[fragmede]

There was a study done on survival rates of acute heart patients, and the effectiveness of cardiologists. The data showed that when cardiologists were away at conferences, mortality rates went down 7 percent or so.

Yet the American Heart Association treated this study as a curiosity, and doctors deemed it only worthy of snide comments about using it to justifying conference trips to hospital administrators.

http://freakonomics.com/2015/04/09/how-many-doctors-does-it-...

As for science, the two examples I'll trot out yet again are Pluto, and kibibytes. Pluto, is of course, not a planet, yet many cling to the nomenclature simply because that's what they learned growing up. More relevant for HN's is usage of kibibytes for 1024 bytes and kilobytes for 1000 bytes. Uptake on replacing the more traditional usage is similarly slow, for some reason.

[ceejayoz]

> There was a study done on survival rates of acute heart patients, and the effectiveness of cardiologists. The data showed that when cardiologists were away at conferences, mortality rates went down 7 percent or so.

The ramifications of that aren't clear, though, as it could be something so simple as the risky/difficult procedures being delayed when the best docs are out of the hospital.

[nitrogen]

The difference with kibibytes is that they weren't discovered scientifically, someone (or some group) just decided to force SI prefix normalization on inherently binary units. Most people I know don't use power of two units anyway, though, so those two factors alone would definitely slow uptake.

Pluto is sort of the same, though at least there's an empirical justification in that there are other similarly sized bodies that shouldn't qualify as planets, so a consistent definition was chosen to exclude them and Pluto. But because it's a definitional change rather than a new discovery, people can be slow to adopt it.