DHR vs eDHR: Choose the Record Format by Evidence
Compare four record arrangements using the same fictional evidence packet.

In Short
- DHR vs eDHR is not a contest between a folder and a screen.
- The scope is US finished devices intended for human use.
- Section 11.1(b) applies to electronic records created, modified, maintained, archived, retrieved or transmitted under FDA regulatory record requirements, and to the…
Decide what each preserves, what a reviewer can actually retrieve, and what remains unproven before choosing a format.
Choose by Evidence, Not by Format
DHR vs eDHR is not a contest between a folder and a screen. The useful decision is which arrangement preserves the evidence your process needs, makes its authority understandable, and supports the people making decisions from it. Paper, controlled scans, hybrid records and structured electronic records can present the same production history very differently.
Start with the required records and their intended use, including Part 11 applicability, before selecting a format. Then compare arrangements against identical evidence. A paper signature does not establish that an attachment is complete; a searchable PDF does not establish which instruction revision governed production; a structured completion flag does not establish an authorized disposition.
This guide uses one fictional packet throughout. It contains a named instruction revision, acceptance evidence, a corrected observation, an externally held report, a documented hold and an access problem. Each arrangement receives the same underlying facts. Differences arise from how those facts are maintained, presented and handed to another reviewer—not from giving one option an easier production case.
For terminology, use the DHR reference and eDHR reference. Here, the question is format choice: what can a reviewer establish, what remains inaccessible, and what additional evidence would justify relying on the proposed arrangement?
Separate Current and Historical Requirements
The scope is US finished devices intended for human use. Section 820.1 covers the methods, facilities and controls used for the listed device activities, including manufacturing, packaging and labeling. Manufacturers performing only some covered operations comply with the requirements applicable to those operations. This is not a universal record-format guide for every regulated product or market.
The transition date supplied for this comparison is 2 February 2026, when QMSR became effective. Current requirements are assessed under that framework. Former §820.184 is used only to interpret historical DHR terminology; its old list is not presented as the current production-record checklist.
| Reference | What the Selected Text Establishes | Consequence for This Comparison |
|---|---|---|
| Current §§820.7 and 820.10(a) | ISO 13485:2016 is incorporated by reference; the documented QMS must comply with its applicable requirements and other applicable Part 820 requirements. | Do not equate an application name or document title with fulfillment of the QMS obligations. |
| Current §820.35(c) | UDI must be recorded for each medical device or batch of medical devices. | An internal packet number is not, by itself, evidence that this recording requirement has been met. |
| Current §820.45(b)–(c) | Labeling release for use and results of the specified labeling inspection must be documented. | Changing the record medium does not remove these documentation obligations. |
| Historical §820.184, viewed on 1 February 2026 | The former DHR provision required records for each batch, lot or unit and allowed listed information to be included or referenced by location. | Use this provision for historical interpretation, not as current law or a newly mandated electronic schema. |
Read the selected current Part 820 provisions separately from historical §820.184. This comparison does not establish the detailed contents of the incorporated ISO clauses.
As evaluation practice, map your required evidence before comparing representations. A faithful reproduction of an incomplete packet remains incomplete. Conversely, adding fields merely because a software template contains them does not establish that those fields address the operations your organization performs.
Scope Part 11 Before Format Selection
Section 11.1(b) applies to electronic records created, modified, maintained, archived, retrieved or transmitted under FDA regulatory record requirements, and to the electronic submissions described there. It excludes paper records that are or have been transmitted electronically. The assessment therefore begins with the required record and its electronic use—not with the words PDF, cloud, scan or eDHR.
For each evidence item, ask what requirement it supports and what role the electronic representation performs. Is it a convenience copy of an authoritative paper record, the maintained record used for review, an electronic original, or a reference to another record? Describe actual practice as well as the written designation. Calling a file a convenience copy does not answer what happens when reviewers rely on it instead of the stated original.
- Identify the Required RecordName the underlying record requirement and the evidence item that supports it.
- Describe Its Electronic UseIdentify creation, modification, maintenance, archiving, retrieval or transmission, including reliance during review.
- Resolve Record and Signature ScopeAssess §11.1 and distinguish record controls from any applicable signature controls.
- Identify System BoundariesDetermine who controls access and whether closed- or open-system provisions apply.
- Compare Candidate FormatsEvaluate how each arrangement will satisfy the identified controls and demonstrate its intended behavior.
Section 11.3 distinguishes closed and open systems by whether access is controlled by persons responsible for the electronic records' content. Hosting location alone does not answer that definition. Section 11.10 establishes closed-system controls; §11.30 addresses open systems, including additional measures as necessary under the circumstances.
| Selected Applicable Control | Regulatory Anchor | Evidence to Request in a Format Evaluation |
|---|---|---|
| System validation | §11.10(a): accuracy, reliability, consistent intended performance and discernment of invalid or altered records. | Demonstrated behavior for the intended record functions, not merely a product label. |
| Copies and retrieval | §11.10(b)–(c): accurate and complete human-readable and electronic copies; protection enabling accurate and ready retrieval throughout retention. | A retrieved package with its supporting records, plus an explanation of any exclusions. |
| Access and authority | §11.10(d), (g): authorized access and authority checks. | Evidence that reviewers can reach needed records while users cannot perform excluded actions. |
| Electronic changes | §11.10(e): secure, computer-generated, time-stamped audit trails; changes must not obscure earlier information. | The original entry, subsequent change and relevant audit-trail evidence—not only the latest displayed value. |
| Applicable signed electronic records | §11.50: signer name, signing date/time and meaning, including human-readable forms; §11.70: signature-to-record linkage. | The signed record state and signature details, with evidence of linkage rather than a detached signature image. |
The final column contains proposed evaluation requests, not additional regulatory wording. These are selected controls, not a complete Part 11 assessment. Section 11.1(d) permits qualifying electronic records to be used instead of paper, subject to its conditions and unless paper is specifically required; it is not a mandate to buy an eDHR application. See the supplied Part 11 text.
Open One Common Evidence Packet
Authority register AR-508 assigns responsibility by evidence type. The controlled instruction and transition record establish the production baseline; the observation source and correction record establish the reading history; the external report establishes its own findings; and DISP-508 establishes disposition. No summary may silently replace those sources when they disagree.
| Evidence Item | Same Facts in All Four Arrangements | Question It Leaves Open |
|---|---|---|
| Instruction baseline | Execution identifies WI-MX-12 revision C. Revision D was approved later. Transition record EC-27 applies D to new jobs and leaves this execution on C. | Can another reviewer retrieve C and EC-27 rather than only today's D? |
| Acceptance method | TP-MX-09 revision B specifies an illustrative acceptable range of 0.40–0.60 N·m. | Can the reviewer connect the criterion to this observation, rather than to a later method? |
| Observation source | QL-508 records 0.52 N·m for MX4-508. The production entry initially transcribed it as 0.25 N·m. | Is the source available to substantiate the correction? |
| Correction | CR-508 identifies Elena Park, preserves 0.25 and 0.52, records correction timing and explains transposed digits by reference to QL-508. No new measurement occurred. | Has the effect of the incorrect entry on prior review been assessed? |
| External attachment | ER-508 revision 1 has two pages. Only page 1 is held locally. Its summary says pass; page 2 contains scope and approval information required by the fictional review procedure. | Does the complete approved report support this device and the stated conclusion? |
| External access | The stored portal reference resolves to latest, whose metadata now identifies revision 2. The former contact's account is disabled; the assigned QA reviewer cannot open the report. | Can the relevant revision 1 be obtained without substituting an unassessed revision? |
| Disposition | QA reviewer Rowan Hale signed DISP-508: hold pending complete external evidence and correction-impact review. Neither dependency is closed. | Will the handover preserve the hold and its reasons? |
| Completion summary | A separate cover or screen says operations complete. It refers to recorded operations, not distribution authorization. | Will someone mistake completion for release? |
The observation is within the fictional criterion after correction, but that does not settle disposition. The external report remains incomplete and the correction-impact review remains unfinished. These are common unresolved facts, not defects assigned selectively to a particular medium. No arrangement receives permission to release the device during this exercise.
PKT-MX-508 and AR-508
Identify the device and direct the reviewer to the controlling source for each evidence type.
WI-MX-12 C and EC-27
Together establish why revision C governs this execution despite the later revision D.
QL-508, CR-508 and TP-MX-09 B
Connect the original observation, transcription correction and applicable acceptance criterion.
ER-508 Revision 1
Supports external findings only when the relevant complete report can be retrieved and assessed.
DISP-508
Records the hold and depends on unresolved external-evidence and correction-review questions.
Compare Four Record Arrangements
The following are four hypothetical representations of the same packet, not four competing originals in a live process. Their stated weaknesses are deliberately chosen demonstration conditions, not claims about how paper or software normally behaves. The comparison does not require running four systems in parallel.
| Arrangement | Authoritative View in This Exercise | What It Preserves | What the Handover Loses or Leaves Unavailable |
|---|---|---|---|
| Paper | Indexed original local documents, with controlled copies of C, EC-27 and the method; DISP-508 controls status. | Visible original and corrected entries, source observation, correction explanation and signed hold. | The supplied folder is readable, but access to the controlled master archive depends on an absent custodian. External page 2 is unavailable. |
| Controlled scan | An image collection designated as authoritative for local records after a documented local-page reconciliation. | Images of the local source, both correction values, historical instructions and disposition. The image is distinguished from OCR text. | OCR still returns 0.25. The demonstrated download omits QL-508, although the internal image collection contains it. External page 2 remains unavailable. |
| Hybrid | Paper controls the observation; electronic records control CR-508, instruction references and DISP-508, as specified in AR-508. | The original observation and electronic correction/disposition history remain distinct and linked in the native review arrangement. | The demonstrated electronic handover omits the paper observation. A recipient cannot substantiate 0.52 from that handover alone. External page 2 remains unavailable. |
| Structured electronic | Identified data events control the observation and correction; pinned documents hold C and the method; DISP-508 controls disposition. | The native view retains both values, their relationship, source reference and hold. Instructions are pinned to identified revisions. | The demonstrated PDF shows only 0.52 and operations complete, omitting correction history and disposition. External page 2 remains unavailable. |
Paper: Readable Evidence, Demonstrated Custody Needed
The paper folder answers several questions directly. A reader can compare the source observation with the crossed-through transcription and its correction, then read the hold. That is meaningful evidence, not a consolation prize for lacking software. However, the folder's availability during a demonstration does not establish continuing access to the controlled master archive. Ask who can retrieve and authenticate the referenced baseline when the usual custodian is absent. An internal retrieval objective should reflect the actual process, not an invented FDA stopwatch.
Controlled Scan: Separate Image Fidelity From Search Behavior
Here, the local-page reconciliation supports the stipulated image collection, but not the completeness of the separate download. Those are different demonstrations. The stale OCR value is also a different issue from image fidelity: the image preserves the correction while search presents the earlier value. Decide whether OCR is only a locator or influences review conclusions. Verify the proposed use accordingly. Neither a successful scan nor searchable text recovers the external page that was never available locally.
Hybrid: Make the Split Explicit to the Recipient
The hybrid arrangement is not inherently contradictory: AR-508 deliberately assigns different evidence to different media. Its weakness appears at the handover. The electronic package looks self-contained but leaves the controlling observation on paper. A recipient needs either that evidence or a controlled, usable route to it. Record which part of the history the package contains and which part remains elsewhere; do not let an electronic correction appear to be an electronic original measurement.
Structured Electronic: Test Both the Native View and the Output
The native structured view preserves useful relationships in this example. Its demonstrated PDF does not. This is an output-design finding, not proof that structured records inherently lose history. Ask whether the recipient is intended to rely on that output, what companion records accompany it, and how completeness is established. A richer database is not a substitute for an interpretable handover when the recipient cannot use the database.
Run the Evidence Handover Exercise
Work with the packet and representation table before reading the findings. Assign one participant to act as the sender and another as a reviewer unfamiliar with the record. The reviewer receives only the demonstrated handover for the selected arrangement. Any explanation supplied orally must be identified as an explanation, not treated as a missing record suddenly appearing.
Produce a short reconstruction with three labels: established from available evidence, contradicted by available evidence, and not yet established. Cite packet identifiers beside each conclusion. This is not a numerical compliance score: several correct answers cannot cancel an unresolved disposition dependency.
- Authority: Which item controls disposition? Explain why operations complete cannot override it.
- Version: Does revision D's later approval prove that this execution used the wrong instruction? Identify the evidence that answers the question.
- Access: Separate the internal retrieval problem from the external portal problem. What can the assigned reviewer actually reach?
- Correction: Is 0.52 a retest result, a new observation or a corrected transcription? Identify the supporting source and the unresolved review question.
- Retrieval: For the selected handover, identify exactly which missing content prevents reconstruction. Distinguish missing-from-output from missing-at-source.
- Unresolved risk: Can the device's hold be removed on the supplied facts? What evidence would change that answer?
Next, exchange arrangements without changing the underlying facts. Compare which conclusions survive the exchange. If your conclusion changes only because one display looks more polished, revisit its evidence basis. Keep the participant's initial answer: the difference between that answer and the supported reconstruction can reveal an ambiguous status label or an undisclosed dependency worth addressing.
Read the Answerable Findings
| Question | Supported Finding | What Still Needs Evidence |
|---|---|---|
| Authority | DISP-508 controls disposition under AR-508 and records a hold. Operations complete has a narrower meaning. | Closure of the stated dependencies and a subsequent authorized disposition decision; no such decision is supplied. |
| Version | Revision C is the stipulated baseline. EC-27 explains why D applies to new jobs, not this execution. | Successful retrieval of the relevant controlled records through the intended access arrangement, not merely a link to D. |
| Access | The paper master's custodian is unavailable in the exercise. Separately, all four arrangements lack usable access to complete ER-508 revision 1. | An authorized retrieval route that returns the correct historical report and supports the required internal access. |
| Correction | 0.52 is a corrected transcription supported by QL-508, not a retest. The earlier 0.25 remains part of the history. | Completion of correction-impact review, including whether any earlier assessment relied on 0.25. |
| Retrieval | The scan download omits QL-508; the hybrid handover omits its paper observation; the structured PDF omits correction and disposition context. Those items exist in their native arrangements. | A demonstrated handover that includes or reliably exposes the omitted evidence. External page 2 is a separate source-access gap. |
| Unresolved risk | The supplied facts do not support removing the hold. A passing corrected observation does not establish the missing external approval or finish correction review. | Complete relevant external evidence, its assessment, completed correction-impact review and the resulting disposition record. |
Two distinctions matter. First, missing from the output is not the same as nonexistent. The structured correction history is available natively; the sample PDF fails to convey it. Second, a document named latest is not proof of the historical version needed. Recovering revision 2 might provide useful information, but it does not establish what revision 1 said or authorize replacing the earlier review basis without assessment.
The OCR mismatch likewise needs a use-specific response. If the text only helps locate an image and reviewers assess that image, the immediate concern differs from a workflow that uses extracted 0.25 to determine acceptance. The exercise does not supply evidence about such automated use. Record that uncertainty rather than inventing either a harmless outcome or a safety consequence.
The appropriate conclusion is therefore narrow: each arrangement has identifiable strengths and unproven dependencies; the common hold remains. These findings support targeted investigation of the proposed record arrangement. They do not establish overall compliance, device safety, a universal format ranking or readiness to replace the current process.
Identify the Decision Evidence Still Missing
A format decision requires more than identifying a correctable demonstration defect. Ask whether the proposed operating arrangement can maintain the evidence over its intended use. Separate an observed failure from an untested behavior and from a known limitation that the proposed scope deliberately excludes. All three deserve a clear description, but they support different decisions.
| Candidate | Evidence Not Yet Supplied | Useful Next Demonstration | Decision That Remains Premature |
|---|---|---|---|
| Paper | Alternative custody, controlled-master retrieval and a documented response to missing external evidence. | Have an authorized substitute retrieve the baseline and assemble the full review package without the usual custodian. | Concluding that today's readable folder establishes sustainable retrieval. |
| Controlled scan | Control of OCR use, complete export behavior and handling of later corrections to the designated image collection. | Retrieve the source image from a conflicting search result, export all indexed local evidence and demonstrate a later linked correction. | Accepting image fidelity as proof that search, export and correction processes are adequate. |
| Hybrid | A complete recipient-facing package and workable responsibility for evidence spanning media. | Give a different reviewer the paper observation and electronic history through the proposed handover route; reconstruct the correction without oral assistance. | Assuming the native reviewer's access will automatically be available to every intended recipient. |
| Structured electronic | Complete output behavior, applicable authority/signature controls and assurance for the proposed configuration. | Reconstruct the same correction and hold from the intended output; challenge an unauthorized disposition attempt in an evaluation environment. | Treating a well-linked native record as proof of complete exports or demonstrated control of all intended uses. |
| All four | Complete ER-508 revision 1, correction-impact review, applicable retention decisions and sustained access arrangements. | Obtain and assess the report, complete the review and test later reconstruction through the proposed retained record. | Removing the hold, declaring a universal winner or promising a time saving. |
For a buyer evaluation, request retained outputs and an explanation of their boundaries rather than an affirmative answer to whether a feature exists. Identify the configuration, roles and evidence used in the demonstration. Distinguish what was observed from what was described as possible. Apply the same standard to an internal archive proposal, an outsourced scanning service and a software supplier.
If review effort or retrieval time influences the decision, measure representative tasks under comparable conditions and record the exceptions. Include the work of preparing, checking and repairing the package, not only opening it. This guide supplies no basis for predicting that any format will always be faster, less expensive or less error-prone.
Apply the Selected Software Assurance Recommendations
The supplied FDA computer software assurance scope passage concerns computers or automated data-processing systems used as part of medical-device production or the quality management system. It expressly says the guidance is not a complete description of all software-validation principles. The discussion here is limited to guidance scope and the risk-based assurance approach.
The risk passage describes computer software assurance as establishing and maintaining confidence that software is fit for its intended use. It considers the risk of compromised device safety or quality if software fails to perform as intended when determining appropriate assurance effort and activities. These are nonbinding recommendations, distinct from §11.10(a)'s applicable closed-system validation requirement.
For this evaluation, describe software functions separately. Locating an image, extracting a numeric result, selecting an instruction revision and rendering disposition history have different intended uses. Ask what could be misunderstood if each function failed, which decisions depend on it, and what evidence would establish confidence in its behavior. These questions are proposed practice; they are not additional quotations from FDA.
Do not infer a prescribed test method, fixed number of successful records, automatic acceptance of supplier evidence or mandatory parallel operation from the two selected passages. They also do not establish that a generic assurance package covers your configuration. Keep the assurance question attached to the specific reliance being proposed—for example, whether the recipient can reconstruct DISP-508 from the actual export—not to the broad label electronic record.
Write a Conditional Format Decision
A defensible comparison ends with a conditional decision, not a winner's badge. Describe the review tasks the arrangement must support, the evidence it preserves, the dependencies it introduces and the demonstrations still missing. A candidate may remain worth evaluating even when this packet exposes a defect. That is different from approving it for operational reliance.
- Consider paper where document-based review fits the intended tasks and custody, access and historical retrieval can be supported with evidence.
- Consider controlled scans where faithful document retrieval is the intended use and image completeness, indexing, corrections and output boundaries can be demonstrated.
- Consider hybrid records where evidence genuinely spans media and ownership, cross-references and recipient access can be made explicit.
- Consider structured records where reviewers need explicit relationships among observations, revisions, corrections and decisions—and those relationships remain usable in the required outputs.
For PKT-MX-508, a suitable decision note would say: retain the device hold; do not infer release from operations complete; obtain the complete historical external report; finish correction-impact review; and continue evaluating the proposed arrangement against its identified access and handover gaps. The note should distinguish the product disposition from the format-selection decision. Neither closes the other by implication.
Treat historical conversion as a separate assessment. Determine which records would change authority or location, what context could be lost, which retention and retrieval obligations apply, and how discrepancies would be handled. This comparison does not supply a blanket instruction either to convert all historical paper or to leave every historical record unchanged.
Once format choice has an evidence basis, implementation planning belongs in the eDHR implementation guide. Nothing in this comparison calls for a routine parallel run, automatic reuse of validation evidence, a fixed rollout schedule or changes to existing holds. The immediate deliverable is a better-supported decision about record reliance.
Questions
Does FDA require an eDHR application instead of paper?
Do not treat the selected provisions as an application-purchasing mandate. Section 11.1(d) permits qualifying electronic records in place of paper, subject to its conditions and unless paper is specifically required. Current §820.10(a) requires a documented QMS meeting applicable incorporated and supplemental requirements. Establish the required evidence and controls first, then assess the proposed medium. Neither a paper folder nor an eDHR application name establishes compliance.
Is a controlled scan the same as a structured electronic record?
Not for this comparison. A controlled scan preserves a document representation; a structured record represents observations and relationships as identified data. Either may support a useful review, depending on intended use and demonstrated controls. In the example, the scan image preserves the correction while OCR misrepresents it; the structured native view preserves the history while its sample PDF omits it. Evaluate the actual representation and handover, not the category alone.
Does every digital file associated with production fall under Part 11?
Do not classify files solely because they are digital. Section 11.1(b) ties scope to electronic records under FDA record requirements and the specified electronic submissions, and excludes paper records merely transmitted electronically. Map the required record and actual electronic use. A PDF, database entry or attachment needs that assessment; its file extension does not settle it. Assess applicable system and signature controls separately after determining scope.
Does the evidence have to live in one system?
For this evaluation, count unresolved evidence dependencies rather than applications. A single application may still point to an inaccessible external report; several controlled repositories may provide an understandable history when references, authority and access work. PKT-MX-508 needs the exact external report regardless of where its local records live. The comparison does not establish a legal architecture rule: request evidence that the proposed arrangement supports complete, accurate retrieval for its intended use.
Should all historical paper DHRs be converted?
Make a record-specific decision rather than adopting a blanket conversion or no-conversion rule. Identify applicable obligations, the proposed authoritative record, continuing retrieval needs and the information a conversion could omit. A scan may preserve page appearance while an export loses relationships or attachments. Conversely, an inaccessible physical archive also needs attention. Demonstrate the proposed retained package and resolve discrepancies before relying on a changed arrangement; this article supplies no universal conversion mandate or exemption.
What should a reviewer see when an acceptance value is corrected?
In this exercise, the reviewer needs the original 0.25 entry, corrected 0.52 value, Elena Park's explanation and timing, QL-508 as supporting evidence, and the status of correction-impact review. This is a transcription correction, not a retest. For applicable closed electronic systems, §11.10(e) requires the specified audit-trail controls and changes that do not obscure earlier information. A latest-value-only display does not answer the exercise's historical review question.
Will a structured eDHR make review faster?
The supplied evidence does not establish that. Structured relationships may suit certain review questions, but the demonstrated output still has to preserve the required context. Compare representative tasks using the same evidence and roles, including preparation, missing-attachment follow-up and correction review. Record observed effort and unresolved work rather than extrapolating a universal benefit. In PKT-MX-508, no format resolves the inaccessible external report merely by presenting the local record differently.
Which arrangement wins the fictional comparison?
None receives an overall approval from the supplied facts. Paper exposes a custody dependency; the scan download omits a source; the hybrid handover leaves controlling evidence on paper; and the structured PDF omits correction and disposition context. All four lack complete external evidence and completed correction-impact review. The useful result is a set of targeted next demonstrations and an unchanged hold—not a format ranking or permission to replace the current process.
Sources
- 21 CFR Part 820 — Selected §§820.1, 820.7, 820.10, 820.35 and 820.45; View Dated 21 October 2026
- Historical 21 CFR 820.184 — View Dated 1 February 2026; Not Current Law
- 21 CFR Part 11 — Selected Scope, Definitions, System Controls and Signature Provisions; View Dated 21 October 2026
- FDA Computer Software Assurance Guidance — Supplied Scope and Risk Passages Only; Nonbinding Recommendations
Practical orientation, not legal advice. Rules differ by jurisdiction; your quality and compliance team decides what applies.
Electronic Records & Inspection Readiness Buyer Guide
For quality, regulatory, validation and IT teams evaluating electronic records, signatures and inspection readiness. Use it with or without V5.
- See what a signature must show.
- Check corrections keep the original.
- Know where Part 11 applies.
We ask for your name and email so we can handle this request. The full guide then opens on this site, ready to translate, print or save. Privacy policy. Practical orientation, not legal advice; your quality unit decides which rules apply.
UltimateFree guide
Electronic Records & Inspection Readiness Buyer Guide
For quality, regulatory, validation and IT teams evaluating electronic records, signatures and inspection readiness. Use it with or without V5.

