration are clearly defined.
- The installed version, audit trail settings, and required rights are known.
- Raw data, method versions, calculations, and result histories are professionally attributable.
- The filters employed do not inadvertently exclude relevant prior activities.
- Reintegration and other outcome changes have a verifiable rationale.
- System-wide events with data involvement are communicated to the responsible reviewers.
- Suspicious items are described neutrally and appropriately investigated.
- Open questions have responsible parties and a decision on further use of the data.
- Review and sign-off refer to a specifically defined state.
- The procedure is reviewed upon relevant changes to versions, roles, or processes.
Project experience and collaboration
Tailor the audit trail review to your laboratory workflow
With laboratory staff and QA, we establish which Empower changes matter to the assessment of an analysis. Using a known change, we examine event selection, the reviewer, the information they need and the escalation route. This can clarify the scope of an existing review. If an event cannot be traced to the decision on the result, we examine the whole affected handover with your team. The record should show what was reviewed and how anomalies were handled.
Frequently asked questions about Empower Audit Trail Review
What distinguishes the Data Review from the Audit Trail Review in Empower?
The data review evaluates the investigation itself, such as sample assignment, chromatograms, integration, calibration, and calculations. The associated Audit Trail Review assesses relevant actions and modifications to these records. Both are professionally linked but represent distinct review steps. An event log does not replace an analytical assessment; a plausible results report does not automatically establish the data history and completeness.
Is the Result Audit Viewer sufficient for all audit trails?
No. It consolidates key information related to a result set and supports result and method comparisons. System-wide activities such as user management should be included in a separately defined control area. Additionally, verify that selection and filtering criteria cover the clinically relevant population. The suitability of a view is derived from the specific inspection question, not merely because it displays audit trail data.
Does "Created Manual Result" indicate that a peak was improperly modified?
No. Waters explains this entry also through the individual handling of injections in Review. Clarify which specific action took place and what impact it had. Only by comparing data, method, result, and rationale can an assessment be made. The event name may serve as a starting point for investigation, but it does not constitute evidence of a defective integration or intentional manipulation.
May we retain only the new result during a reintegration?
For the regulated laboratory process under consideration, the final result alone is insufficient. The result history must remain traceable. The FDA explicitly emphasizes the retention of every result in the case of reprocessed chromatography. Additionally, it must be clearly defined which result is functionally used and the rationale for such selection. The non-use of a previous result constitutes a justified decision and must not be concealed.
How do we verify that the correct method version has been used?
Link the specific result to the actually used method version and compare relevant changes. A mere method name is insufficient when multiple versions are involved. Empower provides method history and parameter comparison features. Assess differences based on your approved analytical method. A technically visible version difference does not constitute evidence that the change was permissible or that the result was affected.
Can an earlier result timestamp represent normal system behavior?
Yes, in certain workflows. Waters documents an Empower 3 scenario in which individual processing followed by saving the results together explains this sequence of times. Check whether the explanation fits the actual event, workflow and system version. If the sequence remains implausible, document and investigate the finding before closing it.
Sources: [6]
Who is responsible for reviewing the System Audit Trail?
Designate a technically qualified responsible individual with appropriate authorities and a clearly defined scope of responsibility. This role may differ from the analyst responsible for evaluating analytical results. System events with potential data impact must be communicated to the laboratory. Additionally, establish an appropriate control over administrative activities to ensure that the same technical authority does not simultaneously serve as the sole reviewer of its own modifications.
Does Annex 11 require a monthly Empower review?
No, Annex 11 specifies regular reviews, but does not establish a general monthly interval for empowerment. Justify scope, frequency, and additional triggers based on risk. In the FDA context, the Guidance states that data-related audit trails must be reviewed in conjunction with the associated records and their prescribed frequency. A periodic system review shall not simply replace a required review specific to the particular data usage.
Does the audit trail identifier prove the complete review?
It documents a confirmation within the intended sign-off context. The function in Empower 3.8.0 requires specific settings and rights. It neither evaluates the selection of data scope nor automatically assesses the quality of your review. Therefore, specify which records the signer has reviewed, where findings are documented, and what final status has been confirmed. Review the actual display after changes to results or configuration.
When will a abnormality become a deviation or an OOS examination?
An out-of-specification (OOS) result must be investigated according to the designated OOS procedure. An unexplained change, missing documentation, or unauthorized access may constitute another quality-related investigation trigger. Document the observation and assess its potential impacts in accordance with your deviation investigation procedure. A later plausible value does not invalidate the original finding.
Sources and review date
As of October 6, 2026. Product functionalities are supported by open Waters documentation, particularly Empower 3.8.0 and 3.9.0; settings and actual customer status must be verified separately. Regulatory framework comprises Human Medicinal Product GMP and the applicable FDA jurisdiction. Tables, examples, and checklist are standalone work aids.
- Waters: Audit Trails in Empower – System Audit Trails, Tip 368
- Waters: Audit Trails in Empower – Project Audit Trails, Tip 369
- Waters: Audit Trails in Empower – Method Audit Trails, Tip 370
- Waters: Audit Trails – Result Audit Viewer, Tip 373
- Waters: Created Manual Result in Empower 3
- Waters: Result time prior to method time for individual processing
- Waters: Empower 3.8.0 Release Notes
- Waters: How do you sign off results in Empower?
- Waters: Empower 3.9.0 Release Notes
- European Commission: EU GMP Annex 11, January 2011 revision
- European Commission: EudraLex Volume 4
- eCFR: 21 CFR 211.194 – Laboratory records
- eCFR: 21 CFR Part 11 – Electronic Records; Electronic Signatures
- FDA: Data Integrity and Compliance With Drug CGMP, December 2018
- FDA: Investigating OOS Test Results for Pharmaceutical Production, Revision 1
Sources checked: .
Tailor the Empower-Review to your data and roles
We support you in aligning the review scope, responsibilities, and evidence based on your actual laboratory workflow. Daniel Herrmann Consulting works with you to identify the relevant settings and verification steps for a traceable decision.
- Audit and inspection preparation →
Go into an audit or inspection with accessible evidence and a prepared team. We begin with a walkthrough, rank open issues by risk and rehearse the inspection with your team.

