Aside from a virus, the sample set has hospitalization as a possible factor. Perhaps some have picked up an infection (bacteria), or another contributing virus?

Aside from a virus, the sample set has hospitalization as a possible factor. Perhaps some have picked up an infection (bacteria), or another contributing virus? Also, what about pre-existing conditions? The already compromised having to be hospitalized might mean they left still compromised. I'm not seeing how this analysis rules out these other possibilities. Lots of correlation here, but seemingly not much proven causation (i.e., it's all virus driven).