This chart is a bit unsatisfying since the methods that underlie it are not apparent in the AHA source document (PDF).
Still, it can’t be denied that dying reduces one’s risk of readmission. Yet, in general, one would not welcome death in place of a readmission. In the literature it’s standard to examine readmissions alone, not readmissions or death. But, to the extent the former makes sense to examine (e.g., as a measure of quality), I wonder if the latter makes even more sense.