Question special
Resident

The power calculations required a sample size of 1360 to achieve adequate power; this, however, was hampered by an observed rate of altered mental status about 75% less than previously assumed (3.5% vs 15%). Which factors went into deciding not to continue to enroll more patients to achieve the anticipated power, especially in light of the fact that the confidence intervals on the primary outcome (0.44-1.33 and 0.46-1.40) cannot exclude the clinically meaningful difference threshold from the protocol of 37.5% (RR 0.63 or 1.6, i.e. 7.5%/20%)? I imagine that some--if not most--of the explanation pertains to funding, in which case, maybe the question can be more existentially asked as why the NIH didn't extend funding in order to answer the question more definitively? We should make conclusions about the secondary outcomes with some hesitancy, and appreciate that more data might move the point estimate of the primary outcome towards the null, but there at least appears to be a signal favoring rapid hydration with hypotonic fluids. Given that this potential signal is the opposite of previous conventional bedside wisdom, that fluid type and rate administration are frequent clinical considerations, and that recent research has revealed some intriguing hypotheses for biological plausibility, narrowing the error bars on this estimate would appear to be quite valuable.