Local silence is not comfort: PV needs a playbook for foreign safety signals

The UK is acting on deaths that happened elsewhere

The most instructive item in this week's records is a short one. On 5 November the MHRA issued interim advice on Duvyzat (givinostat) after international reports of deaths in patients with Duchenne muscular dystrophy who were taking the drug. The regulator proposes more frequent ECG and electrolyte checks. It also states that no deaths have been reported in the UK.

Read that last sentence twice. A national regulator is proposing monitoring changes without a domestic fatality. The signal arrived from abroad, and the response is being built around it. The records do not say how many deaths were reported, in which countries, or whether givinostat caused them. They do not say what the independent review will conclude. The MHRA says implementation of the proposed measures depends on that review. Anyone claiming more than that is guessing.

Why this should change how PV teams work

Many pharmacovigilance systems are built around the case that lands in your own database. Local case counts drive local escalation, and local silence tends to read as reassurance. The Duvyzat notice shows that logic is incomplete. A QPPV whose product is on the UK or EU market may have no local deaths and still face a regulator's proposed risk-mitigation measures, and the questions that follow from them. Prescribers will ask what to do. Pharmacists will ask whether to monitor differently. Patients and families, in a rare disease community that talks to itself constantly, will ask what is happening.

The MHRA's choice of tool matters too. It is interim advice, tied to an ongoing review, and aimed at monitoring. It is not a withdrawal and not a conclusion about causality. That is a recognisable pattern: act proportionately and reversibly while the evidence is still being assessed. Teams that cannot move at that pace, because they have no agreed route from 'foreign report' to 'decision', will be reacting to the regulator rather than informing it.

The strongest counter-argument

The best objection is that acting early on unverified reports is its own risk. Death reports from abroad may lack denominators, clinical detail or any clear link to the drug. Extra ECG and electrolyte visits carry a burden for patients and clinics. Premature measures can also generate noise that later has to be walked back. There is a strong case for waiting for the independent review.

I accept the point about causality, and nothing here argues for assuming it. But the argument confuses two decisions. Whether the drug caused the deaths is for the review. Whether your organisation has a ready process for assessing the reports, pulling local data, drafting communications and deciding on proportionate interim steps is yours, and it cannot be improvised once the notice is public. Preparing is not the same as concluding.

What a foreign-signal playbook looks like

It is a short, owned procedure. It names the sources that count as triggers, including other regulators' safety communications, not only your own case flow. It sets who convenes the safety review and within what timeframe. It lists the data to pull straight away: local and global cases, relevant monitoring parameters, and the exposed population. It pre-clears the types of interim measures the company would support or recommend, and who must approve them. It also assigns the first draft of any healthcare professional communication, so that the first version is not written under pressure.

None of this needs new technology. It needs a decision, taken before the next notice arrives, about what 'no local cases' will and will not be allowed to mean in your organisation.