Ratios Proportions And Percents

If you want to prove that open surgery cures 95% and endoscopic treatment cures 80% with a beta of 0.20 and alpha of 0.05 then you will need n=73 patients per group, or 146 for the whole study. For a one-sided study which would be ok to do in this case (Alternative hypothesis: open surgery is superior to deflux) you will need 60 patients in each arm, 120 total .

But drop down your pen because here is an online calculator for sample size. Where it says mean 1 put p1 expressed as proportion (80%=0.80), and where it says mean 2 put p2 again expressed as proportion. Leave SD1 and SD2 empty, leave allocation ratio as 1. You can change power and alpha if you want. If your plan is to compare two groups use the Two sample method. The results will be slightly different from the ones obtained with the above  formula because the formula takes some shortcuts, but the difference should not be by that much.

In order to detect a 15% difference (80 vs 95% success rate) in between the surgical and endoscopic groups, 73 patients are needed per group. If you want to detect a 5% (90 vs. 95% success rate) difference in between the groups then you need 435 per group. With only about 20 patients per group, the study mentioned above was only able to detect a big difference of about 35%  from 60-95%. A one-sided (one-sided means you presume superiority on one of the treatments) comparison for a difference in between 70 and 95% needs 28 per group. To be fair, I was not there during the presentation of the abstract and I am sure the researchers had an explanation for their power calculations. In any circumstance, you will find that negative studies are always harder to publish and are more scrutinized because current scientific standards are designed to err on the side of not finding differences that are real (high Beta) instead of finding differences that do not exists (that is why alpha is small).

Conclusion

When evaluating a study claiming no difference or equality, use an online sample size calculator to determine if the study was powered appropriately. Negative studies are scrutinized more heavily than positive studies because of current scientific standards determined by the scientific community dictate for a higher beta and a smaller alpha.