Treatment of missing data in follow-up studies of randomised controlled trials: A systematic review of the literature