The study you're probably talking about focuses on hospital survivors, not on the broad group of "naturally infected people".