The thumb joint in this trial was the CMC joint (base of thumb) so the antenna probably covered part of the wrist. But really, I think the problem was simply that the study was underpowered. These 60-patient trials (30 active, 30 placebo) are so small that the results are just a crapshoot, IMO.
That's why I think Ilfeld's SofPulse study was successful while his RecoveryRx study was not. Could have gone either way, IMO. The Insurance Industry certainly believes that -- every insurance company decision against covering PEMF "personal sized devices" states that the existing evidence is inconclusive because the trials are underpowered.
There was a previous wrist study that was successful: