Earlier quoted context omitted.
I think stating it as "ventilators are killing people" is wrong. It's an intervention because these people don't have enough oxygen in their blood, the intervention helps them put oxygen in their blood. It's all well and good to look at alternatives as you put it in your last sentence, but the way you put it at the outset sounds like you think ventilation is worse than nothing on the whole, which is probably not true…
That isn't what I said. You misquoted me and I posted my source. To reiterate: Some doctors are concerned the ventilators are actually killing people. They are concerned because the death rate is shockingly high for ventilator usage, much higher than is typical with ventilators. The evidence is mounting that use of ventilators should probably be an absolute last ditch treatment option for people with Coronavirus and…
No inference can be drawn from data points such as death rates with and without ventilators, since ventilators are not randomly assigned. This is just basic statistical inference and the reason why you don't have ANY official sources or guidelines following your directive.
It would still be likely that the people dying with ventilators would die without them, and there is indication that people that survive with ventilation would not have survived otherwise.
Consider, for example, the correlation of death rate for countries without enough ventilators and those (such as Germany) with enough ventilators available. The difference is very large.
Consider also that Chinese doctors recommend the exact opposite to what you say. A lot of people die from heart or organ failure because of the stress oxygen distress induces.
There are a lot of cases where recovery would be possible if it weren't for organ failure. In these cases, ventilation seems to help.
In that sense, there is clear evidence that ventilators are useful and, at least on a population level, outweigh the risks.
We can not (ethically) run experiments in these situations, so it would be wise to refrain from making questionable analyses based on anecdotal data.