Medical Affairs & Governance

Clinical advisory board

We are appointing an independent clinical advisory board to review editorial policy, source hierarchy, and non-promotional governance. Advisors have no commercial edit rights and no access to individual user queries.

Planned expertise: clinical pharmacology, internal medicine / SNS practice, and cardiometabolic guideline alignment.

Expressions of interest: medicalaffairs@smartprescribe.app

Hospital / society pilot

We offer an 8-week unbranded pilot for hospitals, faculties, or specialty societies: verified HCP onboarding, Play + Ask + Prescribe access, and an anonymised aggregate report. No individual chat data is shared.

• Target: 30–50 active verified clinicians
• Success metrics: onboarding, Ask AI engagement, pre/post confidence survey
• Non-binding letter of intent available on request

View pilot LOI template →

Pre/post survey: baseline · post-pilot

Platform positioning

• Audience: Verified healthcare professionals (manual or institutional verification). Ask AI and treatment-planning tools are not open to the general public.
• Purpose: Continuing professional education — pharmacology, prescribing considerations, guideline-aligned management, and drug safety.
• Regulatory stance: Educational software for HCPs. Not marketed as a medical device or regulated clinical decision support product. See Terms of Service.
• Primary surfaces: Daily clinical cases (Play), evidence-based Q&A (Ask AI), and structured treatment planning (Prescribe — educational).

Editorial policy & source hierarchy

Clinical content is generated and retrieved using a fixed source hierarchy:

1. National SmPC / RCM — INFARMED (Portugal), AEMPS, ANSM, MHRA, etc., matched to the clinician's practice location.
2. EMA EPAR / EU product information where applicable.
3. National guidelines — e.g. DGS normas (Portugal), NICE, ESC, KDIGO.
4. Curated interaction & pharmacology datasets with evidence grades and verifiable references (PMID/DOI).
5. Peer-reviewed literature — only when higher-tier sources are insufficient; low-evidence types (case reports, letters) are flagged.

Update cadence: Regulatory documents and guidelines are ingested on a scheduled basis. Guideline drift is monitored; material older than policy thresholds is flagged for reverification. See How we verify answers.

Human oversight: High-traffic clinical outputs (e.g. drug–drug interaction cards) follow a documented pre-publication QA checklist. Admin review applies to user feedback and flagged content.

Non-promotional commitment

• AI does not rank, prefer, or recommend branded medicines for promotional purposes.
• Therapeutic suggestions follow guideline class/level of evidence and generic/active-molecule framing unless the user asks about a specific product.
• No sponsored AI answers. Funding partners cannot edit, approve, or influence individual chat responses.
• Fair balance: benefits and risks are presented where clinically relevant; SmPC contraindications and interactions are prioritised over summary marketing claims.
• All Ask AI replies display an educational disclaimer and verifiable source links where available.

Off-label & indication policy

• Default responses address approved indications in the user's jurisdiction (SmPC/RCM-first).
• Off-label or academic questions may receive educational context when explicitly requested; such content is labelled as outside approved labelling and requires independent verification.
• SmartPrescribe does not encourage off-label promotion and does not substitute company medical information services.

Sponsorship & partnership policy

SmartPrescribe may enter unbranded disease-education partnerships with pharmaceutical Medical Affairs teams, subject to written agreement and compliance review.

• Permitted: Funding of independent educational modules (e.g. hypertension management aligned with DGS/ESC); grant support for guideline ingestion and QA; anonymised aggregate insight reports on knowledge gaps (no individual query data).
• Not permitted: Branded chat prompts; pay-to-rank molecules in AI output; sponsor review or approval of individual Ask AI responses; access to identifiable user queries or patient data.
• Labelling: Supported modules display: “Supported by [Company] — content independently verified by SmartPrescribe.”
• Firewall: Clinical editorial team and QA pipelines are separate from commercial discussions.

Partnerships are structured to align with EFPIA / APINFPMed educational principles. Legal review on both sides is required before launch.

Pharmacovigilance

SmartPrescribe is not a adverse-event (AE) reporting system. Users who mention suspected adverse reactions should follow national procedures:

• Portugal: INFARMED — Notifica espontânea
• EU: National competent authority or marketing authorisation holder medical information line.

We do not forward user chat content to pharmaceutical companies. Product safety concerns should be reported through official channels.

Data & independence

• We do not sell personal or clinical data to third parties, including sponsors.
• Anonymised, aggregated usage metrics may be reported to funding partners under contract (e.g. module completion rates, topic themes).
• Details: Privacy Policy.

Contact Medical Affairs

For governance documentation, pilot proposals, or unbranded education partnerships:

medicalaffairs@smartprescribe.app