SoylentNews is people
SoylentNews
A study published in the BMJ found that medical errors may be the third leading cause of death in the United States:
The IOM, based on one study, estimated deaths because of medical errors as high as 98,000 a year. Makary's research involves a more comprehensive analysis of four large studies, including ones by the Health and Human Services Department's Office of the Inspector General and the Agency for Healthcare Research and Quality that took place between 2000 to 2008. His calculation of 251,000 deaths equates to nearly 700 deaths a day — about 9.5 percent of all deaths annually in the United States.
And from the airplane analogy, a simple fix: checklists.
Is it time for a system theory approach to medicine?
Medical error—the third leading cause of death in the US (DOI: 10.1136/bmj.i2139)
(Score: 4, Interesting) by Justin Case on Saturday May 14 2016, @12:05PM
Poor ... procedural execution
People just hate to follow procedure. Nope, I'm too smart. Telling me how to do my own job is insulting.
Try getting your local MCSE to set up three PCs the same way by following a written checklist. Better chance you'll win the lottery.
This is why entry level jobs are being replaced by machines. Machines follow instructions.
I have a relative who works in healthcare. Even when the nurse's own life is on the line (Ebola, for example) they won't do what they're told.
(Score: 2) by hendrikboom on Saturday May 14 2016, @03:15PM
Following instructions correctly is a superpower -- especially when the instructions are wrong.
(Score: 2) by Justin Case on Saturday May 14 2016, @03:40PM
Instructions can have bugs. That's what software is -- instructions. The computer mindlessly and exactly obeys them for better or for worse.
But imagine a compiler that magically "cleaned up" your bugs... 30% of the time... and also injected its own bugs... 90% of the time. How would you ever find the flaws in the source code and get them corrected?
If a human notices a flaw in their job instructions, ideally they should be reported and fixed. Alternately they may be followed so that the instruction writer can discover the bug and fix it. But don't just say "Oh, this doesn't look right to me. I think instead I'll leave the flaps down throughout the entire flight."
(Score: 0) by Anonymous Coward on Saturday May 14 2016, @04:09PM
But imagine a compiler that magically "cleaned up" your bugs... 30% of the time... and also injected its own bugs... 90% of the time. How would you ever find the flaws in the source code and get them corrected?
No need to imagine, it happens often enough. Ever tried to debug a segfault in C/C++? Compiling with the debug switch turned on often "fixes" the problem... Ergo, most of my code is littered with commented out printf's on the pattern of "entered function xyz", "for loop start" etc.
<mutter>Goddamn Heisenbugs...</mutter>
(Score: 2) by bitstream on Saturday May 14 2016, @06:51PM
When a human finds an error in the list of instruction that can.. Halt and burn the management ;-)
A flamethrower makes the task easier :p