As pointed out in the comments on the site, the key issue here is that it is not a blind study, but a case series study where the demographics of the control gr

As pointed out in the comments on the site, the key issue here is that it is not a blind study, but a case series study where the demographics of the control group are unreported: >The main problem with this work is that the demographics of the control group are not reported, so we have no idea if they are similar. This them means that we have no idea if the differences in outcomes are because of the treatment, or if they would be expected because of the demographies of the groups. >A subtler point is that the reported age profile of the treatment group is impossible. Group A is aged >60 years, and almost half of the treatment group was in that group. But the IQR was reported as 40-60, so only 25% of the treatment group had an age >60. I hope this was due to excessive rounding, or a similar error that can be corrected. It appears that the researchers selected the treatment group based on a set of quite specific criteria and then compared the outcomes of those patients with outcomes in the general population (see figure 1).