Practical Benefit–Risk Assessment in Pharmacovigilance From Safety Signals to Regulatory Decisions

Every periodic safety report ends with the same sentence: the benefit–risk balance remains favourable. Regulatory guidance is excellent at describing what that conclusion must look like on the page. ICH E2C(R2) tells you where the integrated analysis goes. ICH M4E(R2) sets out the subsection structure. GVP Module V defines which risks belong in the safety specification. The FDA framework gives you a grid and explains what belongs in each cell.

None of them tells you how to reason your way to the conclusion. The templates are excellent; the reasoning is left as an exercise — and the exercise is the whole job.

This book teaches the reasoning, method by method, and commits to a single worked case rather than a gallery of illustrations. Therapex — an invented oral kinase inhibitor for rheumatoid arthritis — is authorised in Year 0, accumulates a thromboembolic signal by Year 3, and reaches a regulatory decision in Year 4. Every method chapter runs its method on that same data set. When the effects table, the PrOACT-URL analysis, the BRAT display, the numbers needed to treat and harm, the multi-criteria model, the probabilistic re-run and the patient preference study all rest on identical inputs, the differences between them belong to the methods, not to the examples.

It also delivers the moment the field is quietest about: the point at which the same data, in the same model, return a favourable conclusion for the population as a whole and an unfavourable one for an identifiable subgroup — and someone has to decide what follows.

Inside

Foundations — benefit versus efficacy, risk versus adverse event, balance versus profile, and the four conditions every benefit–risk statement silently carries.

Building the evidence base — framing the decision problem, measuring benefits and risks, choosing sources, and handling uncertainty honestly.

Methods — nine approaches, each taught to the point of execution: structured effects tables, PrOACT-URL, BRAT, NNT and NNH, difference and population-impact measures, multi-criteria decision analysis with swing weighting, probabilistic and Bayesian methods, Q-TWiST, and patient preference studies. Every chapter follows one spine: concept, mathematics, worked example, presentation, interpretation, limitations, and how to say it.

Day-to-day practice — signal management, the integrated analysis in the PBRER, risk management plans, safety governance, reassessment after a new serious signal, and the full ladder of regulatory action.

Special problems — oncology, vaccines, rare disease, pregnancy, paediatrics and older adults, accelerated approval, and subpopulations.

Ten worked case studies and a practical toolkit — templates, worksheets, checklists, and a conclusion-wording library with copy-ready wording for every conclusion type, from favourable and unchanged through to no longer favourable.

Written for pharmacovigilance physicians and safety scientists, benefit–risk leads, pharmacoepidemiologists, medical writers preparing PBRERs and risk management plans, regulatory affairs professionals, biostatisticians working on safety, regulatory assessors, and graduate students in pharmacovigilance and regulatory science. Readers are assumed to know what an individual case safety report is; the book spends its space on the reasoning, not the primer.

Current to August 2026, including the CIOMS Working Group XII consensus report on benefit–risk balance (2025), the FDA's final benefit–risk guidance (2023) and ICH E22 on patient preference studies. An interactive companion workbench implements the calculators and worksheets.

Foreword by Lisa Beth Ferstenberg, MD.