I am not sure if I am fully understanding your question or not, but if you are asking about how to define your treatment and control groups then yes having two categories: (hospitalized: treatment group) and (non-hospitalized= control group) would be good enough. You can use age and duration of sickness as covariates.