Editorials

Increasing power in randomized controlled trials* Conducting adequately pow-ered randomized controlledtrials in critically ill patientsis a challenging business. Without even starting to consider issues of ethics and consent, the critical care researcher is limited by the relatively small and extremely heterogeneous avail-able patient population. The result is that researchers often aim to detect an unre-alistically large treatment benefit; conse-quently, we see too many negative trials in the critical care literature (1). Some authors have proposed that the solution to this is to consider intermediate or composite end points to trials, rather than seeking a reduction in mortality