I'm not sure, the data seems oddly and unnecessarily skewed in terms of the group breakdown. - The placebo group was almost 10 years older on average. You mig

I'm not sure, the data seems oddly and unnecessarily skewed in terms of the group breakdown. - The placebo group was almost 10 years older on average. You might think that this would be fine, or even make the effect demonstrated even larger than the study suggests, except 8 patients were removed from the placebo group due to being moved to the ICU before the start of the study. This introduces a massive amount of selection bias into the data when you only had 48 in the placebo group to start with. - The placebo group had about half as many women as a % Of course there is an attempt to remove these confounding factors in some of the data, but the big results appear to be un-adjusted numbers (number of days in hospital being reduced, reduction of the need for oxygen, etc). There is also an odd coincidence that they selected 53 patients for the Placebo group, and 48 for the L-Arginine group, and the groups lost 8 and 3 patients to the ICU respectively, so the final groups were both 45. Is this normal in these types of studies? Why wouldn't they have made the intention to treat population 50 for both cases?