Yeah my point was more that the JHU estimates are purposely a lower bound. E.g. they are using only data from hospitals, even though most procedures are happen

Yeah my point was more that the JHU estimates are purposely a lower bound. E.g. they are using only data from hospitals, even though most procedures are happen in outpatient or nursing home settings. So whatever the real number of accidental injuries and deaths from sleep deprivation, you can't get it as just a percentage of the JHU estimate.