Use of the CDC dataset as controls for a non-randomized trial is very problematic. Hard to know what to make of this at all. It has a modicum of bioplausability

Use of the CDC dataset as controls for a non-randomized trial is very problematic. Hard to know what to make of this at all. It has a modicum of bioplausability but I don’t think my posterior probability estimate of effectiveness has changed at all after reading the paper.