We know that New York's hospitals are overrun. We also know that in countries where the hospitals aren't at capacity the mortality rate is much lower. Look at Wuhan vs the rest of China. Or Singapore/HK/Taiwan/South Korea vs Italy, Paris, New York.
So if ventilator access is low and doctors are triaging ventilators to go to the most urgent cases, is the data skewed towards severe cases and late access to ventilation support?
This anecdotal account[1] seems to suggest ventilate as early as possible to maximise the chances of saving a patient. So again, a global stat could be masking a time element and context to the data
