For your reference, here are the questions we asked:
[Razavian et al, 2015] The trained model used in this paper might not identify high risk patients well for a rural hospital in Wyoming. Name two potential reasons why.
[Razavian et al, 2015] In Figure 1, we see that Patient G is excluded because they experienced diabetes onset during the gap period. Why is there a gap period for this prediction task?
[Pozen et al, 1984] Based on the paper’s results, name three ways we would expect the CCU to change if we rolled out the model completely and physicians followed the model’s recommendations.