Agreement: I Agree Body: Dear Editor Cave and Harnden are right that passive pharmacovigilance is no longer adequate for modern medicines regulation.[1] The burden of adverse drug reactions makes this a patient safety problem, not merely a data-management inconvenience.[2] The unresolved question is how the proposed Health Data Research Service will convert faster access to data into earlier regulatory action. A national data service, as envisaged in the Sudlow Review,[3] could improve linkage, governance, and access. But linkage is not surveillance. If exposure definitions, outcome validation, denominator coverage, analytic choices, and escalation thresholds are settled only after a concern has arisen, “always on” monitoring may still produce signals that are delayed, inconsistent, or difficult to act on. The service therefore needs a public operating model for priority drug-event pairs, not only a technical front door to datasets. At minimum, this model should define data latency, outcome coding and validation, treatment switching, confounding by indication, thresholds for signal review, and the expected timeline for regulatory response. It should also specify when imperfect but concerning evidence is sufficient to justify strengthened warnings, targeted communication, or temporary restrictions while further evidence is generated. This distinction matters because patients are not protected by observational capacity alone. They are protected when a credible signal has a defined route to scrutiny and action. Proactive pharmacovigilance should therefore be judged not by whether datasets can be linked earlier, but by whether the pathway from signal detection to regulatory decision is predefined and auditable. References 1. Cave A, Harnden A. Proactive drug safety surveillance should be at the heart of the new Health Data Research Service. BMJ 2026;393:s908. doi:10.1136/bmj.s908. 2. Osanlou R, Walker L, Hughes DA, Burnside G, Pirmohamed M. Adverse drug reactions, multimorbidity and polypharmacy: a prospective analysis of 1 month of medical admissions. BMJ Open 2022;12:e055551. doi:10.1136/bmjopen-2021-055551. 3. Sudlow C. Uniting the UK’s Health Data: A Huge Opportunity for Society. The Sudlow Review. Health Data Research UK; 8 Nov 2024. No competing Interests: Yes The following competing Interests: Electronic Publication Date: Sunday, May 24, 2026 - 04:57 AI use: No, I have not used AI Highwire Comment Subject: Proactive drug safety surveillance should be at the heart of the new Health Data Research Service Workflow State: Released Full Title: Proactive pharmacovigilance needs predefined action rules Highwire Comment Response to: Proactive drug safety surveillance should be at the heart of the new Health Data Research Service Check this box if you would like your letter to appear anonymously:: Last Name: Nie First name and middle initial: Biao Email: biaonie2@163.com Address: Guangzhou, Guangdong Province Occupation: Physician Other Authors: Jinghong Liu, Cunli Affiliation: The First Affiliated Hospital of Jinan University BMJ: Additional Article Info: Rapid response
Proactive pharmacovigilance needs predefined action rules
