Several of you have asked me about about the “ATLAS tool” used at Tulane to rank applicants.

Until the details of Dr. Dennar’s lawsuit emerged, I’d never heard of it. But I have now confirmed some of the details.

(thread)

First: as others have noted, this is an internal tool, not a commercial product.

Essentially, ATLAS is just a fancy name for an Excel spreadsheet calculator. You put in certain variables from the ERAS application, and it scores/weighs those variables to create a rank score.

The primary ATLAS input variables are:

-USMLE Step 1 score
-USMLE Step 2 CK score
-Interview scores
-Class rank from the MSPE, modified by school strength
(according to the US News & World Report rankings)

It’s easy to see how these could be problematic.

For instance, there are well-described systematic differences in USMLE Step 1 performance by the test taker’s sex and race/ethnicity.

The lawsuit highlights the impact of the USNWR rankings - which are horrible measures of medical school educational quality.

By weighting NIH funding as 40% and measures of ‘reputation’ as 30% of a school’s score, the USNWR rankings calcify a certain hierarchy in medicine.

Knowing that the odds are stacked against them, many schools refuse to provide data to USNWR - and get listed as “unranked.”

Among HBCUs, only Howard was ranked (in the bottom quartile) of the 2021 survey.

(Interestingly, Tulane also refused to provide data and is unranked.)

A couple of final thoughts about this:

1) I don’t know the precise weighting of the variables in ATLAS. There should be an external audit to determine the impact of the formula chosen on the rank order list.

(Unofficially, Tulane *may* be willing to allow this. Stay tuned.)

2) I strongly suspect that many programs outside of Tulane use similar tools to rank applicants.

If your program is one: think carefully about your methodology.

We all crave objectivity. But just slapping a number on something doesn’t make it objective, true, or free from bias.