How to weigh the current evidence for clinical practice

This article presents a template for judging trials of tight glucose control in critically ill patients. It reviews threats to both internal validity and generalisability using examples from the current literature. When judging internal validity, it is important to consider factors specific to trials of glucose control (particularly the methods of glucose control, measurement and reporting) in addition to factors common to all randomised controlled trials (such as treatment allocation, losses to follow-up and protocol violations). Judging generalisability requires the identification of differences between the trial population and the population for whom the intervention is being considered. These may relate to the setting, the patients or the practical delivery of tight glucose control or other interventions. Once identified, a judgement must be made for each difference of whether it is likely to modify the effect of tight glucose control.

Keywords: intensive care, insulin, glucose, hyperglycaemia, internal validity, external validity, generalisability