The small sample size looks like a 'going thru the motions' strategy knowing the outcome is already predetermined. Large enough to have some sort of significance (especially if survival rate is >67%), yet small enough that the study will be over and done with quickly.