That number comes from a single paper, and that notably is the “high” estimate that it provided in the range. Why not take the low estimate or the middle? Second, that paper uses global trigger tools, not direct measurement of error to come up with that estimate.
-
-
Why does that matter? Because the tools have poor interhospital consistency and are not appropriately validated. They look for “proxies” for error in administrative data - an unreliable data set, that suggest an error may have happened.
2 replies 0 retweets 2 likes -
Worst of all while it may be “validated” for being sensitive to error, I have yet to see that it is specific enough. That is, can someone show me that the tools are not detecting death period? All the proxies are events one would also expect in a dying patient without error.
2 replies 0 retweets 2 likes -
Replying to @MarkHoofnagle @gorskon
Throughout our 30+ interviews we encountered people who argued just as fervently that 440k was too low. That the number of deaths across the board caused by easily preventable systemic flaws could be higher.
1 reply 0 retweets 0 likes -
Replying to @ToErrIsHumanDoc @gorskon
Fanstastic. 440 would be over half of all deaths happening in the hospital and thats too low. There are people who won’t be happy until it’s 100% I guess but that doesn’t seem a good reason to listen to them. Do any of them actually treat patients?
2 replies 0 retweets 3 likes -
Do you not get how that is prima facie absurd? I have no doubt that there are people with careers invested in making this the number 1 priority, but there is significant pushback for a reason and it’s not that we are defensive - it’s that measuring this wrong causes real harm.
2 replies 0 retweets 2 likes -
Replying to @MarkHoofnagle @gorskon
All I can really say is that we worked for 3 years on this film, using interviews from leaders in medicine and policy. We worked hard to provide a balanced look at what is being done to improve the flaws in our health care system.
1 reply 0 retweets 0 likes -
If the use of this numbers means you will not give our film a chance, I can’t make you watch it. But I do believe our film’s intent is good & helpful. We can’t tackle every area of this issue, so I am happy that people like you both are working hard to promote more accurate data.
3 replies 0 retweets 0 likes -
If a new study is released, and is widely cited and uses new data, we will absolutely share it and promote the widespread dissemination of it, whether it is 600k or 60k. I’m sorry you feel our intentions are bad.
2 replies 0 retweets 0 likes -
Replying to @ToErrIsHumanDoc @gorskon
No, the problem is everyone’s intentions are good, it’s their understanding of fundamental scientific principles, such as randomization, ITT analysis, and bias that are bad.
1 reply 0 retweets 3 likes
Correct, and those errors have real world consequences and cause real harm.
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.