Responding to a health authority's questions in review
A health authority raises questions while it reviews a dossier, and the sponsor answers them in writing against a deadline. The work covers what those questions are and what answering them achieves, which kind of question has arrived, when each agency asks and how long it allows, who answers and what can be prepared before the questions come, how the response documents are set up, how the response cycle is driven to its deadline, how each question is read, answered and pitched, and what a medical writer adds.
What the agency's questions are and what answering them achieves
In every country, reviewers working through a dossier will have questions. The questions go by different names in different regions (request for information, request for clarification, list of questions, list of outstanding issues), but the principle is the same everywhere: every region raises questions during dossier review.
These responses are part of the dialogue between the sponsor and the agency. They clarify data or address weaknesses in the drug's development programme, and they are the last step before the agency can decide on the drug's benefit-risk ratio. They are also an opportunity to build trust, to show responsiveness to the authority, and to demonstrate scientific integrity. The response-to-questions document is a central part of the post-submission interaction with the health authorities.
Submitting the application, or a variation or supplement for a drug already approved, is an important company milestone, but plenty of work remains after it. After the agency validates the submission it reviews the documentation, and a back-and-forth begins in which positions are negotiated and concessions may be made. The details of this post-submission interaction vary with the type of application and the agency, but the general approach is the same: a list of questions is issued, sometimes called a request for supplementary information, and the company prepares its responses and submits them to the agency for review. More than one round of questions may be needed before the process ends, at which point the agency either approves the application, usually with certain conditions, or rejects it, or the company withdraws it. The process can be considered analogous to peer review, in which authors prepare responses to the reviewers' comments.
Kinds of question
Questions arrive under clinical headings covering efficacy, safety, and clinical pharmacology; under chemistry, manufacturing, and control (CMC), covering drug substance and drug product manufacturing; under non-clinical; and about labelling or the risk management plan (RMP). CMC questions are often handled separately by the CMC teams.
Questions can concern any aspect of the submission, and they range from fairly simple ones, such as a request to provide a certificate of analysis, to complex ones, such as a fundamental challenge to some aspect of the interpretation of the results.
Questions also differ in weight: a major question, or major objection, can block approval if it is not resolved, and it affects the benefit-risk conclusion. Major objections are often anticipated but are tricky to answer, because they require many resources and meetings. A major objection is sometimes resolved by a conditional approval, with the applicant asked to generate additional data to convert it to a full approval, and sometimes the agency limits the indication to a subgroup with a better benefit-risk profile. A minor question is a clarification that does not necessarily affect approval. For instance, where module 3 states that the drug can be stored at 18 degrees for 24 months and the label states 36 months, the agency will ask for the discrepancy to be clarified, and the clarification will not affect approvability.
Nonclinical submissions do not elicit a straight yes or no from the health authorities, given their complexity. Health authorities frequently have questions about the submission data and their interpretation, which must be answered within a limited timeframe — anywhere from one day to several months, depending on the country, the submission, and the number and type of questions. These may include technical queries such as a missing report signature, proposed alternative interpretations of findings, and requests for additional data to support a particular claim. Answering them often requires delving deeper into the original study data or the scientific literature, and sometimes even planning a new study.
When each agency asks and how long it allows
Under FDA review of a new drug application or biologics license application, questions can arrive at any point in the review, and FDA is less bound by a pre-specified schedule. FDA sometimes allows two days for a reply, sometimes weeks and sometimes a month, depending on the question's complexity. For FDA there is no predictability, and the team needs to stay responsive and proactive throughout the review, because FDA questions often have short turnarounds.
FDA often stipulates a deadline when it sends questions; where it does not, teams typically still answer as soon as possible, which is often within about 14 days. The number of questions in one set varies, they are often grouped by topic — a request for information carrying the statistical questions, for instance — and several requests can arrive in quick succession, depending on where the different reviewers are, so the team always has to be prepared and standing by to work on them.
EMA has well-defined timelines, including so-called clock stops. Its procedure has predictable timelines and checkpoints, according to which resources can be assigned to the responses. The marketing authorisation application is submitted on day 0. The first set of questions comes at day 120 and the applicant has two months to answer, during which the review clock stops and then restarts once the responses are submitted. Any outstanding issues are issued as a list of outstanding issues at day 180, with one month to respond and a second clock stop. EMA usually provides all its questions at the end of the review procedure as part of an assessment report.
Before the official questions, the rapporteur's assessment report is expected at day 80, written to EMA's own templates. These are fairly long documents carrying draft questions that the response team can already begin working from. They give only a short glimpse, and further questions may come later, but they are a very good point at which to start. The dates are defined but not guaranteed: a day 80 report has arrived two weeks late while a whole team stood by waiting for it. If the reports are a few days late, regulatory may politely ask EMA where things stand and when they can be expected, but otherwise there is little to do but wait. The co-rapporteur's assessment follows about two weeks later, adding further assessment comments based on the rapporteur's report and possibly further questions. Preliminary assessment reports may also be sent to the company, and they can give some indication of the thinking of the agency reviewers.
The rapporteur's and co-rapporteur's reports carry more than a proposed list of questions: they include a critical assessment of the submission with discussion and conclusions, so EMA questions arrive with context that sometimes shows where they came from. Quality, non-clinical, and clinical material typically comes as separate reports. Within a report, a factual summary of the submitted data is followed by the rapporteur's and co-rapporteur's comments at the end of each subsection, with an overall assessment below the larger sections such as pharmacokinetics, efficacy, and safety. The preliminary questions themselves sit at the end of the subsections, and although it is not part of the official EMA template, a separate section collecting the list of questions is often present as well.
The day 120 list of questions is agreed at a CHMP meeting, where other EU countries can add new questions or change the rapporteurs' and co-rapporteurs' proposed ones before one set is agreed. That document too provides context — an executive summary, a scientific overview, and discussion — with the questions to be addressed in writing numbered in the subsections of its section seven.
After the responses are submitted, EMA's assessment of them comes back as a day 150 joint assessment report, which states a rationale and conclusion for each and indicates whether the issue is considered resolved, not resolved, or may have triggered a new question. Based on that outcome, a day 180 list of outstanding issues is issued, typically with one month to respond using the same response template, and EMA comments on those responses in a day 195 joint assessment report. Issues still unresolved at that point will likely bring a further clock stop and possibly another list of outstanding issues requiring more written responses, and optionally an invitation to an oral explanation at a CHMP meeting.
Most other authorities either follow one of these two shapes or sit somewhere between them: Health Canada follows FDA's approach, while the MHRA and Swissmedic follow EMA's. Timelines differ, but generally an agency either raises questions at any time during the review as they come up or takes a more structured approach, which need not have EMA's two clock stops but often has more defined timelines.
The response team
Answering a health authority's questions is a collaborative task, not one person's job: it is a team effort within the response team, with the subject-matter experts and every function involved, and with the health authority itself. The large team that does it is sometimes called the rapid response team. Regulatory affairs leads coordination and submission of the responses. Medical writing drafts and formats the responses and ensures clarity and consistency of content. Medical writers also help regulatory affairs with coordination, updating the tracker and giving the team status updates. It is helpful to have a medical writer as coordinator, pulling in the input from the contributing functions and maintaining the tracking sheet, and on larger projects with several writers assigned to subsets of responses, it is important that one writer has oversight across all the activities and coordinates them across all the areas. Subject-matter experts from the clinical, CMC, clinical pharmacology, and non-clinical teams provide input during drafting and again during review. The statistics team are the experts on statistical questions and prepare new tables, figures, and listings as required, generating with the programming team the additional analyses and outputs that some questions need. Quality control checks the responses for accuracy of data and compliance with regulations. Translation covers countries that do not accept an English submission, and publishing combines the response document, outputs, appendices, and literature into the single PDF that is submitted.
Before the request for supplementary information arrives, it may be helpful to put together a response team whose members are able to dedicate sufficient time to the responses. A standardised approach to getting ready centres on being proactively prepared with a team able to produce focused responses rapidly — clinical operations, biostatistics, clinical pharmacology, preclinical operations, chemistry, manufacturing and controls, regulatory affairs, medical writing and document management — each member with an assigned role so that it is clear who does what once the questions come in.
Preparation before the questions arrive
Start preparing as early as possible. Some questions will be a surprise, but others can be anticipated, and the homework for those should already be done when they arrive. Although timelines differ significantly between agencies, a standardised approach to getting ready for and writing responses can make the process more efficient.
Building a list of potential questions and issues during CTD preparation is common practice in regulatory affairs, and anticipating the questions lets the teams prepare clear and robust drafts in advance. The idea is to have draft responses ready, or at least the responsible teams informed of the potential issues, which saves a lot of time later. While preparing the CTD, identify gaps in the data and try to fill them proactively, if possible, before they become major objections, finding the weaknesses of the clinical development programme and discussing them with the team. For instance, a pivotal study that met its primary endpoint with efficacy results that were not robust prompts a discussion with the team of whether additional sensitivity analyses are needed and which would be most meaningful. And where a toxicity risk is mitigated by dose reduction or temporary treatment interruption, the agency can ask whether efficacy was maintained in those participants, and the team discusses this and plans a subgroup analysis to collect more data.
The company may also have already made a critical assessment of its application, identified the weak areas where questions are likely to be asked, and decided on a high-level strategy for each should those issues be raised. The team first runs a self-critical gap analysis of the challenges the authorities may raise, such as weaknesses of the drug or its class or deficiencies in the development programme or data analysis, and plans the deliverables likely to be needed for each.
Once the dossier is submitted, hold a kickoff with the response team while waiting for the questions. Assign roles and responsibilities, review the list of potential questions, and decide the action items it implies — whether outputs can be produced at risk, for instance. Regulatory gives an overview of when questions are expected, and the team agrees the procedures for starting the work: who schedules the kickoff when questions arrive, who schedules the weekly check-ins with the experts. After this kickoff, response templates are prepared. It is a good idea to be in close touch with the global and the local regulatory departments, who can give more insight into when to expect what. Because FDA questions can arrive at any time, agree in advance the process for action, the meeting frequency after questions arrive, and goals for time to completion.
For a marketing authorisation application to EMA, the day 80 and day 150 assessment reports that arrive ahead of the consolidated lists of questions at day 120 and day 180 help identify issues before the clock stop. Once an area of concern is identified in the preliminary questions, the team can begin preparing its response strategy, which is then refined after the final questions arrive. Additional analyses the anticipated queries require can be started then, so that they are ready if the question remains in the final list.
The response documents
Some agencies provide their own response templates, such as EMA and the MHRA, and some do not. Where none exists, prepare one, keeping its formatting and layout consistent with the dossier, and always double-check with the local regulatory lead that it can be used in that region.
FDA sends its questions as a short communication sheet that refers to the submission, states that the division has the following requests for additional information, and numbers the questions. One such first question read: "We were unable to locate some information in your submission. To facilitate timely review, we request that you provide XYZ." FDA usually provides little context with its questions, and it prescribes no response template.
A company may have its own FDA template, whose heading states that the document is the company's response to an FDA information request. One experienced writer always recommends showing the date the request was received prominently, which makes it much easier later to identify at a glance which questions a document answers. A one-statement introduction refers to the request and its receipt date and explains that, for ease of reference, the request is given in italics followed by the sponsor's response. Each item then repeats the question first, copied as received and in this template in italics, though the style can differ by template, followed by a heading for the response and a placeholder for its text. The header commonly carries the drug name and the submission number, and may name the eCTD module the document will sit in; the footer carries page numbers and commonly the date the responses are sent, often as month and year.
For EMA, take its response template, customise it a bit if needed, and copy the questions into it. Putting every response into one document is possible in theory, but one experienced writer would not recommend it. At a minimum, mirror the division EMA used for its assessment reports and keep quality, non-clinical, and clinical responses in separate documents, since different teams typically work on them. Clinical responses can be divided further into separate documents for pharmacokinetics, efficacy, and safety, and, depending on the number of questions, further still, such as a separate document for major objections on efficacy apart from the other efficacy concerns.
Within the EMA template, each area such as clinical aspects opens with the major objections for its topics, and each item gives the EMA question number, the copied question text, and a placeholder for the applicant's response. The grey box in the template is left exactly as it is: it exists for the rapporteurs and co-rapporteurs to give their assessments of the responses. The day 150 joint assessment report gives its assessment of each response in those grey boxes.
The response cycle
The cycle runs from regulatory affairs receiving the questions and their timelines and logging them in a tracker, through the kickoff that assigns roles, agreement of timelines, drafting, review, finalisation including quality control, translation, and publishing into a single package, to tracking each response's completion and regulatory affairs submitting them to the agency.
When the questions arrive, separate those already addressed in preparatory work from new ones. Triage the more complicated ones that may need longer — perhaps because external experts must be involved or fully new analyses are needed — so they can be prioritised, assign content authors to lead the discussion of strategy on the unanticipated ones, and identify issues that may need further contact with the regulator to clarify uncertainty. Once the final request is available, finalise the overall strategy as soon as possible. Where the list is extensive and the timelines short, it may be helpful to classify the questions by level of complexity, so that drafting of the responses to the easy ones can begin straight away in a staggered approach that avoids a log-jam at the end of the process. It is also important to identify early the questions that may require additional statistical outputs, because producing those may well be a rate-limiting step.
When the EMA day 120 list arrives, compare it against the preliminary questions already being worked on. Look for questions that are new, which is the more typical change, and questions that have been removed; for changes in wording, since even the slightest change is copied across into the response document; and for changes in numbering, which matter particularly when a detailed tracker is in use. Restructure the draft response documents around the actual day 120 numbering before continuing with the team.
Populate the template with the questions before the kickoff, because the meeting discusses each question and its writing strategy. Regulatory often schedules that kickoff, and it is led by both regulatory and medical writing. For each question the meeting identifies the subject-matter expert, discusses a high-level writing strategy, assesses whether additional analysis or justification is needed and who can provide any additional sources, and aligns on timelines and next steps. Regular check-ins with the experts follow, because the drafting is done with them.
Drafting the timelines means weighing several constraints. Company processes and standard operating procedures may fix the number of reviews and specific quality-control requests, such as whether quality control runs on the final draft alone or on the initial draft and the final one. Questions that demand new outputs need more time. Translation, where required, adds time, and whether translators can start from a draft version matters. Publishing time depends on the size of the document, the number of questions, and the number of hyperlinks. Once these are addressed, draft timelines for two scenarios: one for different response times, which is important for FDA, and one for whether questions need new outputs. Timelines for who delivers what must be agreed, and realigned as needed as challenges arise during the writing.
The tracker carries a row per question with its category — efficacy, safety, clinical pharmacology — its number, the person responsible, the drafting deadlines, its status, and notes from the writers or team members. A tracking sheet recording each response's status — in draft, in review or finalised — the open action items, and whether it will have appendices or literature is an essential tool, because the team often writes several responses in parallel.
For questions with very short turnarounds, try to update the tracker daily and send the team a daily update by email. It assures them that the work is progressing, keeps them engaged, and reminds them of their action items and responsibilities. FDA turnaround times are short, so the same status update may be worth using for FDA responses, to keep the team engaged and working collaboratively to finish them on time.
Maintain clear documentation and version control: the responses, their supporting data, and every updated document should be clearly documented and organised. That helps reviewers see what has been updated, and it prevents confusion between previous and revised versions.
Reading each question
Work out what the agency is actually asking for: additional data, a clarification or a rationale, or a further analysis. Identify the regulatory concern behind it — the drug's safety margin or exposure, the interpretation of efficacy, an inconsistency in the manufacturing process, a labelling claim — before drafting a response. Then assess whether the question is critical or minor, which helps the team prioritise its internal resources and timelines.
EMA's assessment reports and its day 120 list can sometimes show where a question came from and what the rationale behind it may be, which should be considered when answering it. The questions in the final request are not always clear and unambiguous, and they should always be interpreted in the context of the full assessment report, which may give further clues about the reviewers' concerns in case of doubt.
Where a question is really unclear, it is possible to go back to the authority and ask for clarification, rather than working on a response for two months and only then finding that it was not the question they asked.
Writing the response
The key principles of writing responses are to be direct and answer the exact question first, to be data-driven so that findings are supported by data, to be concise by summarising first and elaborating later, and to be consistent across all responses and the original submission. As a rule of thumb, a response should be as long as required and as short as possible. In the absence of official regulatory guidance on the format and content of responses, it is advisable to keep them brief, well structured and consistent with the layouts and presentations of the dossier, so that they are reviewer-friendly and easy to read.
Restate the agency's question first, clearly and in bold, verbatim as the regulatory team received it. Do not change it, even to fix a typo. Then give a clear, factual response without unnecessary detail: for responses, more does not mean better.
Add cross-references to the documents relied on, such as clinical study reports or modules of the original submission. Cross-references to any newly generated outputs likewise make it easy for the reviewer. Refer to the identifying item number and location in the original briefing package, so that reviewers can easily locate the initial positions.
Be comprehensive but concise, answer only what was asked, and stop as soon as the question is answered. It is imperative to focus only on answering the questions asked and to add no new data unless requested. Additional details the agency has not asked for may in the end trigger more new questions.
Stay open without over-interpreting or speculating. Acknowledge what the data actually say, whether positive or inconclusive, distinguish the known from the unknown, and always support findings with data.
Be consistent across the responses and with the original submission. Reviewers evaluate the responses in parallel with the original submission, so even a minor inconsistency can raise questions about data integrity. Reporting a mean exposure as 24.43 ng/mL in the submission and 24.4 ng/mL in the response can trigger another question, such as whether the datasets were updated; switching between micrograms and nanograms per mL across responses is another such inconsistency.
Conclude each response with a clear statement of whether the issue is resolved and, if not, whether a commitment such as a post-marketing commitment is being made, in which case give the agency its timelines. Provide clarity without introducing unnecessary complexity, and do not over-explain, because over-explaining can backfire and trigger more questions.
Responses to labelling questions should be coordinated with the draft label, and responses to RMP questions with the draft label and the draft RMP. Update the label first, then state in the response that it has been updated as requested and give the changes; they can be presented in a table giving the section number, the old text, and the revised text. For instance, where a Summary of Product Characteristics gives a recommended dose without a dosing frequency and the agency asks for the frequency, the old text "the recommended dose is 100 mg" becomes "the recommended dose is 100 mg once daily".
Tone of the responses
Be collaborative rather than defensive, and try to keep a respectful and collaborative tone throughout. Use language that signals partnership, such as opening with "to support your assessment", or "in alignment with previous agency feedback" where earlier discussions with the agency exist. The company should sound confident and sure of its position without being dismissive of the reviewers' comments and questions.
Demonstrate responsiveness: show that the question is understood, that the reviewer's concern behind it is respected, and that the sponsor is willing to provide the justification or the solution. Maintain traceability by always mentioning earlier discussions with the agency, such as a Type C meeting with FDA or scientific advice from EMA. Prioritise clarity, consistency, and scientific robustness in the responses. Use the questions as an opportunity to strengthen relationships with agencies, and treat them as part of a constructive scientific dialogue: responding clearly, respectfully, and cooperatively helps build trust and credibility with the health authority. The questions are also an opportunity to strengthen the benefit-risk narrative for the drug.
What a medical writer adds
Medical writers are often involved in preparing submission documents such as clinical overviews and clinical summaries.
For a trivial question, such as a request to provide a certificate of analysis, little medical writing support may be required. For more complex issues, the writer may be able to offer valuable assistance, for several reasons. The writer will likely have been closely involved in preparing the initial submission and so be familiar with the details of the project. Where the list of questions is extensive, the writer's ability to manage and oversee complex projects, to coordinate input from a variety of sources and to ensure consistency can be valuable to ensure high-quality responses. The process often requires working to tight deadlines, which medical writers are used to. And the tone of the responses needs careful consideration, so the language expertise of medical writers can also be important.
The medical writer is ideally positioned to identify the parties responsible for the supporting data and analyses the responses require. To guide an efficient and effective preparation of position statements and responses, the writer must understand the development process well and work closely with the team.
Preparing responses to requests for supplementary information can be stressful, but it is also rewarding. Preparing the initial submission is only the start, and involvement in the post-submission process can give the writer, who will likely have worked hard within a team, the satisfaction of seeing the job through. It can also serve as feedback on how the original submission documents were prepared and give some enlightenment on what goes through a reviewer's mind. All this deepens the writer's knowledge of the approval process and helps make a more complete writer.