this is exactly why you have to have a controlled trial, against...

this is exactly why you have to have a controlled trial, against either placebo or 'standard of care'. The fact that a patient is entered in a clinical trial means that everyone involved is teed up to follow protocols correctly, and 'sloppy practice' is minimised (and the hospitals involved are already top line trauma units). This is itself may be enough to significantly reduce death rates. This is in fact the more likely cause for a noticeable overall reduction in mortality, as there are very few examples of any drug that is so successful that it could skew data significantly across the entire study cohort - take aspirin in acute MI, for instance: everyone agrees that it is one of the most important early interventions, yet the number needed to treat to prevent one death is 26. It reduces death by 4%, so if you were blinded in an aspirin vs placebo 1:1 study, the reduction in deaths overall would be 2%, probably not noticeable.