Earlier quoted context omitted.
> As they acknowledge, this factor is extremely difficult to isolate from demographics, geography, religion, etc. No? They used "a within-facility differences-in-differences design to address nonrandom targeting of facilities" so they're measuring the change in outcomes at the same facility before & after takeover.
That is not conclusive either. If PE had not acquired them, may be they would have been worse or better? This analysis doesn't answer that.
Regardless, in this case, where the ongoing harm could be very large, we have to at some point go with the best evidence available.