Now do it outside the country, and I anticipate the results don't replicate, meaning your cited study probably fails replication.

I'm not sure what you mean by this. Let's imagine MCAT scores directly determine malpractice rates. If one country practices race based affirmative action in medical school applications, and another country does not the we should expect the former to see a disparity but not the latter.

Why would we expect this trend to replicate across countries which have different medical training systems?