MSB Trading - 2020, page-10382

Yeah so my methoding is why i'm cautious about the number and that i could be wrong lmao. I'm still getting the hang of statistical data interpretation.
Instead of a two-tailed test with upper and lower tails like you mentioned, I thought the clinical trial is only a one-sided test. ie. the hypothesis is that treatment with remestemcel will lead to an increased survival rate (> 12%). For the null hypothesis to be rejected, the result can only be higher than 12%, not lower, therefore a lower tail does not make sense ergo one-sided test.

Also as the trial variable is: treated vs not treated, and the responses are: survived or not. There's only 2 states for the response so I used a Binary Response approach, and ran a Logistic Regression Analysis using the estimated survival numbers of the non-treated cohort, and kept changing the treated cohort's survival rate until the chi-squared statistical p-value was reached. 13/45 was just above 0.05, 14/45 was the first it goes below 0.05.
These were my assumptions and again, new to statistical analysis at this level!