A letter to the medical journal The Lancet from Professor Ian Roberts and colleagues in London examined the potential effects of time delays related to obtaining written informed consent on the ability to detect beneficial therapies for trauma patients.
The CRASH trial tested whether steroids administered after serious head injury would reduce brain swelling and inflammation to improve patient outcomes. The CRASH study enrolled 10,008 subjects with head injury in a study of steroids versus placebo, and was also published in the The Lancet (Lancet 2004;364:1321-28). There was no observed benefit from steroids.
Some hospitals in the CRASH trial used principles like Exception from Informed Consent (the community provides permission to conduct emergency research), and some hospitals required the patient's family or representatives to provide written informed consent. Hospitals using the Exception initiated treatment about 1.2 hours faster than hospitals requiring written informed consent.
The authors argue that a one-hour delay, like that observed in the CRASH trial, would have reduced the the absolute proportion of subjects benefitted by a treatment in the more recent CRASH-2 trial by 14%. Such a delay might have prevented the trial from detecting a beneficial effect. The CRASH-2 trial found that administration of a pro-clotting drug, tranexamic acid, to bleeding trauma patients improved survival, and that this effect depended on how quickly the drug was initiated.