Abstract
This paper discusses several approaches to improve quality of research in intensive care medicine. The baseline standard of care is important in randomized controlled trials. If standard of care is low, trialists could consider improving this before starting the trial. Implementation studies and efficacy trials should not be mixed up. Trialists could further try to increase prognostic as well as predictive enrichment, e.g., through biological phenotyping. Robustness of statistical findings can increase by enrolling sufficiently high numbers of patients and minimizing loss to follow-up.