Here is the decision curve. See? You can just look at it and get your answer. The model has extremely tiny benefit at the sort of thresholds that are of interest here and so we shouldn’t use it.
Thank you for this series of exercises! It’s nice to see these applications of decision curves in different settings where regardless of whether the model itself is ‘good’ this does not mean it is clinically useful.
Hope to apply them in my own project soon!