A new analysis of data from trials of oseltamivir (Tamiflu) has reopened the debate over whether the drug reduces deaths from H1N1 influenza.
Conducted by the Cochrane Tamiflu team, the study found no effect on mortality,1 in contrast with a 2014 meta-analysis conducted by the PRIDE consortium at the University of Nottingham, which concluded that the drug did save lives.2
The two groups have already crossed swords when the PRIDE study was published and was immediately criticised by the statistician Mark Jones, who is the lead author of the relevant chapter in the new Cochrane analysis. Those who have closely followed the controversy may not, therefore, be surprised at the Cochrane findings.
The difference between the two teams hangs on the methods they used to analyse the data and, in particular,