Statisticians Wolkewitz and Schumaker using methods to avoid survival biases on observational data from 1391 patients hospitalised in UK during 2009-2010 fail to find an effect of Tamiflu on mortality but show a shortened duration of hospital stay.

View the paper here.