Calculation of MIMIC-III SOFA AUC score over time in original paper

[HarrisonWilde]

Prerequisites

• [ X ] Put an X between the brackets on this line if you have done all of the following:
  • Checked the online documentation: https://mimic.mit.edu/
  • Checked that your issue isn't already addressed: https://github.com/MIT-LCP/mimic-code/issues?utf8=%E2%9C%93&q=

Description

Description of the issue, including:

• what you have tried
• references to similar issues
• queries demonstrating your question (if applicable)

I am trying to reproduce Figure 2 in https://academic.oup.com/jamia/article/25/1/32/4259424 p.3 / 34 (online / pdf) and unsure as to how to go about achieving the graph as the AUC calculation I do on the SOFA score relative to mortality is lower. Please could you assist / provide code for how that figure was generated? Thanks!

[tompollard]

@HarrisonWilde Have you looked in https://github.com/MIT-LCP/mimic-code-paper/? The code should be in there, but from a quick glance I couldn't see it. If not, we will need to do some digging.

[HarrisonWilde]

Thanks for the quick response @tompollard , I have had a look and cannot find it there either. I noticed a few of the other figures either are there or look like they are generated there but there doesn't seem to be anything that would line up with that figure.

[tompollard]

I'm sorry, this was an oversight on our behalf. It's been several years since we wrote that paper, so we will need to do some searching. If your issue is that the SOFA score in the MIMIC-Code repo doesn't match the plot, the first step would be to (1) make sure that you are using code from around the time of the paper, which should be doable by searching the commit history and (2) making sure that you are using the same version of MIMIC-III.