Manual, Hybrid and Electronic Approvals: Compare the Controls
Manual vs electronic approvals is the wrong first question for validation records. An approval is only as good as what a later reader can establish from it: who approved, with what authority, meaning what and of which version, whether the record is paper, hybrid or electronic.
Written by CSV/CSA Maturity Editorial TeamPublished 10 min read
On this page8 sections
Whether an approval is handwritten, electronic or a mixture of the two tells you little about how well it is controlled. What matters is whether you can show who approved, that the person was authorized, what the approval meant, exactly which version it covered, and that the record has stayed intact and retrievable since. Each medium can meet those tests and each can fail them. A hybrid process adds one more place to look: the handoff between media.
This comparison is for record owners and quality owners who approve, or rely on, the records behind computer system validation (CSV) and computer software assurance (CSA) decisions: requirements, test summaries, release decisions and change approvals. CSV here does not mean comma-separated values. The article compares controls. It does not decide which regulations apply to your records, and it is not a determination of Part 11 or Annex 11 applicability.
Start with the authoritative record and the purpose of approval
Two definitions come before any comparison of media.
The authoritative record is the version your organization relies on and would produce if asked to show what was decided. There is one per document. Everything else is a copy, however convenient. In a paper process that is usually the signed original. In an electronic process it is the version held in the controlled system. In a hybrid process it has to be decided and written down: the signed paper, the scan or the electronic file.
The purpose of an approval is the specific decision the signature records. "Approved" can mean that the author stands behind the content, that a reviewer checked it technically, that quality released it for use, or that a system owner accepted an open issue. A signature whose meaning has to be guessed is weak in any medium.
With those two fixed, an approval should let a later reader answer four questions: who approved, with what authority, meaning what, and of exactly which version.
Criterion table: the same controls in three media
Illustrative examples. Each cell describes one way a control can be met. The cells are not a list of requirements and they are not product features.
| Criterion | Manual (paper) | Hybrid | Electronic |
|---|---|---|---|
| Identity of the approver | Handwritten signature with printed name. A signature log lets a reader match the two | Ink signature on a printed page that is later scanned. The scan is tied to the signer through the paper original or the signature log | The approval is recorded against an individual user account. A shared account breaks the link |
| Authority to approve | A role list or delegation record that was in force on the signing date | The same, plus clarity about who may approve at each stage of the handoff | Approval rights assigned by role, with a record of who held them and since when |
| Meaning of the approval | The signature block states it: author, reviewer or approver | The meaning is stated on the signed page and survives when only that page is scanned | The meaning is shown with the approval entry |
| Tie to the exact version | Version and page count on every signed page | The signed page names the document version, so it cannot be filed in front of a different file | The approval is attached to one stored version. A later edit creates a new version that needs its own approval |
| Changes after approval | Corrections keep the original entry legible and are dated and attributed | A rule says which medium is corrected first and how the other copy is brought into line | Edits are recorded with who and when. Approved versions are locked or superseded |
| Access and protection | Controlled storage and a sign-out record | Both the paper store and the file location are controlled | Access by role, with a periodic check of who holds which rights |
| Retention and retrieval | An indexed archive and an occasional retrieval test | An index that links the paper original to its electronic copy | Backup, a restore test and export in a readable form |
Nothing in the electronic column arrives by installing software. Account management, role assignment, configuration and periodic checks are your organization's work. Nothing in the manual column is inherently weaker either. A paper process with a current signature log, controlled storage and a tested archive can be better controlled than an electronic workflow with shared accounts and approval rights nobody has reviewed. The reverse is just as possible.
Inspect the handoffs in a hybrid process
A hybrid process is one where a record and its approval do not stay in the same medium for their whole life. Typical handoffs are:
- an electronic document is printed, signed in ink, scanned and filed;
- an approval is given by email and attached to the record afterwards;
- an approved file is exported from a controlled system to a shared location for daily use;
- a paper form is transcribed into a system by someone other than the signer.
At each handoff ask the same four things. Is the content identical on both sides? Which side is the authoritative record? Who performs the handoff, and when? Could a reader a year from now tell all of that from the records alone?
Synthetic sample: the stored version differs from the approved version
Synthetic sample. The record, the roles and the days below are invented to illustrate a review. They do not describe a real system or organization.
| Day | What the records show | What the reviewer needs to clarify |
|---|---|---|
| Day 1 | A test summary, version 1.2, is printed. The authorized approver signs and dates the last page in ink | Does the signed page state the version it belongs to? |
| Day 1 | The signed page is scanned and stored in the team repository as a separate file | Is the scan the whole document or the signature page only? Which copy is the authoritative record? |
| Day 3 | The electronic file is edited and saved as version 1.3. The change note reads "corrected test case reference" | Who made the edit, and under which procedure? Does that procedure allow such a correction without a new approval? |
| Day 4 | The release decision record cites "test summary, version 1.3" | Did the approver see version 1.3? Did the decision rely on 1.2 as signed or on 1.3 as stored? |
From these records the reviewer can state two things. The version cited in the release decision is not the version that carries the signature. The records at hand do not show an approval of version 1.3.
The reviewer cannot state, from these records alone, that the edit was improper, that the difference in content matters or that any requirement was breached. An allowed administrative correction that was never cross-referenced, a second approval filed somewhere else and a real gap all look identical at this point.
The useful output is therefore an open question with an owner and a date, a side-by-side comparison of versions 1.2 and 1.3, and a note of which control would have removed the doubt: a version identifier on the signed page, a written rule for edits after approval, or a release decision that cites the approved version. What follows, if anything, is decided through your quality system's own process. An internal check like this one is an evidence review, not an audit, and the difference between a self-assessment, an evidence review and a formal audit matters when you report what you found.
Applicability boundaries: what the medium does not decide
The medium of an approval does not decide which rules apply to it, and a comparison of controls is not an applicability decision. Three scope points are worth having in front of you.
Part 11 in the United States. 21 CFR Part 11 is the FDA regulation on electronic records and electronic signatures. FDA's 2003 guidance on the scope and application of Part 11 [S06] is nonbinding. It says the agency will interpret the scope of Part 11 narrowly, and that it intends to exercise enforcement discretion for the validation, audit trail, record retention and record copying requirements under the conditions the guidance describes. It also says Part 11 remains in effect and that records must still meet the underlying predicate rules (Part 11 scope guidance [S06]). Our reading is that the first questions are which records a predicate rule requires and whether those records are kept in electronic form.
Medical device production and quality management system software. FDA's final guidance on computer software assurance, issued in February 2026, gives nonbinding recommendations for software used as part of medical device production or the quality management system. It says Part 11 generally applies when a document required under 21 CFR Part 820 is kept in electronic form, and it refers readers to the 2003 guidance (Section V.B of the FDA guidance [S03]). It also says the Part 11 enforcement discretion on validation does not extend to the validation requirements for production and quality management system software (FDA guidance on Part 11 and validation [S03]). The same document lists review and approval "when appropriate" among the contents of the assurance record (Section V.A.6 [S03]).
EU GMP for medicinal products. Annex 11 [S36] has a clause on electronic signatures (clause 14) and addresses the recording of who entered or changed data (clause 12.4). As of November 12, 2026, the EudraLex Volume 4 index [S04] lists Annex 11 as the January 2011 revision.
In our view, two conclusions do not follow from any of this:
- That handwritten signatures are always sufficient. A paper approval still has to meet whatever rules apply to the record, and a process that looks manual may contain electronic records at some stage.
- That an electronic signature automatically meets Part 11 or Annex 11. Applicability depends on the record and its context, and the controls that count are the ones your organization configures and operates.
Choose an internal evidence sample
A small sample tells you more than a description of the process. Five steps are enough to start:
- List the record types that carry assurance decisions in your scope, for example a requirements approval, a test summary, a release decision and a change approval.
- For each medium you use, pick a few recent records. Include at least one that crossed a handoff. Your quality system may set its own sampling rules; if it does, follow them.
- For each record, answer the four questions (who, authority on that date, meaning, exact version), then note any change since approval and where the authoritative copy is kept.
- Ask someone who did not file the record to retrieve one older example without help.
- Write down what could not be shown. That is an unknown to resolve, not yet a failure.
A plain worksheet is sufficient. The first row is a synthetic sample based on the handoff above.
| Record sampled | Medium | What the sample showed | What could not be shown | Owner of the open question |
|---|---|---|---|---|
| Test summary, version 1.3 (synthetic) | Hybrid | Signature, date and meaning on the page for version 1.2 | An approval that covers version 1.3 | Record owner |
Keep the worksheet with your own records. Do not paste system names or record contents into a public tool.
How this maps to the assessment
In the CSV/CSA Maturity Assessment, the practice area "Approvals and record controls" has two questions. Q14 asks how responsibility for review and approval is controlled. Q15 asks how the records used for assurance decisions are protected and kept trustworthy. Each is answered with one of four practice descriptions, scored 0 to 3, or with "I do not know yet".
Three points from the model are relevant here:
- The questions describe controls, not media. The note on Q14 reads: "Apply this to paper, electronic or hybrid records. Buying an e-signature product does not itself earn points."
- The area score is a simple ratio. With two questions, it is 100 times the sum of the two answers, divided by 6. Answers of 2 and 2 give 66.7. Answers of 3 and 0 give 50.0.
- Q15 carries a priority review prompt. An answer of 0 asks you to review the practice independently of the overall score, and an answer of "I do not know yet" asks you to verify the evidence. Neither changes the arithmetic.
The methodology page sets out how area scores and review prompts are calculated. If you have a sample in hand, the evidence line for each question tells you what to compare it with, and you can read the approvals and record controls area of the assessment before you start. The assessment asks for a scope, takes you through seven practice areas and a review step, then asks for your first name, last name, work email and company before it shows the profile, which you can download as a PDF.
Limitations
- The criterion table gives illustrative examples. It is not a complete control set for any medium, and it is not a checklist against any regulation.
- The handoff is a synthetic sample. A real case needs the actual procedures and records, read by someone qualified to judge them.
- The regulatory section reports scope statements from the linked sources as checked on November 12, 2026. It does not interpret 21 CFR Part 11 or Annex 11 for your records.
- A self-reported answer about approvals says what the team believes. Only an examination of records can show what is in place.
Related reading
- Self-assessment vs quality audit: what an internal sample can and cannot establish before you report it.
- Review validation traceability with one evidence chain: how to identify the exact evidence version behind a decision, the same question the handoff example raises.
- CSV/CSA Maturity Assessment methodology: the formulas, answer states and priority review prompts referred to above.
This article describes assessment model csa-maturity-2.0.0.
Sources and references
- S03Computer Software Assurance for Production and Quality Management System Software: Guidance for Industry and Food and Drug Administration Staff. U.S. Food and Drug Administration.
Final guidance document (PDF), issued February 3, 2026; nonbinding; supersedes the final guidance issued September 24, 2025.
- S04EudraLex - Volume 4 - Good Manufacturing Practice (GMP) guidelines. European Commission.
Official index of EU GMP guidelines; listed Annex 11 as 'Computerised Systems (revision January 2011)' when checked on November 12, 2026.
- S06Part 11, Electronic Records; Electronic Signatures - Scope and Application. U.S. Food and Drug Administration.
Final guidance (nonbinding) from 2003; the cover is dated August 2003 and FDA's web listing shows September 2003.
- S36EudraLex Volume 4, Annex 11: Computerised Systems. European Commission.
EU GMP guideline, revision 1, with a stated deadline for coming into operation of June 30, 2011; the version linked from the EudraLex index on November 12, 2026.