- Scope
- Validation of computerised systems in regulated life sciences
- Key requirements
- Risk-based software categorisation (Cat 1–5), validation plans, IQ/OQ/PQ, change control, audit trails
- Applies to
- Pharma, biotech, medtech manufacturers
- Overlaps
- Overlaps with 21 CFR Part 11 & Annex 11 on electronic records and audit trails
- What sets it apart
- Risk-based categorisation of software; the only framework specifically guiding how to validate GxP systems
About this guide
This guide is for leaders, quality, regulatory affairs and IT managers in life sciences and healthcare companies: biotech, pharma, medtech, laboratories, clinics, e-health. It presents the main regulatory frameworks that apply to their processes and data, the friction points between these frameworks, and how digitizing processes makes compliance simpler.
It does not replace advice from a qualified regulatory affairs consultant: it is a map to find your way, ask the right questions and set priorities.
Contents
Chapter 1
Compliance: a foundation, not a burden
In life sciences and healthcare, regulatory compliance is not an administrative formality: it is the condition for market access. GxP, GDPR, HDS, ISO 13485: each framework translates a safety, traceability and quality requirement that protects patients and guarantees the integrity of scientific data.
The cost of compliance rarely comes from the requirements themselves. It comes from the way they are met: evidence rebuilt by hand before every audit, non-conformities tracked in spreadsheets, paper signatures, documents whose latest version nobody can find.
Digitized processes are traceable by design. Compliance stops being a layer added afterwards: it becomes a property of the processes, and the evidence is already there on audit day.
Chapter 2
AI raises the bar
Adopting artificial intelligence in R&D and healthcare processes creates new regulatory challenges. An AI model used in a medical or regulatory decision must be transparent, auditable and validated: this is required by the EU AI Act (2024), but also by existing GxP rules on computerized system validation (GAMP 5, 21 CFR Part 11).
Poorly traced input data can invalidate every conclusion of a study. Conversely, a traceable, well-governed data infrastructure makes AI usable in high-stakes regulatory contexts, including submissions to the FDA or the EMA.
That is why, at Terminus 2π, AI agents embedded in processes work with human approval, and every action is traced: who proposed what, who approved it, on which data.
Chapter 3
Overview: 13 frameworks in 7 domains
Thirteen frameworks shape the compliance of biotech, pharma and medtech operations. They cover seven domains: quality systems, data integrity, privacy / gdpr, health data, it systems, security / gxp, environment. The table gives the overview; each framework is then detailed.
| Framework | Domain | Region |
|---|---|---|
| GAMP 5 | IT / Quality | Global (ISPE) |
| GLP | GxP / Safety | OECD / Global |
| 21 CFR Part 11 | Data integrity | USA (FDA) |
| RGPD / GDPR | Privacy | EU (applies globally) |
| HDS | Health data | France (ANS/ANSSI) |
| ISO 13485 | Quality | Global (ISO) |
| ICH Q10 | Quality | Global (ICH) |
| GMP / EudraLex Vol.4 | GxP / Safety | EU (EMA) |
| GCP / ICH E6(R3) | GxP / Clinical | Global (ICH) |
| ISO 27001 | IT security | Global (ISO) |
| ICH Q9 (R1) | Risk management | Global (ICH) |
| REACH / CLP | Environment / Safety | EU (ECHA) |
| HIAPP | eHealth / Apps | France (ANS / HAS) |
- Scope
- Non-clinical safety studies (toxicology, ecotoxicology)
- Key requirements
- Study director accountability, SOPs, raw data archiving, test facility inspection, specimen retention
- Applies to
- CROs, labs conducting non-clinical studies for regulatory submission
- Overlaps
- Shares data integrity & audit trail goals with 21 CFR Part 11 / GAMP 5
- What sets it apart
- Covers the physical conduct of studies (animals, reagents, notebooks), not just electronic systems
- Scope
- Electronic records & electronic signatures in FDA-regulated industries
- Key requirements
- ALCOA+ data principles, audit trails, e-signature controls, system validation, access controls, record retention
- Applies to
- Any US FDA-regulated entity (pharma, biotech, device, food)
- Overlaps
- Mirrors EU Annex 11 (EudraLex); both enforced alongside GAMP 5 guidance
- What sets it apart
- Legally binding US regulation; specifically governs e-signatures as legally equivalent to handwritten ones
- Scope
- Personal data processing of EU residents
- Key requirements
- Lawful basis, data subject rights, DPO appointment, DPIA, data minimisation, breach notification ≤72h, cross-border transfer rules
- Applies to
- Any entity processing EU personal data, including patient data in trials
- Overlaps
- Overlaps with HDS on health data; pairs with ICH E6 on clinical trial data governance
- What sets it apart
- Broadest personal data law; unique right-to-erasure tension with GxP data retention requirements
- Scope
- Hosting of personal health data on behalf of third parties in France
- Key requirements
- Mandatory HDS certification, ISO 27001 base, data localisation (FR/EU), subcontractor control, reversibility plan
- Applies to
- Cloud providers, SaaS vendors, CROs hosting French patient data
- Overlaps
- Built on GDPR + ISO 27001; stronger than GDPR alone for health data hosted in France
- What sets it apart
- Requires certified host — cannot self-certify; contractual chain must be fully certified end-to-end
- Scope
- Quality management system for medical devices
- Key requirements
- Design controls, risk management, CAPA, post-market surveillance, supplier qualification, traceability
- Applies to
- Medical device manufacturers & supply chain
- Overlaps
- Overlaps ISO 9001 (QMS base); aligns with EU MDR / IVDR and 21 CFR 820
- What sets it apart
- Device-specific QMS; mandatory for CE marking under MDR — not applicable to pure pharma/biotech
- Scope
- Pharmaceutical quality system across product lifecycle
- Key requirements
- Management responsibility, continuous improvement, CAPA, knowledge management, lifecycle approach (dev → discontinuation)
- Applies to
- Pharmaceutical & biotech companies globally
- Overlaps
- Complements ICH Q8 (design) & Q9 (risk); implemented via EU GMP Part III / FDA guidance
- What sets it apart
- Lifecycle view — explicit coverage from early development to market withdrawal, unlike site-focused GMP
- Scope
- Manufacturing of medicinal products for human & veterinary use
- Key requirements
- Premises & equipment, personnel, documentation, production, QC, QA, outsourced activities, complaints, recalls; Annex 11 (computerised systems)
- Applies to
- EU pharmaceutical & biotech manufacturers, importers
- Overlaps
- Mirror of US cGMP (21 CFR 210/211); Annex 11 ≈ 21 CFR Part 11
- What sets it apart
- Legally enforceable EU manufacturing law; Annex 11 is the authoritative EU counterpart to GAMP 5 recommendations
- Scope
- Design, conduct, recording and reporting of clinical trials
- Key requirements
- Protocol, IRB/IEC approval, informed consent, investigator responsibilities, data management, monitoring, adverse event reporting, TMF
- Applies to
- Sponsors, CROs, investigators running clinical trials
- Overlaps
- Requires GAMP 5-compliant CTMS/EDC; patient data governance intersects with GDPR
- What sets it apart
- Solely governs the clinical trial process; E6(R3) adds risk-based monitoring and data governance modernisation
- Scope
- Information security management system (ISMS)
- Key requirements
- Risk assessment, 93 controls (Annex A), ISMS scope, incident management, business continuity, supplier security, access management
- Applies to
- Any organisation handling sensitive data — prerequisite for HDS
- Overlaps
- Foundation for HDS; supports GDPR Article 32 (technical measures); underpins GAMP 5 infrastructure controls
- What sets it apart
- Horizontal security standard — not industry-specific; must be combined with sector regulations (GMP, HDS, GDPR)
- Scope
- Quality risk management across pharmaceutical operations
- Key requirements
- Risk assessment tools (FMEA, HAZOP, FTA), risk review, risk communication, integration into QMS and regulatory decision-making
- Applies to
- Pharma, biotech — any GxP operation
- Overlaps
- Provides the risk methodology used by GAMP 5 for software categorisation and by GMP for process validation
- What sets it apart
- Pure methodology standard — defines how to do risk management; referenced by virtually every other GxP framework
- Scope
- Registration, evaluation, authorisation of chemicals; hazard communication
- Key requirements
- Substance registration (REACH), SVHC authorisation, safety data sheets (SDS), hazard labelling & packaging (CLP/GHS)
- Applies to
- Biotech companies manufacturing, importing or using chemical substances
- Overlaps
- Intersects with GLP (ecotoxicology data feeds REACH dossiers); links to GMP for excipient safety
- What sets it apart
- Covers the chemical substance itself (not the product/drug) — distinct regulatory pathway from medicinal product authorisation
- Scope
- Referencing and certification of digital health applications integrating the national ecosystem (Mon Espace Santé, ENS)
- Key requirements
- GDPR compliance, HDS hosting if personal health data, FHIR/HL7 interoperability, CE marking if SaMD, clinical or usage validation, consent management, accessibility (RGAA)
- Applies to
- Digital health app publishers, mHealth and DTC solutions targeting the French market or the national digital health bouquet
- Overlaps
- Built on GDPR, HDS and ISO 27001 as a foundation; aligns with ISO 13485 for apps with a medical claim (SaMD)
- What sets it apart
- National framework specific to Mon Espace Santé; jointly evaluates security, interoperability, clinical validity and user experience
Chapter 4
Four tensions to anticipate
Frameworks overlap, and sometimes contradict each other. Four friction points come up often; they are best handled when processes are designed.
1. Right to erasure (GDPR) versus retention (GxP)
The GDPR gives people the right to have their data erased; GxP rules require some records to be kept for up to 15 to 30 years. These two obligations are in direct tension. They call for case-by-case legal analysis and pseudonymization strategies that separate a person’s identity from the data that must be kept.
2. GAMP 5 validation of SaaS and cloud tools
Validation obligations (GAMP 5, Annex 11) also apply to SaaS tools: EDC, LIMS, MES. Choosing a cloud provider means checking that it can supply the required validation documentation.
3. An HDS-certified hosting chain
Even if your own systems comply, the host of personal health data must hold HDS certification (French health data hosting). The obligation extends to every subcontractor in the chain: self-certification is not allowed.
4. Two pathways: REACH and medicines regulation
REACH covers the chemical substance itself; the authorization of the medicine or biologic (EMA, FDA) covers the product. A biotech may need to run both procedures in parallel for the same molecule.
Chapter 5
What digitized processes change
Most of the requirements listed in chapter 3 come down to a few questions: who did what, when, on which data, with what approval, and can it be proven? A process digitized end to end answers them natively.
- Audit trail: every entry, change and approval is recorded with its author and date (ALCOA+, 21 CFR Part 11, Annex 11).
- Access management: everyone sees and changes what their role allows (ISO 27001, GDPR).
- Information entered once: no more re-keying between tools, so fewer errors and consistent data.
- Versions and signatures: documents have a reference version and traced electronic approvals.
- Retention and reversibility: retention periods are enforced by the system, not by people’s memory.
- Evidence ready: audit files are prepared in hours, from the processes’ own data.
AI agents fit into this framework: they prepare, check and follow up, but their proposals go through human approval, and their actions leave the same trail as the teams’.
Terminus 2π helps you bring your processes into compliance; certifications are still issued by accredited bodies.
Chapter 6
Choosing hosting to fit your constraints
There is no compliant hosting in the absolute, only hosting suited to your data and your obligations. Health data of French patients requires an HDS-certified host; clinical trial data adds GCP requirements; some organizations require their data to stay on-premises.
Liquid instances are deployed with global hosting providers (Google Cloud, AWS, Scaleway, OVH) or on-premises for organizations that prefer it. In our engagements, we help you choose the hosting that meets your compliance constraints: data location, certifications of the host and its subcontractors, reversibility, validation documentation.
Chapter 7
Where to start
Three steps are enough to turn compliance into an advantage:
- Map your processes and identify the frameworks that apply to each.
- Prioritize the processes where compliance costs the most: re-keying, manual evidence, audits prepared in a rush.
- Digitize these processes end to end, so traceability becomes automatic.
The Express Process Diagnostic does exactly that in ten days: a map of your key processes, a measure of your time, quality and cost losses, and a costed action plan.
Let’s talk about your constraints
Terminus 2π · Digital process first. AI at the core.
Website: terminus2pi.com · E-mail: [email protected]
LinkedIn: linkedin.com/company/terminus-2pi
This guide is provided for reference and is not legal or regulatory advice: check your situation with a qualified regulatory affairs consultant. Framework matrix and tensions after a synthesis published by Gencovery in 2025, reviewed and extended by Terminus 2π.