Hacker News Reader

Comment by DanielBMarkham | Hacker News Reader

"... Imagine if we optimized every single facet of medical care, from patient admission to discharge, and all intermediary processes, top to bottom..."

Let's separate standardizing and optimizing. A checklist lets you standardize. It does not optimize anything.

As the author continues, he gets into a scheme of various agencies all working together in various ways -- this is the core of the optimization problem. I'd suggest he doesn't have a clue here as to how to actually accomplish this. But still, overall this is a great idea. Standardization in many rote procedures, especially those involving health, is desperately needed.

View on HN

I think that those things are not even new. There are already decision support systems(medical expert systems) and protocols of evidence based medicine - for standardization.

And research works around the world is flowing into those systems and improving them.

And the line-infection rate checklist the author talks about, does exist for something like 10 years.

Those are the relatively easy parts.

The hard part is: how do you change the culture of medicine, from "cowboy style" that comes with a lot of authority for the doctor, to a more boring job , with much more subservience to machines and protocols ?

The question to ask is, how is it that pilots made the switch from "cowboy style" to more reliance on machines and protocols? The answer is probably cultural.

Which brings me to another thing that occurred to me: Is medicine ready for "5 whys?" Checklists aren't going to work until the culture of the medical profession is ready for it. I get the impression that doctors are treated as "functionally infallible" (so long as they consult with other doctors when they need to) and mistakes are treated as flukes and aberrations. (If not entirely swept under the rug.) It's pretty clear that medicine's capacity for self improvement is not entirely mature from the performance of infection control. There are still hospitals that spread infections. (http://www.ehso.com/ehshome/washing_hands.htm)

It's more than cultural. All airplanes of a type are basically the same. They have the same parts, the same operational and maintenance requirements, and the same procedures work to fix them when something fails.

Humans aren't so standardized. Not everyone with a disease responds to the same treatment in the same way. They may not even display the same symptoms. Some people have allergies to certain medications, etc. Medicine is just way more complicated than ticking through a pre-flight before takeoff.

IN the case of pilots , few factors combined together to build this culture:

1. Checklists were introduced when flying was very young,with very few trained pilots.

2. Pilots want to live, so they use checklists.

3. Pilots don't have a choice. All airlines companies enforce checklists. Probably it's an FAA requirement.

4. Checklists are a part of training.

Complexity of operations. Simple planes don't really need checklists. There are not that many knobs to turn and not many things can go wrong. As the planes got bigger, faster, and got more equipment, it became impossible to just "follow the flow".

Cowboy style? Do you actually believe this? On what basis are you suggesting that medical errors are related to the reckless behaviour of doctors? Do you have some secret research that goes against the current thinking on medical errors and how they are system problems, not related to a given individual? Do you have any evidence that subservience to mythical machines will improve health care? Do you work in healthcare? Do you deal with doctors and patients?

Atul gawande which is a surgeon coined the word cowboy in a new yorker article and in a lecture before new doctors.

There's some research that shows that decision support systems improve medicine. Too tired to look it up.

OP does eventually talk about A/B testing in order to optimize the checklists. I assumed he was just getting ahead of himself.

Herself. Judging from the final line, A girl can dream.

Himself, actually :P

You would know best, I was just looking for clues...

Reply on news.ycombinator.com