Regulatory Writing Services for Drug, Device and IVD Submissions
Pubrica provides regulatory writing services for pharmaceutical, biotechnology, medical device and IVD sponsors, CROs and regulatory affairs teams. Our regulatory medical writers prepare the clinical, nonclinical, safety, device and lifecycle documents that make up a submission, including CTD Modules 2 to 5, clinical study reports, protocols, investigator’s brochures, PBRERs, DSURs, clinical evaluation reports, technical documentation and health-authority correspondence. Each document is written to the guideline that governs it, and that guideline is named so that you can verify it.
PSURs and DSURs attributed to ICH E2EWrong
ICH as “International Conference on Harmonization”Renamed 2015
ISO 14155:2020 and ISO 20417:2021Both replaced in 2026
Published turnaround bandsAccurate · kept
Every citation named, so you can check itCheckable
Product → stage → region → document set → draft → submission
- Written to a named instrument
- Version checked at the start
- QC by a second reader
- Submission-ready format
- Turnaround band published
Our Services
Regulatory Writing Services by Document Type
Each service can be ordered individually. Most sponsors need three or four of them, and ordering only those costs less than a package built around them.
Who We Serve
Sponsors and Organisations We Support with Regulatory Writing
We work with six types of client, each of whom must meet a fixed filing date.
Scope of Our Regulatory Writing Service
Pubrica writes regulatory documents. We do not provide regulatory strategy consultancy, hold your regulatory intelligence or submit on your behalf; the submission is made by the applicant, through the applicant’s account and under the applicant’s responsibility. We are not a notified body, an auditor or a qualified person, and any sign-off by your QP or review by your notified body remains a separate function. Our role is to turn evidence into documents that state what the evidence supports, in the structure the governing instrument requires, with every claim traceable. Where we have concerns about the strategy behind a submission, we will raise them and then defer to the person responsible for it.
Compliance and Standards
Six Regulatory Frameworks and Their Current Versions
Four of these frameworks changed within the last two years and two changed this year, so we cite the edition as well as the name of every standard.
ICH E6(R3) (good clinical practice)
The Principles and Annex 1 were adopted by ICH and the CHMP on 23 July 2025 and replace E6(R2); the FDA issued its final E6(R3) guidance in September 2025. Annex 2, covering non-traditional interventional designs including pragmatic and decentralised trials, was adopted in June 2026 and applies in the EU from 15 January 2027, with a consolidated guideline published in July 2026. R3 is more proportionate than R2: quality management is built around the risks that actually matter to participants and to the reliability of the results.
ICH M4(R4) and M11 (the CTD, and the new protocol standard)
M4(R4) is the current Common Technical Document organisation, Step 4 on 15 June 2016. M11 — the Clinical electronic Structured Harmonised Protocol — was adopted on 19 November 2025 and published on 5 May 2026, and it is three components rather than one: a guideline, a template and a technical specification. It changes how protocols are structured, and it is worth adopting on the next protocol rather than at the next revision. Note that M13 should never be cited bare: only M13A is adopted.
The ICH E2 series (and which one governs what)
E2A for expedited reporting definitions. E2B(R3) for electronic transmission of individual case safety reports. E2C(R2) for the Periodic Benefit-Risk Evaluation Report, which is what most people still call a PSUR. E2D(R1) for post-approval ICSRs — Step 4 in September 2025, EU-effective 18 March 2026, replacing the 2003 version and adding definitions for digital platforms, social media, patient support programmes and market research. E2E for pharmacovigilance planning, and for nothing else. E2F for the Development Safety Update Report.
EU MDR and IVDR (and the current status of MEDDEV)
Clinical evaluation under MDR Annex XIV, with MDCG 2020-6 on sufficient clinical evidence for legacy devices, 2020-5 on equivalence, 2020-13 on the assessment report template and 2020-1 for device software. Performance evaluation under IVDR Annex XIII. On MEDDEV 2.7/1 rev 4: it was written under the Directives and has not been formally withdrawn, but it is not listed as current MDR guidance — MDCG 2020-6 Appendix I identifies the parts still relevant, and that is the accurate way to cite it. Transition: MDR to 31 December 2027 for Class III and implantable Class IIb and to 31 December 2028 for the rest; IVDR to 2027, 2028 and 2029 by class.
The device ISO standards (two of three changed in March 2026)
ISO 14155:2026 for clinical investigation of medical devices — the fourth edition, published March 2026; ISO 14155:2020 was withdrawn on 23 March 2026. ISO 14971:2019 for risk management — unchanged, and confirmed current in the systematic review closed in March 2025. ISO 20417:2026 for information supplied by the manufacturer — the second edition, published 17 March 2026, replacing the 2021 version.
EU Regulation 536/2014 (transition period completed)
CTIS has been the single submission route since 31 January 2023, and the three-year transition of Directive-authorised trials was completed on 30 January 2025, so there is no longer a transition period. Article 37(4) requires a summary of results within 12 months of the end of the trial in all member states concerned, including a summary written so that a layperson can understand it, per Annex V. Where the trial includes paediatric participants, the deadline is 6 months. The summary is written to the Commission’s Good Lay Summary Practice guidance.
Responsible Use of AI
Responsible Use of AI in Regulatory Writing
Regulatory reviewers trace each decision back to its evidence and to a responsible individual. For this reason, we treat source traceability as more important than drafting speed, and every statement in a regulatory document must be traceable to its source.
Why We Do Not Publish an AI Accuracy Figure
Accuracy figures for AI tools are usually reported by the companies that sell them.
Pubrica therefore does not quote a self-generated accuracy figure, because it would carry the same conflict of interest. Instead, we tell you which tool was used on your project, at which stage, and what a specialist changed afterwards.
How It Works
How a Regulatory Writing Project Works
Every project begins with a guideline version check, because two of the standards this work depends on changed in March 2026. A document written against a withdrawn edition is difficult to detect and costly to correct late in a project.
Scope Definition and Guideline Version Check
Completed before we provide a quote. We confirm the product, the stage, the regions, any previous agency interactions and the document set required. We then check that every guideline, standard and template is current at the start of the project rather than carried over from the last one. ISO 14155 and ISO 20417 were both replaced in March 2026, and E2D(R1) took effect in the EU that month.
Source Material Assembly and Reconciliation
We collect the protocol, statistical outputs, data listings, prior submissions and agency correspondence. These are reconciled against each other before drafting, because a figure given as 88.7% in the statistical output and 89% in an earlier summary is a discrepancy that must be resolved rather than rounded, and resolving it at this stage is quick.
Document Structure Agreed Against the Governing Guideline
We map the document to the sections the guideline requires, with the content source for each. E3 governs a clinical study report, M11 a protocol, and MDR Annex XIV with the MDCG guidance a clinical evaluation. The structure is agreed before drafting, because a regulatory document restructured afterwards loses its cross-references and numbering, both of which reviewers check.
Drafting by an Experienced Regulatory Writer
Each writer is matched to the document type and the therapeutic or device area. CER writing and CSR writing require different specialists. Every claim is traced to its evidence, every number to its source output, and every standard is cited with its current version and year.
Quality Control Across the Full Document Set
A reviewer who did not write the document reads it against the other documents in the submission. The review covers claim tracing forward and back, numerical consistency including rounding, references read against the statements they support, and completeness against the guideline’s own section list. Most assessor findings concern disagreements between documents rather than errors within one.
Formatting, Delivery and Support with Agency Questions
Documents are delivered submission-ready in your publishing format, with the source files. If the agency raises questions after delivery, the writer who prepared the document responds to them, and the rationale for every drafting decision is recorded.
Request a Free Submission Requirements Assessment
The assessment takes two business days and is free of charge. It identifies which documents your submission needs, the instrument that governs each one, a realistic timeline against your filing date and our view on whether that date is achievable. If a document you already hold cites an out-of-date edition, we will also tell you.
Agency correspondence is the most useful item on this list and the one most often withheld. A deficiency letter, meeting minutes or a question from a previous cycle shows what the reviewer focuses on for this product. If a previous submission received significant questions, that correspondence is the document we most want to see.
Submission Document Requirements
Regulatory Submission Document Set Builder
Choose your product, the stage you have reached and the region you are filing in. The tool below lists the documents you will need, names the regulation or guideline behind each one, and gives a realistic turnaround for it. It also tells you where your agency currently stands on electronic submission, which varies more between regions than most summaries suggest.
01
The documents, the guideline governing each, the turnaround band and your region’s eCTD position.
Runs in your browser · nothing sent anywhereThe Principles and Annex 1 were adopted by ICH and the CHMP on 23 July 2025 and replace E6(R2). Annex 2, covering pragmatic and decentralised designs, applies in the EU from 15 January 2027.
M11 was adopted on 19 November 2025 and published on 5 May 2026. It is a guideline, a template and a technical specification rather than a single document, and it changes how protocols are structured — worth adopting now rather than at the next revision.
Written to the committee’s own format and at the reading level they specify. The investigator holds responsibility for the consent process; we draft the document.
Module 1 is regional and is the part that differs most between agencies. Modules 2 to 5 are common.
Module 2.6, which is often written last and read first by reviewers.
Turnaround: 7–10 business days. Short regulatory summaries and single documents. These are the bands we work to and they have not changed — short regulatory summaries 7 to 10 business days, full clinical study reports 15 to 25, complex dossiers and multi-module CTDs 25 or more. Expedited delivery is available where capacity allows and is quoted on capacity rather than on urgency.
Submission format in your region
Sources and a correction. ICH is the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use — renamed from “Conference” on 23 October 2015. Guidelines published before then carry the old name on their face and should be cited as printed. Governing instruments above are as published at database.ich.org, on the European Commission’s MDCG index and at iso.org. The correction: “compliance with GVP and ICH E2E guidelines” is sometimes attached to packages containing PSURs and DSURs. E2E covers pharmacovigilance planning; PBRERs are governed by E2C(R2) and DSURs by E2F. This is a planning aid, not regulatory advice — your regulatory strategy, your agency interactions and your submission decisions are yours, and requirements differ by region and by product.
Why We Name the Governing Guideline for Every Document
In regulatory writing, the citation is the evidence of expertise. Reviewers, auditors and notified body assessors check whether a vendor knows which guideline governs each document, and a single wrong attribution is noticed. A common error is attaching “compliance with GVP and ICH E2E guidelines” to packages containing PSURs and DSURs. E2E covers pharmacovigilance planning, including the safety specification and the pharmacovigilance plan, and governs no periodic report: PBRERs are governed by E2C(R2) and DSURs by E2F.
Our Experts
Our Regulatory Writing Team
A named team of regulatory specialists stays with your project from scoping to delivery. Your project coordinator introduces each team member by name before work begins.
Writes the document against the governing template and regulatory guidance, keeps every statement traceable to its source data, and maintains the version history and comment log through each review round.
Reviews the clinical and scientific content for accuracy and confirms that each conclusion is supported by the data. Each expert is matched to your therapeutic or device area.
Checks that statistical methods, tables and conclusions are consistent with the analysis plan, and reviews how the data are presented before the document reaches regulatory review.
Performs an independent check of the final document against source data, the template and the applicable regulatory checklist. Findings are logged and resolved before delivery.
Sample Work
Sample Quality Control Report for a Regulatory Document Set
Most assessor findings are not errors within one document but disagreements between two documents, so our quality control reads the whole set against itself. Use the tabs to view each check, then open the full report below.
Document set QC report
Class IIb device · 4 documents · 34 claims traced
Version currency · claim tracing · numbers · references
4 version findings in 4 documentsThree of them from templates carried forward
- Claims traced
- 34 performance and safety claims, each traced back to the clinical evidence and forward to every other document repeating it.
- Unsupported by the cited evidence
- 3. In each case the cited study supports a narrower claim than the document makes — a different population in two, a different endpoint in the third.
- Present in the IFU, absent from the CER
- 2. A claim a patient will read that the clinical evaluation does not establish is the specific thing an assessor looks for, and it is the commonest way this set fails.
- Stated differently across documents
- 4. Same claim, four wordings, three of which are broader than the evidence. Reconciled to the narrowest defensible version and used consistently.
- Equivalence argument
- Reworked to MDCG 2020-5. The original asserted equivalence on clinical and technical characteristics and did not address biological characteristics at all, which is one of the three required and the one most often skipped.
- Kept unchanged
- 25 of 34. The evidence supports them as written, and saying so is part of the report.
4 numerical findingsNone of them an arithmetic error
- References checked
- 84, each opened and read against the statement it supports rather than matched to the citation.
- Do not support the statement
- 5. Four were inherited from an earlier CER’s citation rather than read at first hand; one supports the opposite conclusion in the population at issue.
- Retracted or subject to an expression of concern
- 1. Retracted fourteen months before the draft. Removed, and the claim it supported re-evidenced from two other sources.
- Search currency
- The literature search re-run to the delivery date. Six new papers, two of which changed the state of the art section and one of which weakened a claim — which is reported rather than left out.
- Search reproducibility
- Every strategy saved with its database, interface, date and result count, so the notified body can re-run it and the annual update starts from a strategy rather than from scratch.
- Verified and retained
- 78 of 84.
Redacted extract · device, class and client changedQC report, as delivered
This report reviews four documents as one set.Use the tabs to see each check.
The versions tab shows the most common source of error. Three of the four version findings came from templates carried forward from a previous submission. This is standard practice, which is why two standards changing in the same month creates risk. ISO 14155 and ISO 20417 were both replaced in March 2026, and only a deliberate version check identifies such changes.
The claims tab shows the most serious finding. Two claims in the instructions for use are not established by the clinical evaluation. Patients read the first document, and assessors read both.
On request, we prepare a sample QC report based on a document set similar to yours, with the same product type, class and submission route.
Regulatory writing · sample QC report
Document set QC report — Class IIb device, MDR submission
A QC report from a real project with the device, the manufacturer, the indication and every figure changed. The checks, the categories of finding and the reporting format are as delivered. On request, we will prepare one based on a document set like yours.
1What was checked, and how
Four documents from one submission, read as a single set rather than one at a time. Most of what a notified body finds is not an error inside a document; it is a disagreement between two of them, and no single-document review can see it.
| Element | How it was handled |
|---|---|
| Documents | Clinical evaluation report, instructions for use, labelling text, and the post-market clinical follow-up plan. |
| Reviewer | Independent of the writing team. A writer checking their own set re-reads what they meant rather than what is on the page. |
| Version currency | Every standard, guidance document and template cited, checked against the issuing body’s current publication as at the review date. |
| Claim tracing | Every performance and safety claim traced back to the clinical evidence and forward to every other document repeating it. 34 claims. |
| Numerical consistency | Every figure compared across documents including its rounding. A value given as 88.7% in one and 89% in another is a claim discrepancy, not a typographical one. |
| References | Each cited work opened and read against the statement it supports, rather than matched against the citation. 84 references. |
| Search currency | The literature search re-run to the delivery date and anything new appraised. |
What this is not: a regulatory review. We check that the documents say the same thing, that each claim has evidence behind it, and that each is written to the standard that governs it. Whether the regulatory strategy is right belongs to your qualified person.
2Version findings
| Document | As cited | Current | Resolution |
|---|---|---|---|
| Clinical investigation plan | ISO 14155:2020 | ISO 14155:2026, 4th edition, March 2026 | The 2020 edition was withdrawn on 23 March 2026. Updated, with the change flagged to the notified body contact rather than made silently. |
| Labelling file | ISO 20417:2021 | ISO 20417:2026, 2nd edition, 17 March 2026 | Updated. Same month, same cause as above. |
| Clinical evaluation report | MEDDEV 2.7/1 rev 4 throughout | MDR Annex XIV with MDCG 2020-6, 2020-5, 2020-13 | MEDDEV rev 4 has not been formally withdrawn but was written under the Directives and is not listed as current MDR guidance. Restructured to the MDR framework, retaining the sections MDCG 2020-6 Appendix I identifies as still relevant. |
| All four | No MDCG 2020-1 reference | MDCG 2020-1 | The device contains software with a clinical function. Clinical evaluation of that software falls under MDCG 2020-1, which was not cited anywhere. Added. |
| Risk management file | ISO 14971:2019 | ISO 14971:2019 | Correct. Confirmed current in the systematic review closed March 2025. The one standard in this set that did not change. |
Three of the four findings came from templates carried forward from the manufacturer’s previous submission. This is standard practice in regulatory teams, which is why two standards changing in the same month creates risk. A normal drafting process does not detect this; only a deliberate version check at project start does, which is why it is stage one rather than part of QC.
3Claim tracing — 34 claims
| Finding | n | Detail |
|---|---|---|
| Supported as written | 25 | Retained unchanged. Saying so is part of the report; a QC report that only lists problems is not a QC report. |
| Narrower than claimed | 3 | The cited study supports a narrower claim than the document makes — a different population in two cases, a different endpoint in the third. Rewritten to what the evidence establishes. |
| In the IFU, not in the CER | 2 | The most serious finding. A claim a patient reads that the clinical evaluation does not establish. Either evidence it or remove it; there is no third option, and an assessor reads both documents. |
| Stated inconsistently across documents | 4 | Same claim, four wordings, three broader than the evidence. Reconciled to the narrowest defensible version and used consistently. |
| Equivalence argument | 1 | Reworked to MDCG 2020-5. The original addressed clinical and technical characteristics and not biological characteristics — one of the three required, and the one most often skipped. |
4Numerical consistency
| Finding | Why it matters | Resolution |
|---|---|---|
| 88.7% in the CER, 89% in the IFU | A performance figure stated to different precision in two documents is a discrepancy an assessor will query, and “we rounded” costs a cycle to answer. | One figure, one precision, used everywhere. |
| n=412 in the summary, n=409 in the analysis table | Three subjects excluded post-hoc for a protocol deviation. Legitimate — but it reads as an inconsistency unless stated. | Stated with the reason, in both documents. |
| “Up to 24 months” and “24 months” | Median follow-up was 19 months. The unqualified version overstates it, and it appeared in the document the patient reads. | Qualified consistently, with the median given. |
| Confidence intervals missing from two claims | Point estimates alone overstate certainty, and the intervals existed in the source output. | Added from source. |
| All remaining figures | — | Traced to a named source output, so a query three months later is answered in minutes rather than reconstructed. |
5Reference verification — 84 references
| Finding | n | Detail |
|---|---|---|
| Verified and retained | 78 | Opened, read, and confirmed to support the statement attached to them. |
| Do not support the statement | 5 | Four inherited from an earlier CER’s citation rather than read at first hand; one supports the opposite conclusion in the population at issue. This is how an error propagates through successive versions of a clinical evaluation for years. |
| Retracted | 1 | Retracted fourteen months before the draft was written. Removed; the claim it supported was re-evidenced from two other sources. |
| Search re-run to delivery date | +6 | Six new papers. Two changed the state-of-the-art section; one weakened a claim, which is reported rather than omitted. |
| Search reproducibility | — | Every strategy saved with its database, interface, date and result count, so the notified body can re-run it and next year’s update starts from a strategy rather than from scratch. |
6Deliverables provided
| Item | Why |
|---|---|
| The four documents, revised, in source format | Editable, so your team owns them and the next update is an amendment rather than a re-commission. |
| The QC report itself | Findings by severity, with the resolution recorded for each. This is the document that answers “what did you check?” at audit. |
| The claim trace matrix | Every claim, its evidence, and every document it appears in. The artefact that makes the next annual update tractable. |
| The search log | Strategies, dates, interfaces and counts. Re-runnable, because a clinical evaluation is updated rather than finished. |
| The version register | Every standard and guidance document cited, with the edition and date checked at project start and again at delivery. |
What this report does not do is guarantee an outcome. A notified body may take a different view of an equivalence argument, and an assessor may ask for evidence nobody anticipated. What a QC report of this kind does is remove the findings that are avoidable, so that the questions you receive are about the science rather than about whether your documents agree with each other.
Prepared by Pubrica. Pubrica is a trading name of Guires®, registered in England, Wales and India.
Request this samplePackages
Regulatory Writing Service Packages
Our regulatory writing prices reflect the length of regulatory documents, the experience of writers who have prepared the same document type before, and the scrutiny of the full-time assessors who review them.
Single Regulatory Document for One Region
A protocol, an investigator’s brochure, a DSUR, a PBRER, a summary, an IFU, or a response to an agency question
From $2,400up to 30 pages · 1 region · 2 revision rounds
Delivery 7–10 business days
- ✓Written to the governing instrument, with the version confirmed at project start
- ✓Every claim traced to its evidence and every number to its source output
- ✓QC by a second reader who did not write it
- ✓Delivered in your publishing format, with source files
- ✓The writer stays available for agency questions afterwards
Optional additions: further regions and QC of documents written elsewhere.
View a sample Select DocumentClinical Study Report or Clinical Evaluation
Sponsors and CROs reporting a study, and device manufacturers whose CER or PER is due
From $9,500up to 120 pages · appendices · literature search included
Delivery 15–25 business days
- ✓Everything in the Document package, at study scale
- ✓CSR to ICH E3, or CER to MDR Annex XIV with the MDCG guidance applied
- ✓Patient narratives written from source data rather than summarised from tables
- ✓A documented, re-runnable literature search where the document needs one
- ✓The claim trace matrix and the search log handed over, not just the report
- ✓Cross-document QC against the other documents in the set
Optional additions: the rest of the submission and annual CER updates.
View a sample Select StudyRegulatory Dossier or Device Technical File
Marketing applications, multi-module CTDs, and full technical documentation for CE marking
From $18,000Modules 2–5 summaries, or a full technical file
Delivery 25+ business days
- ✓Everything in the Study package, across the document set
- ✓CTD Modules 2 to 5 to ICH M4(R4), or MDR Annexes II and III
- ✓The Clinical Overview written as an argument rather than a summary
- ✓Regional Module 1 handled as a separate deliverable per agency
- ✓Set-wide QC: claim tracing, numerical consistency, reference verification
- ✓A version register covering every standard and guidance document cited
Multi-region submissions and health-authority meeting packages are quoted separately.
View a sample Select SubmissionAdditions are quoted alongside any tier: QC of documents written elsewhere, which is frequently the better first purchase and always the cheapest way to find a version problem; annual clinical evaluation updates, which under the MDR are a recurring obligation rather than a project; health-authority meeting packages and briefing books; lay summaries under CTR Article 37(4), measured against a reading grade; translation and multi-language labelling; eCTD publishing and submission-readiness checks; and expedited delivery where capacity allows, quoted on capacity rather than on urgency.
Quality Assurance
Our Service Commitments and Limitations
Approval timelines are set by the agency, so we do not make promises about them. Our commitments and their limits are set out below.
Our Commitment
Revision Support Until Agreed Requirements Are Met
We revise the document until it meets the brief and the instrument it is written to, with no cap on rounds within the agreed scope, and the writer remains available for agency questions after delivery.
01
Version checks at start and delivery
At project start and again at delivery. Two of the standards this work depends on were replaced in March 2026, and a template carried forward is how that gets missed.
02
Agency questions answered by the writer
When the agency raises questions, the writer who prepared the document responds, and every drafting decision is recorded.
03
Working files provided
You receive the claim trace matrix, the search log and the version register as well as the document, which simplifies next year’s update.
04
Correction of our errors at no cost
If a wrong version citation, an untraced claim or an inconsistent number across documents originated with us, we correct it free of charge within the agreed scope and inform you first.
Limits of our service. We cannot accelerate an approval timeline: review clocks are set by the agency, and a writer affects only the submission-preparation timeline. We do not guarantee that a submission will be accepted, that a notified body will agree with an equivalence argument, or that an audit will close without findings, and we publish a testimonial referring to such an outcome only when the client has confirmed it in writing and agreed to be named. We are not your qualified person, notified body or regulatory strategist, and we do not submit on your behalf; the submission is made by the applicant, under the applicant’s responsibility. We will not write a claim the evidence does not support. Where a claim in your instructions for use is not established by your clinical evaluation, we will tell you and then either remove it or add the supporting evidence. Where a document you already hold cites a withdrawn standard, we will tell you, even if you did not ask and the document was not written by us.
Free Guideline Version Check
Send us a document you already hold, and we will tell you which of the standards and guidance documents it cites are no longer current, free of charge and with no obligation. Outdated citations are the most common defect we find, usually because templates are carried forward while standards change. Two changed in March 2026 alone.
Confidentiality and Conflicts of Interest
A non-disclosure agreement is signed before any material moves, and writers and project staff are bound by internal confidentiality agreements as well. Unpublished clinical data, submission drafts and agency correspondence are held under the UK GDPR on access-controlled systems within a quality management system certified to ISO 9001:2015, kept only as long as the project needs them, never used to train anything and never shared outside your project team. We do not take competing products in the same indication concurrently, and we will tell you if a conflict exists rather than managing it quietly.
Working Within Your Processes
We work to your templates, SOPs, QC steps, document management system and naming conventions. Imposing a different process on a sponsor with a validated one would create a deviation, and CROs need overflow capacity to work to the same standard as their own team. Where you have no process, we will propose and document one; where you have one, we follow it.
Authorship and Writing Assistance Disclosure
Where a regulatory document will also become a publication, the ICMJE authorship criteria apply and writing assistance is acknowledged with the sponsor’s role in the writing disclosed. Where the research is company-sponsored, GPP 2022 applies alongside — published in the Annals of Internal Medicine in 2022 and stewarded by ISMPP, superseding GPP3 from 2015. Any use of AI is declared to you and handled to the ICMJE’s January 2026 guidance.
Client Stories
Client Feedback on Regulatory Writing Projects
These quotations are paraphrased from client feedback. None of them claims a regulatory outcome, because that decision belongs to the agency.
“They flagged that two of the ISO standards our template cited had been replaced that March. Nobody in our team had noticed, and we had three submissions running on it.”
“Two claims in our instructions for use were not established anywhere in the clinical evaluation. That is exactly what we would have been asked about.”
“The writer who drafted the CSR answered the agency’s questions four months later. They still had the reasoning written down.”
FAQs
Regulatory Writing FAQs
If you are unsure which documents your submission requires, or whether a document you hold cites a current standard, Ask an expert — no quote is required.
Clinical and nonclinical: protocols to ICH M11 and E6(R3), investigator’s brochures, informed consent documents, clinical study reports to E3, nonclinical summaries, patient narratives and case listings. Dossier: CTD Modules 2 to 5 to M4(R4), with Module 1 handled per region. Applications and correspondence: INDs, NDAs, BLAs, MAAs, IMPDs, briefing books, meeting packages, responses to agency questions, rebuttal letters. Safety: PBRERs to E2C(R2), DSURs to E2F, risk management plans to EU GVP Module V, safety specifications and pharmacovigilance plans to E2E, expedited reports to E2A, post-approval case handling to E2D(R1). Devices and IVDs: clinical evaluation reports under MDR Annex XIV, performance evaluation reports under IVDR Annex XIII, technical documentation, clinical investigation plans to ISO 14155:2026, instructions for use and labelling to ISO 20417:2026, post-market surveillance and PMCF. Software: clinical evaluation to MDCG 2020-1 and lifecycle documentation to IEC 62304. Lifecycle: SmPCs, PILs, variations, renewals. And lay summaries under CTR Article 37(4).
Short regulatory summaries and single documents: 7 to 10 business days. Full clinical study reports: 15 to 25 business days. Complex dossiers and multi-module CTDs: 25 or more business days. These turnaround bands have not changed since we first published them. Two factors affect them more than document length. The first is the completeness of the source material, since a CSR cannot be finalised while it waits on final statistical outputs. The second is your review cycles, which usually take the longest; a document drafted in fifteen days and reviewed over six weeks takes eight weeks in total. Expedited delivery is available where capacity allows and is quoted on capacity rather than on urgency. Where a date is not achievable, we will tell you before accepting the work.
ICH E2C(R2), which defines the Periodic Benefit-Risk Evaluation Report — the PBRER, which is what most people still mean when they say PSUR. E2C(R2) reached Step 4 on 15 November 2012 and extended the report’s scope from safety to benefit-risk. DSURs are governed by ICH E2F, Step 4 on 17 August 2010. E2E is Pharmacovigilance Planning — it covers the safety specification and the pharmacovigilance plan, principally in preparation for the early post-marketing period, and it governs no periodic report. We explain this in detail because the error is common: “compliance with GVP and ICH E2E guidelines” is often attached to packages containing PSURs and DSURs.
No. ICH E6(R3) is current: the Principles and Annex 1 were adopted by ICH and by the CHMP on 23 July 2025 and are described as a replacement for E6(R2), and the FDA issued its final E6(R3) guidance in September 2025. Annex 2, which covers non-traditional interventional designs including pragmatic and decentralised trials, was adopted in June 2026 and applies in the EU from 15 January 2027; a consolidated guideline covering all three parts was published in July 2026. A protocol or SOP written in 2026 that still cites E6(R2) as the current standard is a year out of date. The substantive change is proportionality: R3 asks for quality management built around the risks that actually matter to participant safety and to the reliability of the results, rather than uniform procedure applied regardless of risk — which is generally good news for investigator-initiated and lower-risk studies.
Yes, with care, but not as the governing framework. MEDDEV 2.7/1 rev 4 was issued in 2016 under the Directives, before the MDR applied. It has not been formally withdrawn, and industry does still use it — but it is not listed as current MDR guidance, and MDCG 2020-6 Appendix I identifies the sections of it that remain relevant under the MDR. That is the accurate position and the defensible way to cite it: structure the clinical evaluation to MDR Annex XIV with the MDCG series applied, and reference MEDDEV rev 4 where MDCG 2020-6 says it still holds. The MDCG documents that matter most for a CER are 2020-6 on sufficient clinical evidence for legacy devices, 2020-5 on equivalence, 2020-13 on the assessment report template, and 2020-1 where the device contains software — which is the one most often missing from device CERs that happen to include software.
Not yet. eCTD v4.0 is not mandatory in any region today. The FDA has accepted v4.0 for new applications since 16 September 2024, forward compatibility for existing v3.2.2 applications is not yet available, and no date has been announced on which v4.0 becomes the only accepted version. The EMA is the only agency with a published timetable: optional for new centrally authorised MAAs since 22 December 2025, strongly recommended from Q1 2027, and mandatory for new CAP MAAs from Q1 2028. Health Canada has not implemented it and its Canadian Module 1 guide for v4 remains in draft. The MHRA and TGA continue in v3.2.2. On Japan: PMDA accepts v4.0 and its Japanese implementation guide is at v1.6.0 dated 10 March 2025, but PMDA has published no mandatory transition deadline — a claim of a 2026 Japanese mandate circulates widely in consultancy material and we could not verify it from PMDA’s own publications, so we do not repeat it.
A non-disclosure agreement is signed before any material moves, and writers and project staff are bound by internal confidentiality agreements as well. Unpublished clinical data, submission drafts and agency correspondence are held under the UK GDPR on access-controlled systems within a quality management system certified to ISO 9001:2015 — certificate H3CR9TCU20251212INDQM10N — kept only as long as the project needs them, never used to train anything, and never shared outside your project team. On conflicts, which matter more in this category than in most: we do not take competing products in the same indication concurrently, and where a potential conflict exists we tell you before accepting the work rather than managing it quietly. If you require a writer to work only within your own document management system rather than ours, we do that; several of our clients require it and it is a reasonable thing to require.
Yes. A full sample is available on this page: a document set QC report for a Class IIb device submission, with the device, the manufacturer and every figure changed. You can read it in full without contacting us. If you would like a sample closer to your own work, we can prepare one based on a similar document type, with the same product class, submission route and governing instruments. Please note two limitations. Client work is confidential, so every sample is redacted or composite. A sample also shows structure, citation practice and QC method, but it cannot show whether the writer understands your product, which is the purpose of the free scoping discussion.
No. Approval timelines are set by the agency. Review clocks are determined by regulation and by the agency’s own procedures, and no writer can shorten them. What a writing partner can affect is the submission-preparation timeline and the number of questions raised after submission. When the documents agree with each other, every claim is evidenced and every standard is current, agency questions are more likely to concern the science than the documentation. A deficiency letter can add months, so this is a real benefit, but we do not promise an outcome we do not control.
We revise it until it meets the brief and the instrument it is written to, with no cap on revision rounds within the agreed scope. Two practices keep this manageable. First, the structure is agreed against the governing guideline before drafting, so a revision does not require a restart; most dissatisfaction with regulatory documents traces back to a structure that was never agreed. Second, the scope is documented: a different indication, a second region’s Module 1 or a document type not in the brief is new work, and we confirm this before starting. Where an error is ours, such as a wrong version cited, an untraced claim or a number inconsistent across documents, we correct it free of charge within the agreed scope and inform you first.
Related Services
Related Research and Publication Services
Regulatory documents draw on evidence gathered earlier in development and are often followed by publications.
1EVIDENCE SYNTHESISsearches and synthesis
Clinical literature reviewTargeted and systematic review, including for CERs and HTA dossiers. Literature search and citationAuditable, re-runnable searches with every strategy saved.2SUPPORTING SERVICESdesign, numbers and editing
Biostatistics and programmingAnalysis plans and outputs, with the estimands framework addressed. Editorial and quality supportIndependent QC of documents written elsewhere.3PUBLICATION AND COMMUNICATIONafter submission
Medical writingThe broader writing programme the submission sits inside. Scientific medical communicationWhat may be said about an approved product, and under which code.Request a Free Guideline Version Check of Your Document
We will tell you, within two business days and free of charge, which of the standards and guidance documents your document cites are no longer current. There is nothing to sign. Outdated guideline citations are a frequent finding in regulatory documents.
Guides and articles in this section
Show all 26
- 510(k) and Technical Documentation Writing
- Clinical Evaluation Plan (CEP) Writing Services
- Clinical Evaluation Report Writing
- Clinical Study Protocol Writing
- Clinical Study Report Writing
- CTD Module 2.5 and 2.7 Writing
- CTD Structure: Overview of the Five Modules
- Device Risk Management File | ISO 14971:2019+A1
- GSPR Checklist and Technical Documentation
- Health Authority Query Response Writing
- Informed Consent Form Writing
- Investigator's Brochure Writing
- Nonclinical Study Report Writing
- Patient Safety Narrative Writing
- Performance Evaluation Plan (PEP) Writing
- Performance Evaluation Report Writing
- Plain Language Summary Writing
- PMCF Plan, Survey and Report Writing
- Post-Market Surveillance and PSUR Writing
- PSUR, PBRER and DSUR Writing
- Risk Management Plan Writing
- Scientific Validity Report Writing for IVDR
- SmPC, PIL and USPI Labelling Writing
- SOP Writing Services
- State of the Art Review for Medical Device CER
- Summary of Safety and Clinical Performance Writing