This is two pages long (excluding a page of images). It is not that heavy e.g. no numbers. It touches on some of the issues I brought up in my paper: Unfortunately the article then finishes by looking at one trial in particular, the PACE Trial, and making it out it clearly found CBT and GET were not more risky. I discussed the PACE Trial in particular in Section 6 of my paper, "PACE Trial – A model of excellence in harms reporting?"