Somebody sick dying in hospital, on the other hand, is just something that happens, and draws very little attention - even though way more people are affected than in plane crashes.
So, the clustering and visibility of plane crashes leads to excellent check list discipline and other best practices in aviation, by and large (CRM, Crew Resource Management, is another thing Atul Gawande brings up in The Checklist Manifesto, and also eminently transferable to surgery), while I suspect that in surgery it is easier to drift away from best practices again without anyone noticing.
That's why these large studies and the educational efforts of Dr Gawande and others are so important.
edit: typo
If we saw a midair collision of jumbo jets every day however the public would never accept it.
https://catalyst.nejm.org/medical-errors-preventable-deaths/
Because of that book, I'm not surprised that the checklist reduced deaths, but its a bit shocking that its by a third. Then again, the base rate was very low to begin with, showing that surgery teams were already doing a seemingly good job:
"The death rate fell to 0.46 per 100 procedures between 2000 and 2014, analysis of 6.8m operations showed."
Still, over 6.8 million operations, is that 15,000 survivals? Wow.