Authorisation of a medicine does not mean its safety assessment is complete. Pre-authorisation studies are conducted in a limited number of participants, over a limited period, and usually under criteria that exclude comorbidities. Under those conditions, adverse effects that are rare, that emerge late, or that belong to particular subgroups cannot be detected.
Why post-authorisation surveillance is needed
A study of a few thousand people cannot reliably detect an adverse effect occurring in one in ten thousand. In real use, a medicine reaches a far larger and far more varied population: older people, those with renal or hepatic impairment, pregnancy, and people taking several medicines at once. Pharmacovigilance covers the collection and appraisal of safety signals arising in that wider use, and regulatory action where it is required.
Spontaneous reporting and its limits
The principal source for the system is adverse reaction reporting by healthcare professionals and patients.
The strength of spontaneous reporting is that it can catch an unexpected effect early. Its limits are these:
- Under-reporting — only a fraction of events that occur are ever reported.
- An unknown denominator — because the number of people using the medicine is not known, true frequency cannot be calculated.
- Reporting bias — rates rise for medicines in the news and for recently authorised products.
The number of reports is therefore not a measure of risk on its own; it generates a signal, it does not settle a question.
Assessing causality
Whether an event is related to a medicine cannot be determined from a single report. The elements weighed include the temporal relationship, whether the event resolved when the medicine was withdrawn, whether it recurred on rechallenge, the presence of alternative explanations, and consistency with a known pharmacological mechanism.
From signal to action
Once a safety signal is confirmed, the available measures are graduated: updating the product information, direct communication to healthcare professionals, restricting the conditions of use, requiring an additional surveillance study, and, in the most serious case, suspension of the authorisation. Each step should be proportionate to the strength of the evidence behind it.
Benefit and risk as a continuing judgement
The purpose of pharmacovigilance is not to reduce risk to zero; no effective medicine carries zero risk. The purpose is the continual re-assessment of the benefit-risk balance against current data. Detecting an adverse effect does not by itself require withdrawal: its frequency, its seriousness, whether it can be prevented, and the benefit the medicine provides are weighed together.
The value of a single report
One report seldom produces a conclusion on its own. Yet a signal only becomes visible once reports accumulate. Reporting an adverse reaction is therefore more than an administrative obligation: it is the data that makes the system work at all.
This is a general scientific information article and does not constitute individual medical advice. Decisions about medicines should be made with a clinician.