Typically, companies and regulators receive adverse event reports and maintain separate databases for them. Companies always then submit theirs to regulators, and so those cases are routinely and non-simultaneously duplicated on regulators’ databases. Moreover, in the US, adverse event reports can emanate separately from both clinicians, patients, or mere bystanders, so there is scope for duplication there to, to an extent that about a third of the entire regulators’ databases having originated with non-clinicians (this is unlike in most of Europe, where patients’ reports of serious adverse events (at least) are not accepted by regulators, although it is true for the yellow card scheme in the UK, to the extent of about a quarter of the database).

This leads to:

a) Uncertainty of adverse event type frequency;

b) Messed-up (usually exaggerated) differential frequencies, which are used to sort ’signal’ (i.e. the subset of adverse event types that could be, but are not certainly, truly due to the drug) from ‘noise’ (those adverse events that are merely coincidental or unique idiosyncracies, etc.).

c) Potential over-reaction by regulators who see high frequencies of adverse events and have no idea what proportion are duplicates or even triplicates.

d) Scathing criticism by activists of pharmaceutical companies because the latter see and report fewer adverse events than regulators; this is then falsely and darkly imputed as reflecting deliberate, favourable bias towards the Companies’ products by those who sell them.

e) The consequent risk-benefit assessments often being inadequate.

One, current hoo-har where all this is likely to be at least a component is for rosiglitazone and cardiovascular adverse events in patients with diabetes. Large-scale prospective studies (i.e., the gold standard type of study, e.g., ProActive for pioglitazone by Takeda) have shown that these drugs benefit patients with diabetes and improve their otherwise high rates of cardiovascular complications. But a few boffins with track records for criticising the industry got hold of adverse event reports and have published a paper; luckily, this time, it looks like the FDA is being appropriately more circumspect about these boffins’ claims than was the editor of the New England Journal, and, of course, the lay press. It must be confusing and concerning for the patients.