Free, always
A glossary, bite-sized quick reads on the field and on interviewing, and a hub for the free certifications worth pursuing. No signup required for any of it.
Under a minute each
PV and clinical research bits, interview tips, and quick facts — hand-curated, not an auto-generated feed.
The EMA has been revising its Good Vigilance Practice modules, with Module IX (signal management) getting the most attention. The direction of travel is tighter integration between EudraVigilance data and each marketing authorisation holder’s own internal signal detection process — rather than treating the central database and a company’s own case data as two separate analytical exercises that happen to look at overlapping information.
If an interviewer asks whether a case is "related," don’t just pick a category — walk through your reasoning out loud: the temporal relationship between drug exposure and event onset, dechallenge/rechallenge information if it’s available, alternative explanations (other medications, underlying conditions, disease progression), and which causality method you’d actually apply — WHO-UMC or Naranjo — and why that one fits this case.
MedDRA releases a new version every March and September, on a fixed schedule, rather than pushing updates continuously as new terms are proposed and approved. That fixed cadence is deliberate, not a technical limitation — it exists specifically so companies and regulators can plan validation and re-coding work around a predictable calendar, rather than chasing a constantly moving target that could shift underneath an in-progress submission or audit.
A number with no reasoning behind it invites negotiation from a weak position — it reads as either a guess or a wish, and either way it hands the interviewer the initiative to push back without you having anywhere solid to push from. A range tied to explicit reasoning does the opposite: it signals you’ve actually researched the role and know your own market value, which is a different, stronger conversation to be in.
A growing share of day-to-day case processing now sits with CROs and IT/consulting firms running pharmacovigilance operations on behalf of pharma sponsors, rather than sitting inside the sponsor company’s own in-house safety team. This isn’t a niche trend confined to smaller sponsors either — even large pharmaceutical companies increasingly outsource significant portions of their case-processing volume to specialized vendors, keeping strategic oversight (signal evaluation, regulatory strategy, QPPV accountability) in-house while the high-volume operational work moves outward.
Adverse event reporting for a newly launched drug typically follows a predictable arc: it rises after launch, peaks around the drug’s second year on the market, and then gradually declines from there — and critically, this pattern shows up in the reporting data even when the drug’s actual underlying safety profile hasn’t changed at all. It’s not that the drug got safer over time. It’s that reporting enthusiasm fades: media attention around a new drug’s launch drops off, prescribers and patients become less novelty-motivated to report events for a medication that’s no longer new, and the sheer volume of routine, unremarkable use starts to dwarf the early period when every early adverse event felt more notable.
Most behavioral interview questions — tell me about a conflict, a mistake, a time you were under pressure, a disagreement with a colleague — can be answered convincingly using the same one or two well-chosen real work stories, simply reframed for whatever specific angle the question is asking about. Trying to have five or six completely different, distinct stories ready for every possible behavioral question is both harder to prepare and, counterintuitively, produces weaker answers than going deep on fewer.
Every SUSAR is serious, unexpected, and suspected to be related to the drug — all three conditions have to be true simultaneously for a case to actually qualify as a SUSAR (Suspected Unexpected Serious Adverse Reaction). Drop any one of the three and the classification changes, along with the reporting obligations that follow from it.
Most large safety teams now use some form of AI or NLP tool to triage medical literature for potential adverse event reports before a human reviewer looks at it — screening thousands of new publications for anything that might contain a reportable case is exactly the kind of high-volume, pattern-matching task these tools handle well, freeing up human reviewers to focus on the genuinely ambiguous cases that actually need clinical judgment.
A sharper closing question signals genuine engagement in a way "what’s the culture like?" simply doesn’t anymore — it’s become such a default, low-effort closing question that it reads as filler rather than real curiosity, even when the person asking it genuinely does want to know. The fix isn’t to skip asking a question (never asking anything at the end reads worse), it’s to ask one specific enough that it could only come from someone who’s actually been thinking seriously about the role.
Plain-language reference
ICSR — Individual Case Safety Report
A structured report of a single adverse event experienced by a single patient with a single suspect drug — the basic unit of pharmacovigilance case processing.
SUSAR — Suspected Unexpected Serious Adverse Reaction
An adverse reaction that is serious, not listed in the reference safety information, and suspected to be related to the drug — all three conditions must be true. Triggers 7-day (fatal/life-threatening) or 15-day expedited reporting.
SAE — Serious Adverse Event
An adverse event meeting at least one seriousness criterion: death, life-threatening, hospitalization/prolonged hospitalization, persistent/significant disability, congenital anomaly, or another medically important condition.
ADR — Adverse Drug Reaction
A response to a drug that is noxious and unintended — unlike a general adverse event, an ADR implies at least a reasonable possibility of a causal relationship to the drug.
AE — Adverse Event
Any untoward medical occurrence in a patient given a drug, whether or not it is considered related to the drug — the broadest term in the field.
Narrative —
The written summary within an ICSR that tells the clinical story of the case chronologically — onset, treatment, outcome — in prose rather than structured fields.
Follow-up —
Additional information received after an initial case report, used to complete missing fields or update the case (e.g. an outcome that was previously unknown).
Minimum Criteria —
The four elements required for a report to count as a valid ICSR: an identifiable patient, an identifiable reporter, a suspect drug, and an adverse event.
Dechallenge/Rechallenge —
Dechallenge is stopping the drug to see if the reaction resolves; rechallenge is restarting it to see if the reaction recurs — both are strong (though not definitive) causality evidence.
Medication Error —
A preventable mistake in prescribing, dispensing, or administering a drug — only becomes a reportable adverse event if it actually results in patient harm or a special-situation report.
Case Processing —
The end-to-end workflow of collecting, validating, coding, narrativizing, and reporting an ICSR from source document to regulatory submission — increasingly AI-assisted at individual steps, always with human validation.
Duplicate Detection —
Identifying and linking multiple reports of the same event in the same patient from different sources (patient report + HCP report + literature, for example) so they aren’t counted as separate cases.
MIE — Medically Important Event
An event that doesn’t immediately threaten life or require hospitalization but still jeopardizes the patient or needs medical intervention to prevent a serious outcome — qualifies as serious under ICH E2A and requires clinical judgment to assess.
Batch Reporting —
Submitting multiple ICSRs together in bulk rather than one at a time — a common submission pattern to EudraVigilance and FAERS.
Causality Assessment —
Evaluating whether a drug could plausibly have contributed to an event, weighing time-to-onset, dechallenge/rechallenge, alternative explanations, and supporting literature.
Confounding Factors —
Other diseases, concomitant medications, or lifestyle factors that could explain an event independent of the suspect drug — something causality assessment has to rule in or out.
Consumer Reporter —
A non-healthcare-professional reporter — a patient or caregiver, typically — whose report may need medical confirmation before it can be fully interpreted clinically.
IDR — Initial Data Receipt Date
The earliest date the company actually receives information about a case — the anchor date used to calculate reporting-clock timelines.
Indication —
The reason the drug was prescribed in the first place — a required structured field in every case.
Key Event —
The clinically significant component of an adverse event that most often drives its escalation as a serious case.
Lab Value of Interest —
A laboratory parameter relevant to evaluating the event — ALT, creatinine, or QTc interval, for example — flagged for closer review during triage.
Medically Significant Condition —
An event that doesn’t neatly fit the standard seriousness criteria but still requires medical intervention — allergic bronchospasm treated at home is the classic example.
Outcome of Event —
The resolution status of a case — recovered, recovering, not recovered, fatal, or unknown — one of the required structured fields.
QC — Quality Check
Verification of a case’s accuracy and completeness before it moves forward — the step CROs and sponsors track hardest, since quality-score KPIs (commonly a 95%+ target) are built directly on QC results.
Time-to-Onset — TTO
The gap between starting the drug and the event’s onset — one of the key inputs into causality assessment.
Triage —
The initial sorting step where a case’s seriousness, priority, and reporting timeline are assigned — sets everything downstream in motion.
WHO-UMC Causality Categories —
The six-category scale used to classify how likely a drug-event relationship is: certain, probable/likely, possible, unlikely, conditional/unclassified, or unassessable/unclassifiable.
MedDRA — Medical Dictionary for Regulatory Activities
The standardized medical terminology used to code adverse events, from broad System Organ Classes down to specific Lowest Level Terms, updated twice yearly (March and September) by ICH MSSO.
SOC — System Organ Class
The highest, broadest level of the MedDRA hierarchy — e.g. "Gastrointestinal disorders" — used mainly for high-level data presentation, not day-to-day coding.
PT — Preferred Term
The MedDRA level actually used for most analysis and reporting — a single, distinct medical concept like "Nausea" or "Hepatic failure."
LLT — Lowest Level Term
The most granular MedDRA level, closest to how a reporter actually describes a symptom in their own words — each LLT maps up to exactly one PT.
SMQ — Standardised MedDRA Query
A pre-built, validated grouping of MedDRA terms representing a medical condition of interest (e.g. anaphylaxis) — used to pull all relevant cases together for aggregate analysis, since no single PT would catch them all.
Verbatim —
The reporter’s original, unedited description of the event, before it’s translated into a formal MedDRA term — the coder’s actual starting point.
ATC Classification — Anatomical Therapeutic Chemical Classification
A five-level system assigning every pharmaceutical substance a hierarchy based on the body system it acts on, its therapeutic indication, pharmacological action, and chemical structure — used for class-effect signal detection.
Overcoding —
Incorrectly assigning a more severe or specific MedDRA term than the underlying verbatim actually supports — a common QC finding.
Signal —
Information suggesting a new, potentially causal association (or a new aspect of a known one) between a drug and an event, sufficiently likely to warrant further investigation.
PRR — Proportional Reporting Ratio
A disproportionality statistic comparing how often a specific event is reported for a specific drug versus all other drugs in a database — a common first-pass signal detection metric. Common threshold: PRR ≥ 2, chi-squared ≥ 4, n ≥ 3 cases.
ROR — Reporting Odds Ratio
A disproportionality measure similar to PRR but calculated as an odds ratio rather than a ratio of proportions — commonly used alongside or instead of PRR, including in EudraVigilance analyses.
EBGM — Empirical Bayes Geometric Mean
A Bayesian disproportionality measure that shrinks extreme values toward the mean, making it more conservative than PRR when case counts are small.
IC — Information Component
The Bayesian disproportionality measure behind WHO-UMC’s VigiBase signal detection — expressed as log2 of the ratio of observed to expected reports under independence; a signal is flagged when IC is positive and its lower 95% credibility bound (IC025) exceeds zero.
Weber Effect —
The tendency for adverse event reporting on a newly launched drug to peak around its second year on the market and then decline — a known reporting artifact, not a real safety change.
Signal Validation —
The step after detection where a signal is reviewed for clinical plausibility and data quality before deciding whether it warrants a full assessment — filters noise from real candidates.
Bayesian Analysis —
A statistical approach that combines prior knowledge with current data to estimate the strength of a causal association or signal — the basis of IC at WHO-UMC and EBGM at FDA FAERS.
Bradford Hill Criteria —
Nine considerations (strength, consistency, specificity, temporality, biological gradient, plausibility, coherence, experiment, analogy) used to judge whether an association is likely causal rather than coincidental.
BRAT Framework — Benefit-Risk Action Team Framework
A structured, visual method (developed by TransCelerate) for comparing benefit and risk criteria side by side — widely used in PBRER Section 16 for complex products.
Biological Plausibility —
Whether a known mechanism of action or pharmacological pathway could plausibly explain the observed adverse event — one input into causality and signal assessment.
Literature Screening —
Systematic, typically weekly review of scientific publications to catch reportable cases or emerging safety signals that spontaneous reporting alone would miss.
NNH — Number Needed to Harm
The number of patients who’d need to be treated, on average, before one additional harmful event occurs — a way of expressing risk magnitude in practical terms.
NNT — Number Needed to Treat
The number of patients who’d need to be treated, on average, before one additional beneficial outcome occurs — the benefit-side counterpart to NNH.
PSUR — Periodic Safety Update Report
A periodic aggregate safety report on a marketed product, largely superseded in the EU by the PBRER but still used in some regions.
PBRER — Periodic Benefit-Risk Evaluation Report
The ICH E2C(R2) evolution of the PSUR — a cumulative aggregate report focused explicitly on ongoing benefit-risk balance, covering all safety data since the International Birth Date.
DSUR — Development Safety Update Report
The clinical-trial-phase equivalent of a PSUR/PBRER — an annual aggregate safety report, governed by ICH E2F, covering an investigational drug still in development.
RMP — Risk Management Plan
A living document describing a product’s known risks, potential risks, missing information, and the specific activities planned to characterize and minimize them.
DHPC — Direct Healthcare Professional Communication
An urgent, regulator-mandated letter sent directly to prescribers/pharmacists to communicate an important new safety concern.
Expedited Report —
A report that must be submitted to regulators within a short fixed window (commonly 15 or 7 days) because it meets serious + unexpected + related criteria — distinct from routine periodic reporting.
CIOMS Form —
A standardized international case report form format (from the Council for International Organizations of Medical Sciences), widely used for exchanging ICSRs between companies and regulators.
Data Lock Point — DLP
The cutoff date beyond which no new safety data is included in a specific aggregate report cycle — everything with an outcome date before it is captured, everything after waits for the next cycle.
PASS — Post-Authorisation Safety Study
A study conducted after marketing authorization to characterize risks identified in the RMP, validate signals, measure real-world risk incidence, or evaluate risk-minimization effectiveness — either regulator-mandated or company-initiated.
Benefit-Risk Assessment — BRA
The ongoing evaluation weighing a product’s therapeutic benefits against its potential risks — conducted at approval and continuously updated, formally presented in PBRER Section 16.
Risk Minimization Measures — RMMs
Concrete actions taken to reduce the occurrence or severity of a known risk — educational materials, restricted-access programs, or monitoring requirements, for example.
Missing Information —
A gap in what’s known about a product’s safety — pregnancy or pediatric use data, for instance — documented in the RMP as an area needing further characterization.
Population Exposure —
An estimate of how many patients have actually used the product — the denominator that turns raw case counts into meaningful reporting rates.
GVP — Good Pharmacovigilance Practice
The EMA’s modular guidance defining operational standards for EU pharmacovigilance — sixteen modules spanning system quality (Module I) through signal management (Module IX) to risk-minimization effectiveness (Module XVI).
QPPV — Qualified Person for Pharmacovigilance
The individual a marketing authorisation holder designates as personally accountable for the pharmacovigilance system — an EU-resident role requiring 24-hour access to the safety database and PSMF.
PSMF — Pharmacovigilance System Master File
A detailed description of a company’s entire pharmacovigilance system — organization, processes, quality system — maintained and inspectable at any time, available to regulators within one business day on request.
CCDS — Company Core Data Sheet
A company’s own master safety and prescribing reference for a product, from which local product labels around the world are derived, and against which expectedness/listedness is judged.
RSI — Reference Safety Information
The document (often the CCDS or investigator’s brochure) used to determine whether an event is "expected" — the baseline everything else is compared against.
Expectedness — Listedness
Whether an adverse event is already listed in the reference safety information for the drug — "unexpected" events are what trigger expedited reporting, not just "serious" ones alone.
ALCOA-C — Attributable, Legible, Contemporaneous, Original, Accurate, Complete
The data integrity principles that any regulated PV documentation — case data, audit trails, SOPs — is expected to satisfy.
CAPA — Corrective and Preventive Action
The formal process for investigating a quality deviation, fixing the immediate issue, and preventing it from recurring — standard across regulated industries, not unique to PV.
GxP — Good "x" Practice
The umbrella term for the family of regulated-industry quality standards (GVP, GCP, GMP, GLP) — the "x" stands in for whichever practice area applies.
Audit Trail —
A sequential, unalterable, time-stamped record of every action taken in a safety database — who did what, when, and what changed — required by GVP Module I, 21 CFR Part 11, and EU Annex 11, and can’t be edited or deleted.
CIOMS — Council for International Organizations of Medical Sciences
The international body whose Working Groups (VI through XIV) produce harmonized PV guidance covering everything from signal management to, most recently, AI governance in PV.
De-identification —
Removing or irreversibly masking personal identifiers from a dataset to protect patient privacy while preserving its analytical value — required under GDPR, India’s DPDP Act, and HIPAA for secondary use of patient data.
EMA — European Medicines Agency
The EU regulatory body that maintains the GVP Modules, coordinates PRAC, and manages EudraVigilance.
EU AI Act — Regulation (EU) 2024/1689
The world’s first comprehensive AI regulation. Classifies PV signal-detection and case-processing AI as high-risk, requiring risk-management and technical documentation, human-oversight mechanisms, accuracy standards, and EU AI database registration.
EU Regulation 2025/1466 —
The most significant EU PV rule change since 2012 (applicable February 2026): abolishes standalone signal notifications to EMA, mandates risk-based subcontractor auditing, requires PSMFs to list only open major/critical deviations, makes EudraVigilance monitoring a legal obligation, and requires PBRERs to include risk-minimization effectiveness data.
PVQS — Pharmacovigilance Quality System
The documented quality-management framework governing all PV activities in an organization — SOPs, training, deviation management, CAPA, a validated safety database, audit programs, and QPPV oversight, all required under GVP Module I.
GCP — Good Clinical Practice
The international ethical and scientific quality standard for designing, conducting, and reporting clinical trials involving human subjects.
IB — Investigator’s Brochure
A compiled summary of all known clinical and non-clinical data on an investigational drug, given to trial investigators — often doubles as the reference safety information during a trial.
Protocol Deviation —
Any departure from the approved clinical trial protocol — ranges from minor and inconsequential to serious enough to affect patient safety or data integrity.
IND — Investigational New Drug (application)
The FDA submission required before a drug can be tested in human clinical trials in the US.
NDA — New Drug Application
The FDA submission requesting approval to market a new drug in the US, built on the full clinical trial data package.
Blinded/Unblinded Case —
In a blinded trial, treatment allocation is concealed from some or all parties; a safety event can be unblinded for medical assessment when clinically necessary, following a defined procedure.
Efficacy vs. Effectiveness —
Efficacy is what’s observed under the controlled conditions of a clinical trial; effectiveness is what actually happens once the drug is used in everyday real-world practice.
E2B(R3) — ICH E2B Revision 3
The current international electronic data-exchange standard (an XML structure with controlled vocabulary) for transmitting ICSRs between safety databases and regulators — implemented across EudraVigilance, FAERS, VigiBase, and PvPI.
EudraVigilance —
The EU’s central system for managing and analyzing suspected adverse reaction reports for authorised medicines — mandatory submission gateway for EU MAHs, with its EVDAS module handling PRR/ROR signal analytics.
FAERS — FDA Adverse Event Reporting System
The FDA’s database of spontaneous adverse event and medication error reports, processing millions annually, with signal detection run via MGPS/EBGM Bayesian methods.
VigiBase —
The WHO’s global ICSR database, maintained by the Uppsala Monitoring Centre and holding tens of millions of reports from 150+ countries — analyzed through VigiLyze using IC disproportionality.
Argus Safety —
One of the most widely used commercial safety database platforms (Oracle) for end-to-end ICSR case processing in the pharma/CRO industry.
WHO-DD — WHO Drug Dictionary
A standardized reference of drug names and ingredients, maintained by WHO-UMC and updated quarterly, used alongside MedDRA so the "what drug" side of a case is coded as consistently as the "what happened" side.
EHR — Electronic Health Record
A digital repository of longitudinal patient data from routine clinical care — a primary real-world-data source for pharmacoepidemiology and PASS studies, usually requiring NLP to extract adverse events from unstructured notes.
FHIR — Fast Healthcare Interoperability Resources
An HL7 data-exchange standard increasingly used to connect EHR systems with E2B(R3) case submission pipelines.
IDMP — Identification of Medicinal Products
Five ISO standards providing a globally unique identifier for medicinal products and their components — required in E2B(R3) submissions to EudraVigilance, with EMA’s SPOR implementation still ongoing.
OMOP CDM — Observational Medical Outcomes Partnership Common Data Model
A common table structure and standard vocabulary set (RxNorm, SNOMED CT, LOINC) that lets heterogeneous real-world data sources be analyzed consistently — the foundation of the DARWIN EU and Sentinel networks.
DARWIN EU —
EMA’s federated real-world-evidence network connecting healthcare databases across EU member states on the OMOP CDM, feeding regulatory-grade studies into PRAC signal assessments.
Sentinel System —
FDA’s active-surveillance infrastructure — a federated network of US claims and EHR data covering 100+ million lives, used for near-real-time monitoring of selected product classes.
RWD — Real-World Data
Health data collected outside a traditional randomized trial — EHRs, insurance claims, patient registries, wearables, pharmacy dispensing records — providing population-level context spontaneous reports can’t.
RWE — Real-World Evidence
Clinical evidence about a drug’s usage, benefits, or risks derived from analyzing RWD — with growing regulatory acceptance through networks like DARWIN EU, FDA Sentinel, and CDSCO’s National RWE Hub.
Cloud PV System —
A cloud-hosted safety database platform (Veeva Vault Safety, Oracle Argus Cloud, ArisGlobal LifeSphere, for example) whose architecture enables continuous AI integration and elastic scaling for high case volumes.
VigiGrade —
WHO-UMC’s automated completeness score for ICSRs in VigiBase — checks for key data elements like age, sex, dose, and time-to-onset, and flags low-scoring cases for follow-up.
Human-in-the-Loop — HITL
The design principle requiring a trained human to validate an AI output before any regulatory- or clinical-consequence action is taken on it — mandatory under CIOMS WG XIV, the EU AI Act, and GVP Module IX Rev 2.
Agentic AI —
An AI system that’s given a goal and autonomously plans and executes a multi-step workflow using available tools, without a human in the loop between steps — examples include Veeva Vault Safety AI Agents and ArisGlobal NavaX. Human checkpoints are still mandatory in regulated use.
NLP — Natural Language Processing
AI methods for extracting meaning from text — used in PV for entity extraction from AE emails/PDFs, MedDRA code suggestion from verbatim, literature screening, and narrative-similarity duplicate detection. The most widely deployed AI type in production PV systems today.
GenAI — Generative AI
AI capable of producing coherent text from structured or semi-structured inputs — used in PV for case narrative drafting and PBRER executive summaries, always with mandatory physician review before submission.
CIOMS WG XIV —
CIOMS Working Group XIV on Artificial Intelligence in Pharmacovigilance, whose December 2025 final report is the current authoritative international AI-governance framework for PV — built on seven principles: transparency, validation, human oversight, bias testing, audit trails, accountability, and ongoing monitoring.
Algorithmic Bias —
Systematic error in an AI model’s output that produces unequal performance across demographic, geographic, or clinical subgroups, typically from underrepresentation in training data — testing and mitigation is required under CIOMS WG XIV and the EU AI Act.
Explainable AI — XAI
The principle that an AI model’s decisions must be interpretable in human terms — required for regulated PV use under CIOMS WG XIV and the EU AI Act, via techniques like SHAP values and counterfactual explanations.
Model Drift —
The gradual decline in an AI model’s performance as real-world data shifts away from what it was trained on — requires continuous monitoring, and revalidation once drift crosses a defined threshold.
Confidence Score —
An AI system’s self-assessed probability that its own output is correct — used in AI-assisted coding, seriousness classification, and duplicate detection, but a high score is never a guarantee; low scores trigger mandatory elevated human review.
Failure Mode (AI) —
A way an AI system can go wrong in practice — hallucinating detail, misclassifying an event, or missing one entirely — the kind of risk human review exists to catch.
Federated Learning —
A machine-learning approach where a model trains across multiple data sources without any raw patient data ever leaving its source — only aggregate model updates are shared. WHO-UMC’s AI Hub is the leading PV implementation.
FDA-EMA Joint AI Principles —
Ten Guiding Principles of Good AI Practice in Drug Development, published jointly by FDA and EMA in January 2026 — the first transatlantic AI governance statement covering pharmaceutical development, including PV. Not legally binding, but sets the practical benchmark.
NavaX —
ArisGlobal LifeSphere’s AI-driven case processing platform, built around an Intelligence Agent (case processing), Distribution Agent (multi-jurisdiction submission), and Signals Agent (continuous signal detection).
WHO-UMC AI Hub —
WHO-UMC’s initiative for trustworthy, federated AI in global PV — developing signal-detection algorithms deployable across national PV databases without moving raw patient data.
CDSCO — Central Drugs Standard Control Organization
India’s national drug and cosmetics regulatory authority — oversees pharmacovigilance through PvPI, coordinates with the Indian Pharmacopoeia Commission, and conducts PV inspections under the Drugs and Cosmetics Act.
PvPI — Pharmacovigilance Programme of India
India’s national pharmacovigilance initiative, run by the Indian Pharmacopoeia Commission (IPC) as National Coordination Centre with a network of 1,000+ ADR Monitoring Centres, reporting to CDSCO and linked to VigiBase through WHO-UMC.
Vigiflow —
The case-management system PvPI uses to submit India’s ICSRs into VigiBase.
DPDP Act 2023 — Digital Personal Data Protection Act
India’s data protection law governing how personal data — including patient health data used in PV — is processed, with public-interest exceptions for health research and requirements around consent, minimization, and cross-border transfer.
GPSP — Good Post-Marketing Study Practice
Japan’s rigorous post-marketing safety surveillance ordinance, overseen by the PMDA.
JADER — Japan Adverse Drug Event Report database
Japan’s national spontaneous adverse drug event reporting database.
Want calculators, not just definitions?
Naranjo, WHO-UMC, PRR/ROR, and an AI narrative grader — the PV Toolkit.
Start building real credentials
ICH E6(R3)-aligned Good Clinical Practice — the current 11-principle framework, not outdated E6(R2) material.
Pharmacovigilance fundamentals, signal detection, causality assessment, and statistical reasoning — from the organisation that created VigiBase and the WHO-UMC causality scale.
MedDRA’s 5-level hierarchy and coding conventions, directly from the organisation that maintains the dictionary.
GVP Modules and benefit-risk frameworks, hosted directly on the European Medicines Agency website — developed via the SCOPE Joint Action to strengthen national PV systems.