Thread by @anish_koka

Small plug: Medical errors as 3rd leading cause of death comes from this paper in the Journal of patient safety journals.lww.com/journalpatient…

I believe it’s their most cited.. (1/n

External Tweet loading...
If nothing shows, it may have been deleted
by @roguerad view original on Twitter

John James accompanied by a number of illustrious academics sat together in Congress and testified to the veracity of this new evidence.

The number - 440,000 dying from medical errors in hospitals means more than half of folks dying in hospitals are from medical error

When I looked at the paper- It was based on four trials

1 trial was over 1 week.
1 trial didn’t report which harms were preventable
The largest trial -2341 cases reviewed - 588 harms ID’d, 9 were deemed preventable deaths.

So data says - .4%

James made some assumptions that overreached - essentially conflating all harms with preventable harms..

And then aggravated the problem by saying - since diagnostic errors weren’t included, there were probably 2x as many errors. So he just multiplied his estimate by 2

I reached out to the journal -and asked how this made it through peer review and if they thought the paper should be retracted since this number was being used as the evidence base for all sorts of policy proposals. They demurred, and said it was too late to even publish a letter

I did email with Dr. James- he’s a toxicologist at NASA- smart guy , but zero medical experience. He was galvanized to this by tragedy. His son died of arrhythmic sudden death after being discharged from the ER and seeing a cardiologist.

Not satisfied by his response and with nowhere else to go- I wrote a blog thehealthcareblog.com/blog/2015/11/3…

He responded kindly. thehealthcareblog.com/blog/2015/12/0…

The patient safety movement wasn’t too happy and unloaded on me in comments 🤷🏽‍♂️

I also approached one of my affiliated hospitals and asked if there was a way to generate our own numbers - Zero interest.

I’m in solo practice - the daily struggle is to stay alive , not do research with .25 FTEs.

So agree with Sunil- best response is with data- but reality is that most of us don’t have the infrastructure to respond in kind. There’s a small group of folks that have access to funding/data/stats/phd students/friendly peer reviewers to churn out papers. Hats off to them..

But having this high of an entry barrier isn’t a good thing. So it’s especially important for data to survive critiques - even those on twitter - to establish veracity. Don’t get too annoyed by us Sunil! It implies fragility :) #Antifragile