Oseltamivir (Tamiflu) and causality
admin July 13th, 2007
The Japan Institute of Pharmacovigilance has written to the BMJ concerning a editorial concerning oseltamivir. The letter states the following:
On 16 June 2007 the Japanese Ministry of Health Labour and Welfare announced that by 31 May 2007 it had received 1377 reports of adverse reactions since 2001, when marketing of oseltamivir started in Japan.
Of these, 567 were serious neuropsychiatric cases, 211 showing abnormal behaviour. The number of deaths reported was 71. These are not only “adverse events” but also “adverse reactions” to oseltamivir because many doctors classed and reported them as probably related or that causality could not be ruled out.
The letter is correct to note that spontaneous reports to regulatory authorities are not adverse events (any event that occurs after the use of a drug with no assessment of causality), since the suspicion of the doctor that the drug was the cause of the event makes them adverse reactions (events that are suspected to be causally-related to the use of a drug). However, it is also important to note that such suspicions are not clear evidence of causality. It is over-simplistic to accept the decision of a doctor to report to a spontaneous reporting system as evidence of causality. Firstly, spontaneous reporting schemes specifically ask for doctors to report possible reactions, not those they have clear causality for. Secondly, the assessment of causality is highly variable between clinicians, especially if the judgement is based on non-expert opinions without the use of a standard algorithm. For this reason, the more cautious interpretation of Dr Simon R J Maxwell may well be the more sensible approach:
Before 2007, there had already been more than 100 reports of neuropsychiatric events (including delirium, convulsions, and encephalitis) with oseltamivir in children, almost entirely from Japan, which has the highest usage of oseltamivir worldwide. However, these disturbing events had to be seen in the context of the millions of prescriptions worldwide and the fact that abnormal behaviour could also be due to flu or disease related complications. Indeed, a Food and Drug Administration (FDA) review of clinical trial and postmarketing data concluded that these events were not clearly drug related but might be related to higher rates of flu related encephalitis in Japan. Since last November, the FDA has required that doctors be warned that patients should be closely monitored for signs of abnormal behaviour throughout the treatment period and the European Medicine Evaluation Agency (EMEA) took similar steps in February.
The controversy about oseltamivir is a further reminder that, although common adverse effects of a drug may emerge in prelicensing studies, the detection of rarer and potentially more serious events has to await exposure of large numbers of patients. In the UK, oseltamivir is a “black triangle” drug, so it remains under more intensive surveillance. Doctors and other healthcare professionals should report all minor as well as serious adverse events via the yellow card scheme.
It is also notable that the Japan Institute of Pharmacovigilance is non-profit making organisation that argues against the use of flu vaccine and anti-pyretics, as well as publishing articles suggesting that “High cholesterol level makes a long life“. Things you may wish to bear in mind when assessing their views on drug safety.