A US institute analysed the data gathered between 2013 and 2014 through the Food and Drug Administration's pharmacovigilance system. Of the 847,000 adverse effects reported, around 29,000 reports came directly from patients and healthcare professionals, while most (96.5%) came from pharmaceutical companies. Reports from patients and healthcare professionals were more comprehensive than those from drug companies, for example giving the patient's age and gender and the date in 81% of cases, versus only 46% of the reports from drug companies.

When it comes to reporting the most severe adverse effects, the situation is hardly any more satisfactory: 85% of reports from patients and healthcare professionals are reasonably comprehensive as against 49% from pharmaceutical companies.

Furthermore, there were a number of reports from drug companies giving information that was vague or not very useful, such as: "pain at the injection site", "adverse effect" (with no details), "nasopharyngitis" and even "no adverse effects". In cases of patient death, the pharmaceutical companies' reports were of little use. In 28% of cases, it was unclear whether the drug had played any role in the death. In cases of congenital malformations, only 25% of reports were sufficiently comprehensive to be of any value.

In short, this new study and many years' experience show that we cannot expect pharmaceutical companies to provide quality data on the problems associated with their drugs. And it is unacceptable that governments tend to rely on the pharmaceutical industry to improve pharmacovigilance, sometimes even considering a relaxation of the clinical evidence to be provided before a drug to receive a marketing authorisation.

©Prescrire 1 November 2015

"Adverse effect reporting: pharmaceutical companies can't be trusted" Prescrire Int 2015; 24 (165): 256. (Pdf, free).