Compliance
Data Integrity Gaps Between Paper and Electronic GMP Records
Spot data integrity gaps between paper and electronic GMP records, with seven handoff checks and a pharmaceutical QC example using Excel and SharePoint.
Imagine the morning quality meeting at a pharmaceutical manufacturing site. The report shows a quality control (QC) sample as complete. Then someone asks, "Has the result been reviewed?"
The analyst finished the test. The Excel calculation is saved. The SharePoint List says "Complete". But the worksheet is still waiting for review. Now someone has to follow the paper, calculation file and tracker entry to work out what is actually finished.
If that sounds familiar, the frustration is not just finding a file. It is having to reconstruct the story before anyone can confidently answer a simple status question.
Here, "complete" changed meaning as the work moved between people and records: calculation completed became testing and review completed. The report accurately repeated the List's status, but the status did not reflect the outstanding review.
Start with the handoff, not another tool. To address data integrity gaps between paper and electronic good manufacturing practice (GMP) records, make clear what is being passed on, which source supports it, who checks it and what the next status means. When records disagree, keep the discrepancy visible and refer it to the responsible owner rather than choosing whichever value looks most plausible.
This article is a practical review aid for pharmaceutical manufacturing Quality, QC and Digital Quality/IT owners. It focuses on the gaps at handoffs, rather than a full comparison of paper and electronic systems. The QC example makes those boundaries concrete; the same questions can help teams review handoffs in production logbooks and quality-event records. The questions, template and worked example are recommendations to adapt through your organisation's approved procedures, not a universal regulatory checklist or a determination of compliance.
Start at the boundary between records
Consider a pharmaceutical manufacturer's QC laboratory without an integrated laboratory information management system (LIMS). Samples are logged in a SharePoint List. Analysts record test work on paper worksheets, with supporting instrument files or printouts, then enter selected values into an Excel calculation workbook. The workbook template is held in a controlled SharePoint document library; the completed calculation record is retained at the location designated by the laboratory's procedure. A separate Excel or Power BI report reads sample status from the List.
These are different roles, not two competing trackers:
| Part of the workflow | Purpose in this example | Handoff question |
|---|---|---|
| SharePoint List | Track sample identity, assigned analyst, due date and testing/review status | Which evidence supports each status change? |
| Paper worksheet and supporting instrument records | Record test work and its source observations | Can the values be traced to the right sample, test and source context? |
| Controlled SharePoint document library | Hold the approved Excel calculation workbook template | Which approved version is intended for this calculation, and how is its use recorded? |
| Completed Excel calculation record | Hold entered values, calculations and associated review evidence | Can inputs, units, workbook version and review state be followed? |
| Linked Excel or Power BI report | Present sample-status information from the List | Which List data, filters and refresh point does the report represent? |
The List is the status tracker; the controlled library holds documents and workbook files. Keep the sample's List item connected to the relevant calculation record and review evidence.
Before reviewing the transfer, agree which record is authoritative for each purpose. Do not assume that the newest file, the electronic copy or the most convenient dashboard overrides the controlled source.
Entering a result into Excel or changing a List status does not replace the underlying worksheet, relevant instrument records or review evidence. Establish how the records remain connected under the approved procedure; do not treat copying a value as permission to discard its source.
For the wider record-versus-register distinction, read Beyond Excel: a practical approach to quality registers. Here, the focus is narrower: what crosses each boundary, and what happens when the transfer does not match.
Map one handoff before redesigning the whole process
Choose a real transfer and walk through it with the person who sends the information and the person who receives it. Include the quality reviewer where the transfer supports a quality decision.
Use the following as a starting template. It describes the handoff; it is not permission to change a controlled process.
| Handoff question | What to capture |
|---|---|
| What starts the transfer? | The event or approved process step that makes information ready to pass on |
| What is the source? | The identifiable record, location and relevant version or state |
| Where does the information go? | The destination record and its intended use |
| What needs to travel together? | Values, identifiers, units, dates, status definitions and supporting context relevant to that use |
| Who sends and receives it? | Named roles responsible for entry, checking and exception decisions |
| How is the transfer checked? | The agreed comparison and evidence appropriate to the information's use and risk |
| What if something disagrees? | The discrepancy owner, escalation path and limits on downstream use pending a decision |
| What remains available afterwards? | The source reference, check result, exceptions and recorded disposition |
Responsibility for copying a value is not necessarily authority to interpret it, approve a correction or approve a test result. Make those boundaries clear.
Seven paper-to-electronic GMP handoff gotchas and quick checks
A copied value can look right while its identity, meaning or review state is wrong. Use these illustrative failure modes to start a conversation about your own workflow, not as findings about a particular laboratory.
| Gotcha | Quick check |
|---|---|
| Right result, wrong sample. A value is entered against a similar sample identifier or the next spreadsheet row. | Match the sample and test identifiers across the worksheet, instrument record, calculation and List item. |
| The number travels; the context does not. A value is copied without its unit, dilution factor or method reference. | Compare the value and the context needed to interpret it, including any conversion. |
| Yesterday's workbook becomes today's template. A local copy or previous sample's file is reused. | Identify the approved template version used and check for carried-over sample details or inputs. |
| "Complete" means different things. Calculation finished is treated as result reviewed. | Keep test performed, calculation completed and result reviewed distinguishable in the List. None alone means batch release. |
| A correction stops at the paper. A source value is corrected, but the calculation or status still reflects the earlier entry. | Assign who assesses affected calculations, List entries and reports, and records the follow-up. |
| The dashboard looks current but is not. An old refresh or a filter hides work awaiting review. | Check the report's refresh point and filters against the relevant List items, including pending reviews. |
| Matching counts hide a missing sample. One duplicated entry offsets one omission. | Compare the expected sample identifiers, not just the total number of rows. |
Take away five questions: right sample, right context, right workbook version, right review status, right report view?
The appropriate check depends on intended use and risk. This is not a recommendation to apply the same checking method to every field or to treat a spot check as evidence that all records match.
For background on the record-quality principles behind these checks, read PharmOut's explanation of ALCOA principles for paper and electronic records: attributable, legible, contemporaneous, original and accurate. PharmOut is a QikSolve partner; this is supporting reading, not a substitute for the applicable requirements or your approved procedure.
When the sample is tested but the result is not reviewed
The following is a teaching example, not a customer result, system test or approved procedure.
A pharmaceutical manufacturer's QC laboratory logs sample LAB-026 in a SharePoint List and assigns a test. An analyst performs the test, records observations on a paper worksheet and references the supporting instrument records. The analyst uses the approved Excel calculation template from the controlled SharePoint document library, enters the relevant values and saves the completed calculation record under the laboratory's procedure.
In this example, the worksheet and relevant instrument records support the observations, the completed workbook supports the calculation, and the designated review record establishes review of the result. The List tracks progress; neither the List nor the report replaces those records. The procedure distinguishes "Test performed", "Calculation completed" and "Result reviewed".
During a handoff check, the reviewer follows sample LAB-026:
| Stage | What the reviewer finds | Question to resolve |
|---|---|---|
| Paper worksheet and instrument references | Test performed; result review not recorded | What review remains outstanding? |
| Calculation template in the controlled SharePoint document library | Approved template version identified for use | Does the completed calculation record identify that version? |
| Completed Excel calculation record | Calculation finished; result review still pending | Has calculation completion been mistaken for review completion? |
| Sample item in the SharePoint List | Sample marked "Complete" after calculation | Does the status distinguish testing, calculation and review? |
| Linked Excel or Power BI report | Refreshed after the List update; "Complete" samples excluded from the outstanding-work view | Has work awaiting result review disappeared from oversight? |
The report follows the List in this example, but that agreement does not establish that the result has been reviewed. The discrepancy begins where calculation completion is treated as completion of the whole testing-and-review process.
An illustrative response is:
- Record the disagreement. Link the sample's List item, worksheet, instrument references, completed calculation record and relevant report view. Identify the template version separately from the completed record's location.
- Refer the decision. Ask the designated laboratory/quality owner to establish the correct state and outstanding review through the approved procedure. The person updating the List does not invent a review approval.
- Assess downstream use. Identify whether the "Complete" status removed the sample from the outstanding-work view or affected other decisions. Agree how affected information will be handled while the discrepancy is unresolved.
- Follow the authorised correction path. Once the owner has established the basis, address affected entries without silently overwriting their history. Check the linked report after the appropriate refresh; do not manually patch a report to hide an unresolved List discrepancy. See correcting a GMP record without losing its history for the separate correction question.
- Keep the disposition connected. Retain the decision, supporting references, responsible roles and follow-up through the organisation's approved mechanism.
The discrepancy remains open until the authorised review establishes its disposition. This example does not claim that a correction was executed or that any control passed a test.
Leave evidence the next reviewer can follow
A note saying "saved in SharePoint" may not explain whether a calculation file was saved in a library, a sample status changed in a List, or a review was completed.
A useful handoff note for this example could begin:
Sample LAB-026: test work and calculation completed; result review remains pending. SharePoint List item shows Complete, and the linked report excludes the sample from outstanding work. Worksheet, instrument references, completed calculation record and template version identified for review. Referred to the designated laboratory/quality owner to establish the correct status and assess affected reporting. This handoff check does not approve the result or establish batch release.
This is an example of a discrepancy note, not a completed approval or a replacement for the organisation's records. An actual record would also need the attribution, timing, references and review evidence called for by the applicable procedure.
If similar discrepancies recur, review the handoff design: unclear status definitions, an unassigned receiving role or a source-update step that nobody owns may need attention. Correcting a destination value does not, by itself, address the recurring process problem.
Choose one handoff to review next
Pick a transfer used in a real quality report or decision. Map its source and destination, agree who owns exceptions, and ask whether someone outside the original conversation could follow the check and its outcome.
Where the uncertainty spans teams or systems, use practical data and system governance for quality teams to frame the authoritative-record, ownership and decision-rights questions. That is the next step before selecting a technical fix.
Regulatory context and review limits
For Australian medicinal-product good manufacturing practice (GMP), start with the Therapeutic Goods Administration's guidance on PIC/S GMP adoption and applicability, checked 28 October 2026. The page identifies PE 009-17, dated 25 August 2023, applying from 1 September 2025, except Annexes 4, 5 and 14. Assess the product, activity, exemptions and relevant phased implementation provisions; a newer overseas publication is not automatically an Australian requirement. The handoff checks above are practical recommendations, not clause-specific transfer obligations.
For further primary-source reading, see PIC/S guidance on good practices for data management and integrity in regulated GMP/GDP environments, PI 041-1, dated 1 July 2021. The linked document's identity and date were checked 28 October 2026; it is separate from the Australian-adopted GMP Guide discussed above.
The recommendations above are not presented as quotations or mandatory controls from that guidance. They do not establish that a scan or transcription replaces an original, authorise source-record disposal, or demonstrate that Excel, SharePoint or Power BI is compliant for a particular use.