I would, however, take this data with some caution

I would, however, take this data with some caution. Although I have not read the article, it sounds as if the authors have taken the studies as an aggregate. But the success of trials of different disease classes differ significantly, e.g. cancer vs acne