Hacker News Reader

That doesn't really sound like non-inferiority to me as a layman. That's a difference of over 25%.

Reading further in their additional conclusions, the surgery group also did better in both pain and general health.

Look at the graph in figure 2 - the massive overlap in the interquartlie ranges suggests that any differene in the median values can be attributed to random variation within the samples, rather than a true difference in the populations.

Also its better to quote an absolute rather than relative reduction, so the difference is really 3.6%

Reply on news.ycombinator.com