[jabagawee]

> The researchers divided 500,000-plus cases into four categories: male doctors treating men; male doctors treating women; female doctors treating men; and female doctors treating women. “All of those are statistically indistinguishable except for male doctor–female patient,” says Brad Greenwood, an author on the study and a data scientist at the University of Minnesota. If a heart attack patient is a woman and her emergency physician is a man, he says, her risk of death suddenly rises by about 12 percent.

[PurpleBoxDragon]

I wanted to look at the actual numbers and see how close the findings were, but I don't see the study linked nor the title of the study (I do see a different study, but it isn't the one those numbers are from).

A different article someone else linked seemed to have the study, but it is behind a pay wall.

http://www.pnas.org/cgi/doi/10.1073/pnas.1800097115

This seems to have the actual numbers though.

http://www.pnas.org/content/pnas/suppl/2018/07/31/1800097115...

I'm having trouble grasping how to read the tables (especially given it seems Female Physician Female Patient is repeated twice on tables S2 and S3, but I think that is just a title error.

Here are the numbers, best as I can grasp.

S2 is full, S3 is matched

Mean then standard deviation.

    M/MS2:.881/.114
    M/MS3:.870/.124
    M/FS2:.854/.353
    M/FS3:.861/.121
    F/MS2:.887/.120
    F/MS3:.867/.130
    F/FS2:.857/.116
    F/FS3:.862/.112

Figure S2 looks interesting (M/F and F/F seem equal while M/M seems better than F/M), but I'm not sure what the real axis are.

Edit: formatting