When small degrees of bias in randomized trials can mislead clinical decisions: An example of individualizing preventive treatment of upper gastrointestinal bleeding | Publicación