1. Reactive. Answer from the local label, referenced. Straightforward, and the model for everything else.

2. Reactive, but constrained. A comparison was asked for, so answering is permitted — but only from genuine head-to-head data. If none exists, say so plainly. Do not assemble a cross-trial comparison; that is scientifically invalid and treated as a promotional claim.

3. Neither — this is a signal. Nothing goes back to the enquirers, but the pattern goes to pharmacovigilance. Enquiry clustering is a recognised signal source and surfacing it is part of the job.

4. Reactive. An unsolicited off-label question at a congress is still unsolicited. Handle it under the off-label process: state that the use is not approved, answer only what was asked, document it. Congress does not change the rules, and being face to face makes scope discipline harder.

5. Proactive, and out of scope. Preparing material to support a sales call is not medical information. The publication may be requested through the proper promotional review route by the team that owns it.

6. Reactive, consumer. General interaction information from the patient leaflet may be provided in plain language. Do not advise this individual what to do — direct them to their prescriber or pharmacist.

The pattern: four of six are reactive and answerable. The two that are not fail for different reasons — one is a signal rather than an enquiry, the other is a request from inside the company to do promotional work. Neither is refused rudely; both are routed correctly.

Three parallel obligations from one call.

1. Medical information response. Renal dose adjustment for the product. Answer from the Philippine approved label — not the SmPC, not the USPI. Provide the label position, referenced. Scope is renal dosing and nothing else.

2. Adverse event report. A rash, with a positive dechallenge (settled on withdrawal). This is an adverse event and must be captured and forwarded to pharmacovigilance on the required timeline. “This is probably nothing” does not remove the obligation — and a positive dechallenge is clinically informative.

3. Product quality complaint. Chipped tablets. Separate route to quality, with batch number and pack details if obtainable.

What you record at intake:
· Question verbatim, as spoken
· Enquirer type: healthcare professional (nurse)
· Country: Philippines — determines which label applies
· Product, strength and regimen: 100 mg daily
· Contact details — needed for the response and for pharmacovigilance follow-up on the rash
· Date, time and channel
· Flags raised: AE yes, PQC yes

What you say before ending the call: that you will send the renal dosing information; that the rash will be reported to the safety team, who may contact her for further detail; and that the tablet defect will go to quality, for which you need the batch number. Ask for the batch number now — it is much harder to obtain later.

The trap: the caller explicitly deprioritised both reportable items — “probably nothing”, “anyway, mainly I need the renal dosing”. Following the caller’s priorities is exactly how these get missed, and it is the commonest inspection finding against this function.

1. The label for that market. Not the publication — contraindications are a labelling matter and can differ between markets even for the same product.

2. The primary publication of that trial. Not the label — a label carries approved-use information, not full trial results, and usually not the response rate.

3. The guideline itself, quoted accurately. Not a company summary of the guideline, and no interpretation of what the recommendation implies.

4. Head-to-head data if any exist; otherwise state that none exist. Not a comparison assembled from two separate trials — different populations, different eras, different endpoints. Cross-trial comparison is scientifically invalid and is treated as a promotional claim.

5. Published evidence if any; otherwise data on file, or state that no data are available. Not extrapolation from moderate renal impairment — dialysis is a different question, and inferring an answer would manufacture evidence.

6. The label — and this is an off-label enquiry. The approved dose is 100 mg. 200 mg was studied but is not approved, so the response states the approved dose, notes that 200 mg is not approved, and is handled under the off-label process. Do not simply confirm the trial dose because it appears in a publication.

The pattern: three of six are answered by the label, and in two of those the publication is a tempting wrong answer. Question 6 is the one that catches people — the dose is real, published and studied, and providing it as though it were approved is off-label information given without recognising it as such.

Title: Dose adjustment of [product] in patients with hepatic impairment

Summary answer:

“The approved label recommends no dose adjustment in mild hepatic impairment (Child-Pugh A). In moderate impairment (Child-Pugh B), the dose should be reduced to 50 mg once daily, reflecting an approximately 1.8-fold increase in exposure observed in a dedicated pharmacokinetic study. Use is not recommended in severe hepatic impairment (Child-Pugh C), in whom the product has not been studied. No efficacy data are available in patients with hepatic impairment.”

Five things this summary does:
1. Answers all three severity categories, because the question did not specify one.
2. Gives the reason for the moderate-impairment reduction, so the recommendation is interpretable rather than arbitrary.
3. States why severe is not recommended — not studied — rather than leaving the reader to infer a safety finding.
4. Discloses the absence of efficacy data. The enquirer asked about dosing, but a clinician making this decision needs to know efficacy is unstudied here. This is what balanced means, and omitting it would make the summary read more favourably than the evidence supports.
5. Stands alone. Nothing below section 2 is needed to act on it.

What is deliberately absent: the PK study n = 24 and its design belong in supporting data, not the summary. And no reassurance about tolerability — that was not asked and is not evidenced here.

Rewrite:

“In the pivotal randomised, placebo-controlled trial, the primary endpoint of progression-free survival was met: median progression-free survival was 8.9 months versus 5.2 months (HR 0.66, 95% CI 0.53–0.82; p < 0.001). Grade 3 or higher adverse events occurred in 147/389 patients (37.8%) versus 121/388 (31.2%), and adverse events led to discontinuation in 31/389 patients (8.0%). In the prespecified subgroup aged 65 and over (n = 190), the hazard ratio was 0.77 (95% CI 0.57–1.04); no significant interaction by age was observed (interaction p = 0.58), and the trial was not powered for subgroup comparisons. No head-to-head comparative data with other agents are available.”

The ten problems:
1. excellent efficacy profile — unreferenced judgement.
2. Importantly — editorialising adverb.
3. a clear benefit — replaced with the effect size and interval.
4. only 34% — only editorialises; and progression rate is not the endpoint that was met.
5. tolerability was very good — judgement; replaced with rates and denominators.
6. discontinuation was rare — rare has a defined frequency meaning in labelling; give the number.
7. well established as the preferred option — a positioning claim, outside the function’s scope.
8. compares favourably with alternatives — a comparative claim with no head-to-head data. Replaced with the accurate statement that none exist.
9. Notably, elderly patients did just as well — overstates a non-significant interaction in a subgroup the trial was not powered for.
10. In our experience the product is rarely associated with significant toxicity — anecdote presented as evidence, unreferenced, and from an institutional voice that has no standing here.

Note the rewrite is not longer. Everything removed was judgement; everything added was a number that was available all along.

1. Label, section 4.2 (posology), with market and version date. Not the trial publication — the approved dose is a labelling matter.

2. Primary publication of that trial: authors, journal, year, volume, pages. Specify which trial and which analysis if more than one exists.

3. Label, section 4.3 (contraindications), with market and version date. Note that contraindications can differ between markets, so the market must be stated.

4. Data on file, with a unique identifier and a retained retrievable source. But first check whether a published pooled analysis exists — if it does, use that instead.

5. The guideline itself: issuing body, title, version or year, and the specific section. Not a company summary of the guideline, and quoted rather than paraphrased in a way that shifts the recommendation.

6. Not referenceable, and should not appear. In clinical practice and generally well tolerated are an unreferenced favourable judgement. Replace with adverse event rates and denominators from a cited source.

The pattern: five of six are straightforward once you know which document type answers which question. The sixth is the one that gets written anyway, because it feels like helpful reassurance rather than a claim.

Model standard response document

1. Question
Dose adjustment and available efficacy data for [product] in patients with moderate renal impairment.

2. Summary answer
“The approved label recommends reducing the dose to 50 mg once daily in patients with moderate renal impairment (creatinine clearance 30–59 mL/min), reflecting an approximately 1.6-fold increase in exposure. No dose adjustment is required on the basis of age alone. In the pivotal randomised trial, 88 patients with moderate renal impairment were included in a prespecified subgroup analysis; the hazard ratio for the primary endpoint was 0.71 (95% CI 0.44–1.15), with no significant interaction by renal function (interaction p = 0.62). The trial was not powered for subgroup comparisons, and no prospective efficacy study has been conducted in patients with renal impairment.”

3. Background
Brief: renal elimination pathway and why exposure rises in moderate impairment. Nothing about the product beyond that.

4. Supporting data
Dedicated renal pharmacokinetic study (n = 32) with the exposure findings; the pivotal trial design, population and primary result; the renal and age subgroup results with intervals and interaction p values; explicit statement that patients with creatinine clearance below 30 mL/min were excluded from the trial.

5. References
SmPC sections 4.2 and 5.2 with version date; pivotal trial publication; renal PK study publication.

6. Disclaimers and version
Not individual clinical advice; treatment decisions rest with the prescriber. Document version and date. No off-label statement required — everything asked and answered is within the approved label.

---

Rubric — 16 points

Structure (4)
1. All six sections present
2. Summary stands alone and answers both parts of the question
3. Background limited to what the answer needs
4. Version and date present

Scope (3)
5. Both parts answered — dosing and efficacy. The enquiry had two questions.
6. Nothing volunteered beyond them — no hepatic dosing, no other indications
7. Age addressed, because the enquirer specified a 71-year-old — the question contained it implicitly

Accuracy and balance (6)
8. UK SmPC used, matching the enquirer’s market
9. Exposure increase given as the reason for the dose reduction
10. Subgroup result reported with n, interval and interaction p value
11. Stated that the trial was not powered for subgroup comparison
12. Stated that no prospective efficacy study exists in renal impairment — the honest answer to the second question
13. No promotional language: no only, importantly, well tolerated, reassuring

Referencing (3)
14. Every statement referenced
15. SmPC cited with section and version
16. Subgroup data attributed to the publication, not presented as label content

Scoring
13–16 — pass. This would go through review.
9–12 — close. Note which block you lost points in.
Under 9 — redo 2.1 to 2.3 before retrying.

The three that separate people: item 5, because the second question is easy to lose behind the first; item 12, because “no prospective efficacy study exists” feels like an unhelpful answer and is in fact the most useful sentence in the document; and item 7, because the enquirer never asked about age — they stated it, and a good response notices the difference.

Summary answer:

“[Product] is approved for use in adults aged 18 years and over. Use in adolescents aged 12 to 17 years is not included in the approved label.

Two published reports describe use in this age group. In an open-label, single-arm study of 38 adolescents, the objective response rate was 29%; the study was uncontrolled, and no new safety signals were identified. A case series of seven patients reported mixed outcomes. No randomised controlled data are available in this population.

Treatment decisions remain a matter for the prescribing physician.”

What was deliberately left out, and why:

1. The ongoing Phase III trial and the planned submission. Not asked about, and communicating a planned indication is pre-launch promotion of an unapproved use. This is the most tempting omission to get wrong, because it feels like useful information for a paediatrician deciding what to do.

2. The adult efficacy data. Strong, approved, and irrelevant — the question was about adolescents. Including it would imply extrapolation the evidence does not support.

3. The related product approved from age 16. A different product, and mentioning it is both outside the question and a class-level inference with no evidential basis for this product.

What was deliberately included: the case series with mixed outcomes, and the explicit statement that no randomised data exist. Both make the answer less favourable, and both are required for it to be balanced.

Documentation: logged as an off-label enquiry, separately tracked. If adolescent enquiries cluster over the coming months, that pattern goes to the medical team — and, if the pattern looks prompted rather than spontaneous, to compliance.

1. Adverse event. Transaminase elevation with a positive dechallenge. All four elements present. To pharmacovigilance. Tell the caller the safety team will record it and may contact them.

2. Product quality complaint. To quality, with batch number, expiry and pack details — ask for these now, they are hard to get later. No adverse event unless a patient was harmed.

3. Special situation — lack of effect. Reportable to pharmacovigilance in most systems. Easy to miss, because it sounds like dissatisfaction rather than a safety report. Also answer any information question they have about expected time to response.

4. Special situation — exposure in pregnancy. Reportable even with a normal outcome to date, and follow-up will be sought on the pregnancy outcome. Note the reporter is a family member, not a healthcare professional; consumer-appropriate language, and direct her sister to her own clinician.

5. Information enquiry only. Nothing reportable. Answer from the label.

6. Information enquiry, with care. No identifiable patient and no specific case — this is a general question, not a report, so it is not an adverse event. Answer from the label's undesirable effects section. But if the caller then says “actually I have a patient with it,” that becomes a report — and this is exactly how a general question turns into one mid-conversation.

The pattern: six calls, four reportable, and only one of the four is a straightforward adverse event. Items 3 and 4 are the ones missed most often, because nothing bad appears to have happened.

1. “No head-to-head randomised trial has compared [product] with [competitor]. A published network meta-analysis including both agents reported no statistically significant difference in progression-free survival. As an indirect comparison, this analysis is subject to the limitations of differing trial populations and designs, and should not be interpreted as equivalent to a head-to-head comparison. [Reference]”

2. “A Phase III trial of [compound] is registered on [registry], reference [number], where the publicly available study design and eligibility criteria can be reviewed. No efficacy or safety data are available for this investigational compound, and it is not approved for use in any indication.”
What is left out: the planned submission. Not asked about, and stating it communicates expected future availability.

3. “No head-to-head randomised trial has compared the tolerability of [product] with [competitor], and the rates observed in separate trials cannot be validly compared because the populations, eras and adverse event reporting conventions differ. In the pivotal trial of [product], grade 3 or higher adverse events occurred in 147/389 patients (37.8%) versus 121/388 (31.2%) with placebo. [Reference]”

Why response 3 is the hardest of the three: your number, 37.8%, is lower than theirs, 41%. It is genuinely tempting to place them side by side and let the reader draw the conclusion — without stating any comparison yourself. That is still a cross-trial comparison, constructed by juxtaposition rather than by sentence. The correct response gives your data against its own comparator and states explicitly why the two trials cannot be compared.

Version A — healthcare professional

“[Product] may be administered with or without food; no clinically relevant food effect has been identified. Concomitant grapefruit juice should be avoided: co-administration increases exposure approximately 2.1-fold through inhibition of CYP3A4. [SmPC section 4.5, version dated (date)]”

Version B — patient

“You can take this medicine with or without food — it does not matter whether you have eaten.

However, you should not drink grapefruit juice or eat grapefruit while taking it. Grapefruit can raise the amount of the medicine in your body, which may make side effects more likely.

If you have already had grapefruit, or if you have questions about anything else you eat or drink, please speak to your doctor or pharmacist. They know your full medical history and can advise you properly.”

Four differences that matter:
1. CYP3A4 and 2.1-fold are gone from version B — the mechanism and magnitude help a prescriber and mean nothing to a patient. The consequence is retained.
2. Version B adds grapefruit the fruit, not only the juice. A patient reading “juice” may reasonably conclude the fruit is fine.
3. Version B anticipates the obvious next situation — what if I already have — and routes it to a clinician rather than answering it.
4. Version B closes by directing her to her own prescriber. Every consumer response does.

What is the same in both: the warning itself. Version B is simpler and more specific, not weaker. If a plain-language version leaves the reader less clear about what to avoid, it has failed.

Six adaptation issues:

1. Standard dose differs. Global says 100 mg; locally 75 mg is authorised. The document must state 75 mg. This alone makes the translated global version unusable.

2. The 100 mg strength is not marketed here. Any reference to it must be removed, not merely converted — it is not obtainable.

3. Indication is narrower. First-line use is approved globally and not locally. If an enquiry concerns first-line use it is off-label in this market, and the local document must not present it as approved.

4. An additional national warning exists on hepatic monitoring. It is absent from the global document and must be included, because it is part of the local approved position.

5. The CCDS citation must go. The company core data sheet is an internal reference and is not the approved label in any market. The local label is cited instead.

6. Reduced dose happens to match at 50 mg — but that must be confirmed against the local label rather than assumed, and the confirmation documented.

What the local document must do: state 75 mg as the standard dose; remove all reference to the 100 mg strength; limit the approved indication to second line and flag first-line enquiries as off-label; add the national hepatic monitoring warning; cite the local label with section and version; and go through local medical review and approval before use.

The one that would do most harm: issue 1. A perfectly translated document telling clinicians the approved dose is 100 mg, in a market where 75 mg is authorised and 100 mg is not even available, provides off-label dosing as though it were approved — to every enquirer, repeatedly, until someone notices.

1. Volume up 58%. Context needed — a launch, a guideline change, a publication, or a competitor withdrawal. Not a concern in itself, but it explains the response-time miss.

2. Neuropathy enquiries up from 4 to 31. Escalate to pharmacovigilance now. The event is labelled as common, but an eightfold rise in questions about one event is exactly the pattern signal detection looks for — it may indicate a higher real-world frequency or greater severity than the label conveys.

3. Paediatric off-label enquiries up from 3 to 22, all from one region. Two possible explanations and both need action. It may be genuine unmet need — intelligence for the medical team. It may be that someone in that region is prompting the questions, in which case the enquiries are not unsolicited and the whole chain is non-compliant. Escalate to compliance as well as medical. The geographic concentration is what makes this suspicious rather than merely interesting.

4. SRD coverage down from 84% to 68%. The library is behind the questions being asked. Identify the recurring bespoke topics and commission new SRDs — and this is also partly causing finding 5.

5. Response time 4.1 days against a 3-day standard. Consistent with volume up and coverage down; more bespoke writing takes longer. Addressing finding 4 will partly fix this.

6. Three passing mentions, two AE reports. Investigate the third immediately. Either it was correctly assessed as not meeting the criteria and that assessment is documented, or an adverse event was missed. This is the smallest number on the page and the most serious.

The order of priority: finding 6 first, because a missed AE is a regulatory breach. Then 2 and 3, both escalations. Then 4, which fixes 5. The headline number — volume up 58% — is the least important thing here.

1. Not unsolicited. “I told him you’d have it” means the company prompted the enquiry, so the off-label exemption does not apply. Action: do not provide the material. Explain why to the colleague, and if the customer wishes to enquire directly and unprompted, handle it then. Escalate to compliance — a representative directing customers toward off-label information is an issue in itself.

2. Potential overdose, possibly ongoing patient risk. Action: advise her to contact emergency services or take him to hospital immediately — that is signposting to urgent care, not clinical advice, and it is appropriate. Capture as an adverse event and special situation. Escalate to the medical lead immediately; do not wait for the normal timeline.

3. Not a medical information enquiry. A competitor’s slide is not company material and is not yours to distribute. Action: direct them to the presenting author or the congress organisers. No escalation needed.

4. Not answered by medical information. Action: route to regulatory affairs immediately and do not respond directly. Assemble the enquiry record and the archived document version, since they will be needed — but the response goes through regulatory.

5. Adverse event in a public post. Identifiable patient (the poster), identifiable reporter (same person), suspect product, and an event. All four elements present. Action: capture and forward to pharmacovigilance; reply publicly only to invite them to a private channel; do not answer clinically in public and do not delete the post.

The pattern: only one of the five is a normal enquiry, and it is the one you decline. Two require immediate escalation on different routes, one goes to compliance, and one is not your function at all. This is closer to a real week than the drills that precede it.

The eighteen points

Intake and classification (5)
1. Recorded the question verbatim, all four parts.
2. Enquirer type: healthcare professional. Country: Brazil — which determines the label.
3. Identified four separate matters in one email: off-label enquiry, adverse event, quality complaint, comparative question.
4. Flagged the enquiry as off-label for the adolescent part.
5. Flagged that the enquiry was prompted by a representative — the single most important observation in the whole case.

The adolescent question (4)
6. Recognised it is NOT unsolicited. “Your representative suggested I write to you” means the company initiated it, so the off-label exemption does not apply. The data must not be sent on this basis.
7. Explained this to the enquirer courteously, without implying he did anything wrong.
8. Did not mention the ongoing trial or the planned submission.
9. Did not use the global SRD — it is written for a 100 mg market and is wrong for Brazil.

The adverse event (4)
10. Captured mucositis as an adverse event and forwarded to pharmacovigilance. All four minimum elements present.
11. Recognised the positive dechallenge and rechallenge pattern — resolved on stopping, not recurred at lower dose — and recorded it, as it is clinically informative.
12. Answered the actual question: mucositis is listed as a common adverse reaction in the Brazilian label, with reference.
13. Noticed the patient is on 100 mg, which is not the approved Brazilian dose (75 mg). This is off-label dosing, and it must be flagged in the response and in the safety report.

The complaint and the comparison (3)
14. Cracked vial seals raised as a product quality complaint, with a request for batch numbers — noting the vials were discarded, so obtain details quickly.
15. Comparative question: stated that no head-to-head data exist; cited the network meta-analysis identified as an indirect comparison with its limitations.
16. Did not place the two trials’ numbers side by side.

Escalation (2)
17. Escalated to compliance regarding the representative prompting an off-label enquiry.
18. Escalated the adverse event to pharmacovigilance and the complaint to quality, on their respective timelines.

---

Model response summary

“Thank you for your enquiry.

On mucositis: severe mouth ulcers are consistent with mucositis, which is listed as a common adverse reaction in the Brazilian product information [reference, section, version]. I have passed the details of your patient to our safety team, who may contact you for further information. I note from your message that this patient was receiving 100 mg; the approved dose in Brazil is 75 mg once daily [reference].

On use in patients under 18: [product] is approved in Brazil for adults aged 18 and over only. I understand our representative suggested you contact us on this topic. As information on unapproved uses can only be provided in response to an enquiry you have raised independently, I am not able to send materials on this basis. If you would like to raise the question directly, please write to us and we will respond.

On the comparative question: no head-to-head randomised trial has compared [product] with [competitor]. A published network meta-analysis including both agents is available [reference]; as an indirect comparison it is subject to the limitations of differing trial populations and designs.

On the vials: I have raised a product quality report. Could you provide the batch numbers and expiry dates from the affected packs?”

Scoring
14–18 — pass. You handled a genuinely difficult enquiry correctly, including the two points most people miss.
9–13 — close. Note which block you lost points in and redo that module.
Under 9 — work through Week 3 again before retrying.

The two that separate people: points 6 and 13. Almost everyone answers the adolescent question helpfully, because the clinician has real patients and a real problem — and the single sentence about the representative makes it non-compliant. And almost nobody notices that 100 mg is not the approved Brazilian dose, because the enquirer mentioned it in passing while asking about something else entirely.