What's interesting to me is the statement that the results weren't significantly changed by weighing for invasive ventilation. Whether or not patients were invasively ventilated is part of the defining elements between severe and critical in our trial.

My question then, for how these meta-analysis numbers relate to our trial would be what percentage of patients in the severe category were admitted to ICU?

If a large enough percentage weren't, then the calculus hasn't changed. It's still probably going to come down to how many were on the severe side and how many were on the critical side for the overall mortality rates of the study.