Pharmacovigilance interviews are deceptively technical. On the surface, the role sounds process-driven โ collect adverse events, submit reports, monitor signals. In practice, interviewers at AstraZeneca, Roche, ICON, and Covance want to see that you can apply regulatory judgment under time pressure, code cases accurately, and make defensible benefit-risk decisions.
Here's what they actually test.
The Core Domains in Every PV Interview
Every pharmacovigilance interview, whether for a drug safety associate, PV specialist, or safety physician role, tests the same domains:
- ICSR processing: Case intake, narrative writing, seriousness/expectedness/causality assessment
- Regulatory reporting timelines: 7-day, 15-day, and periodic reporting requirements
- Aggregate safety reports: PSUR, PBRER, DSUR, annual reports
- Signal detection: Disproportionality analysis, data mining, signal evaluation workflow
- Safety databases: Argus, ARISg, Oracle Empirica Signal
- MedDRA coding: SOC, HLGT, HLT, PT, LLT hierarchy โ coding strategy
- Risk management: REMS, EU Risk Management Plan (RMP), PASS, PAES
- ICH guidelines: E2A through E2F โ know them by name and scope
- Benefit-risk assessment: Framework, CIOMS tools, label updates
ICSR Processing: The Most-Tested Area
Individual Case Safety Reports are the operational core of PV. Interviewers probe three things:
1. Seriousness assessment A case is serious if it meets at least one criterion: death, life-threatening, hospitalization (or prolongation), disability/incapacity, congenital anomaly, or medically important event (MIE). Know ICH E2A's definition cold.
Common question: "A patient reports fatigue and nausea after starting your drug. Is this a serious adverse event?" The answer is almost always no โ but the follow-up is: "What if the fatigue led to a fall and fracture?" Now you have hospitalization.
2. Expectedness assessment Expected = listed in the current Company Core Data Sheet (CCDS) or product labeling with comparable nature, severity, and frequency. Unexpected = unlisted or listed with lesser severity.
Common question: "You receive a case of Stevens-Johnson Syndrome for a drug where the label lists 'rash' as an ADR. Is SJS expected?" No โ SJS is a distinct, severe reaction; 'rash' does not capture it.
3. Causality assessment Know the WHO causality categories: Certain, Probable/Likely, Possible, Unlikely, Unassessable/Unclassifiable, Not Related. Also know company-level causality โ most sponsors use "related / not related" with a narrative justification.
Regulatory Reporting Timelines
These questions are asked at virtually every PV interview:
- 7-day expedited report: Unexpected, fatal or life-threatening SAE from a clinical trial (IND safety report to FDA)
- 15-day expedited report: Unexpected serious adverse drug reaction (SUSAR) โ non-fatal/non-life-threatening โ or any serious expected ADR that meets expedited criteria in the EU
- CIOMS I form: Used for expedited reporting internationally; know its structure
- MedWatch 3500A: FDA's form for spontaneous post-marketing reports
- Periodic reports:
- PSUR (Periodic Safety Update Report) โ EU requirement, covers a defined data lock point period
- PBRER (Periodic Benefit-Risk Evaluation Report) โ ICH E2C(R2) aligned, includes benefit-risk evaluation; now the global standard replacing PSUR
- DSUR (Development Safety Update Report) โ ICH E2F, for investigational products in clinical trials; filed annually
Common question: "What's the difference between a PSUR and a PBRER?" PBRER supersedes PSUR under ICH E2C(R2) โ it includes a structured benefit-risk framework and is internationally harmonized. The EU still uses the term PSUR colloquially but expects PBRER content.
ICH Guidelines You Must Know
The E2 series is the backbone of PV practice:
| Guideline | Topic | |-----------|-------| | ICH E2A | Clinical safety data management โ definitions and reporting standards | | ICH E2B(R3) | Electronic transmission of ICSRs (XML format, E2B fields) | | ICH E2C(R2) | PBRER structure and content | | ICH E2D | Post-approval safety data management | | ICH E2E | Pharmacovigilance planning (including risk management) | | ICH E2F | Development Safety Update Report |
Know E2A definitions by heart. Know E2B(R3) because database work and ICSR transmission are tested at CROs and sponsor companies.
Signal Detection Questions
Signal detection separates operational PV from strategic PV. Interviewers at companies with large marketed portfolios (Roche, AstraZeneca) dig into this:
Disproportionality analysis:
- Proportional Reporting Ratio (PRR): cases of drug-event pair / all cases with that event for the drug, divided by the same ratio for all other drugs. PRR โฅ 2 with โฅ 3 cases and chi-squared โฅ 4 = signal.
- Reporting Odds Ratio (ROR): similar concept, logistic model-based
- EBGM (Empirical Bayes Geometric Mean): used in FAERS mining, less sensitive to small counts
Common question: "You run disproportionality analysis on FAERS and find a PRR of 3.8 with 12 cases for a drug-event pair not in your label. What do you do next?" Expected answer: validate the signal โ review the cases for quality, assess causality, check literature, evaluate biological plausibility, convene the signal review team, potentially update labeling or initiate a PASS.
FAERS (FDA Adverse Event Reporting System):
- Know how to access it (openFDA API, FAERS public dashboard)
- Know its limitations: under-reporting, no denominator, confounding by indication, duplicate cases
- Know when FAERS signals are actionable vs noise
MedDRA Coding
MedDRA (Medical Dictionary for Regulatory Activities) coding is tested operationally:
- Hierarchy: System Organ Class (SOC) โ High Level Group Term (HLGT) โ High Level Term (HLT) โ Preferred Term (PT) โ Lowest Level Term (LLT)
- Coding rule: Code to the PT that best represents the reported term โ do not under-code or over-code
- Multiple coding: One event can be coded to multiple PTs if the reporter described distinct components
- SMQs (Standardized MedDRA Queries): Predefined baskets of PTs for signal detection (e.g., SMQ for anaphylaxis, SMQ for drug-induced liver injury)
Common question: "A reporter says the patient had 'an allergic reaction with throat swelling and difficulty breathing.' How do you code this?" Anaphylaxis PT is appropriate; you might also code Throat tightness and Dyspnoea separately if the narrative supports distinct symptoms. Do not collapse everything into one PT if distinct events are described.
Safety Databases
Oracle Argus Safety: Industry standard for ICSR management at large pharma. Know the case workflow โ intake, processing, medical review, expedited submission, periodic report generation. If you've used it, be specific about version and modules.
ARISg (Cluepoints): Used at AstraZeneca and others. Different UI but same ICSR workflow logic.
Oracle Empirica Signal: Signal detection module; integrates with FAERS and internal data.
If you haven't used these systems, say so directly โ then demonstrate that you understand the workflow they support. Interviewers respect honesty more than bluffing.
Risk Management Questions
REMS (Risk Evaluation and Mitigation Strategy):
- FDA tool for drugs with serious safety concerns
- Can include: Medication Guide only; Medication Guide + Communication Plan; ETASU (e.g., certification programs, restricted dispensing, patient monitoring)
- Example: isotretinoin โ iPLEDGE REMS (pregnancy prevention); clozapine โ CLOZAPINE REMS (ANC monitoring)
EU Risk Management Plan (RMP):
- Required for all new marketing authorizations in the EU
- Contains: Safety specification, Pharmacovigilance plan, Risk minimization measures
- PASS (Post-Authorization Safety Study): required when additional safety data needed
- PAES (Post-Authorization Efficacy Study): less common, for conditional approvals
Common question: "What's the EU equivalent of a REMS?" The RMP with Additional Risk Minimization Measures (ARMM) โ but note they are not equivalent structures; REMS is product-specific and often more prescriptive.
Benefit-Risk Assessment
Senior roles and safety physician interviews heavily test benefit-risk framing:
- CIOMS Working Group VI framework: Structured frameworks for benefit-risk decision-making
- PrOACT-URL: Problem, Objectives, Alternatives, Consequences, Trade-offs, Uncertainty, Risk tolerance, Linked decisions โ used in some companies
- The key question: "How do you determine when a safety signal crosses the threshold to label update vs communication vs market withdrawal?"
Know that benefit-risk is comparative, context-dependent, and patient-population specific. A signal that warrants a black box warning in a general population drug may be acceptable in an oncology drug where the alternative is death.
Real Interview Questions by Company Type
Large pharma (AstraZeneca, Roche, Novartis):
- "Walk me through how you would handle a cluster of unexpected hepatotoxicity cases reported in the first year post-launch."
- "Your PBRER is due in 3 weeks and you've identified a potential new safety signal in aggregate data. How do you handle this?"
- "What's your experience with signal detection? What tools have you used and what signals have you escalated?"
CROs (ICON, Covance, IQVIA, PRA):
- "How do you ensure case processing timelines are met when you're managing multiple client accounts simultaneously?"
- "A sponsor disagrees with your causality assessment for a case. How do you handle that?"
- "You're onboarding a new client with a marketed product. What's your process for getting up to speed on their safety profile?"
Biotechs:
- "We don't have a PV system yet โ we're pre-approval. How would you set up a PV infrastructure from scratch?"
- "What would you include in our first PV plan for an IND-stage asset?"
- "How do you handle a serious unexpected adverse event in a Phase 1 study when you have no safety data for the compound?"
Behavioral Questions
PV is high-stakes and deadline-driven. Interviewers probe:
- "Tell me about a time you caught an error in a case that had already been submitted. What did you do?"
- "Describe a situation where you had to make a causality judgment with incomplete information."
- "Give me an example of a time you had a disagreement with a physician on a safety assessment. How did you resolve it?"
Prepare examples that show regulatory judgment, not just process compliance. The best answers describe a decision you made, not just a task you completed.
Preparation Strategy
Technical prep:
- Read ICH E2A and E2C(R2) in full โ they're short and dense with testable content
- Review FDA's published FAERS analysis case studies
- Know the current FAERS dashboard and how to interpret a PRR calculation
- Refresh MedDRA hierarchy with a current version of the coding manual
Company-specific prep:
- What therapeutic areas does this company work in? What are the key safety signals in those classes?
- What safety events have made news for their products? (Google "[company name] + safety + label update" for the last 2 years)
- Is this a CRO role (process-heavy) or sponsor role (strategy-heavy)?
Practice:
- Work through 5โ10 ICSR scenarios with seriousness/expectedness/causality assessments
- Time yourself on signal detection scenarios โ interviewers notice if you can think through a signal workflow in real time
PV interviews reward candidates who combine regulatory precision with clinical judgment. Practice under realistic conditions at CareerLift.ai.