MIT student applies predictive modeling to improve emergency-room wait times | Hacker News Reader

MIT student applies predictive modeling to improve emergency-room wait times

web.mit.edu

34 pointsmuzz18 comments

Not to make light of his work, but this is not that novel. Students have been "improving" the medical field using queuing theory and simulation software for years . I did it in my stochastics class when I got my master's degree in Operations Research. The key translation from the article is "Xu emphasizes that the work is still ongoing" which means that isn't implemented [yet] and the 40 minute savings are just computer estimates.

What I would really like to see is an article titled, "Huge real-world increases in medical efficiency directly attributed to student's work and now saves hundreds of lives per year" -- or something like that. I'm not very good at creating titles.

Again, I want to emphasize that I don't want to make light of Mr. Xu's work (stochastics might be my least favorite subject and gets complicated when you create detailed models) and I wish him luck as he pursues his PhD.

Agreed. My sister is an IE for a major hospital. This is pretty much what she does all day every day.

Academia (meaning anything post-masters) has a very nuanced understanding of the word "novel". Skimming through the paper (linked below) it seems that the novel part is that they "show that the amount of information about the future has a signiﬁcant impact on system performance, in the heavy-traﬃc regime." Not Elon Musk-novel perhaps, but definitely academia-novel.

http://arxiv.org/abs/1211.0618

> “In emergency rooms, you’re typically overloaded all the time. If you look at the statistics of the U.S. average waiting time, the total average time is on the order of a few hours. A lot of people leave without being seen by a doctor because they just get frustrated.”

This suggests that some people use a weird definition of "emergency".

Wait times are dependent wholly on how the those medical professionals you first encounter deem it. Come in walking and generally looking fine, you go on one list. Come in with obvious; read visible; issues needing immediate treatment and your going to get it. Come in with issues that can be easily detected; blood sugar, heart rate, etc; and your going to get seen.

Still my experience has been that far too many people are in there that should not even be there and worse most of them are rude as all get out, the common decency of giving up your seat for the elderly seems to escape them. Got to love it when they put their junk in a chair alternating with each other so they can step outside and smoke. Yeah, been there and seen it.

Perhaps with the rise of more localized clinics some of these numbers can be cut down on, but when the only admission area of the hospital for issues is emergency people are going there. They really need an admission area for non emergency but require medical attention. Somewhere that the medical staff can quick assess the situation and perhaps use RNs to dispense non addictive corrective medications or items.

If you don't have health insurance it is practically impossible to make a doctor appointment, so lots of people just go to the emergency room.

People often feel so horrible they get desperate and stop thinking as logically as they should. For example, a bad flue will make you delirious even though, if you aren't very old or young, the best way to treat it is often some ibuprofen, water, and just to sweat it out (disclaimer: I'm not that kind of doctor).

I went to the emergency room this year because I was in pain and needed a pain killer prescription for a bad abrasion (that I was self treating otherwise) at 3 AM in the morning. But in my case, the emergency room at the private hospital I go to in China was quite deserted, and they even gave me a tetanus shot for the trouble.

> sweat it out (disclaimer: I'm not that kind of doctor).

Yeah, if you're delirious from a flu (assuming you mean the fever), you need to go to either your primary care physician or the ER. Do not try to "sweat it out."

The ER will be more than happy to have you there even if you could have stayed home. They know what they're doing; they're not going to let someone else die to treat your sniffles.

Sweating it out in the ER waiting room sucks versus sweating it out in bed at home. Of course, it depends on how sick you are, and if you ask your doctor they will always err on the side of caution.

...and an IV drip usually doesn't help that much, and its not good to have too many of those.

An IV drip can mean the difference between life and death by dehydration. Seriously, it's why doctors tell you to go to the ER if you can't keep any liquids down when you have food poisoning or the flu. Couple that with the fluid loss you're already experiencing from being sick, and it can turn deadly fast. Not to be rude, but you have to be completely ignorant to say, "an IV drip usually doesn't help that much." Unless you're talking about someone who isn't dehydrated in the first place, in which case I'd agree, but most ERs don't hook someone up to an IV for shits; if you're getting a bag, you're probably dehydrated. (They don't want to waste their bags, IVs, nurses' time, and risk infection on someone who doesn't need it.)

Even if you have a healthy immune system, if you're delirious from a fever, you should call your doctor ASAP.

Note: I'm not a doctor, this is not medical advice, yadda yadda. I've just been living with Crohn's disease long enough to learn some of this from my own visits to the ER.

Most ERs in china will hook you up to an IV drip by default without even a doctor's consultation first. And ya, it sucks.

But really, why do you think flu is so deadly? If you are very young, very old, or have other medical conditions, fine, it makes sense to hit the ER. If you are a healthy adult and just got hit with a bad one/two day flu, you are able to keep your fluids in (beyond sweating them out), you don't need to hit the ER even if you feel like crap... By delirious, I'm talking about shady gradients anyways, I've never had more than sweating dreams where nothing makes sense. Ya, there is a point where you should go the ER, but our thresholds are fairly diverse.

At least a few thousand people die from the flu each year.[0] I've just been responding to your original post. You said "delirious", which is an actual, serious medical condition associated with high fever and dehydration.[1] Then you told people to "sweat it out", which is a myth.[2] I'm just trying to correct some of these misconceptions so someone doesn't read it and think that being delirious is completely normal and you should try to lose even more fluids.

[0]: http://www.cdc.gov/flu/about/disease/us_flu-related_deaths.h...

[1]: http://en.wikipedia.org/wiki/Delirium

[2]: http://www.insweb.com/health-insurance/5-common-flu-season-h...

I'm a little suspicious of how much of this is press-piece fluff, given that they make the assertion that his algorithm reduces times by 40 minutes, which is 10%. This means that the average waiting time is 400 minutes, or over 6 hours.

While I've seen one report[1] that claims this, data from recent sources[2,3] give me median/mean times of around 50-180 minutes, which is much more reasonable.

Does anybody know which is right, and if there's a real paper that lists these findings? I'd be very surprised, as other commenters have noted, if such optimization work wasn't being done already.

[1] http://www.cnn.com/2011/HEALTH/01/13/emergency.room.ep/index...

[2] http://www.medicare.gov/hospitalcompare/Data/emergency-wait-...

[3] http://www.cdc.gov/nchs/data/databriefs/db102.htm

I had worked on analysis of data from National Highway Traffic Safety Administration (NHTSA), the number of traffic accidents based on factors such as weather, time of the day, month, year, weekdays, weekends etc and found very definitive patterns in the number of traffic accidents. Though that was post analysis, it does look quite interesting to use features such as weather, day of the week etc to predict the rush at the emergency rooms.

This is quite similar to the TACO system that started managing operations for all Taco Bell locations in the late 80's by predicting the number of customers, running a queuing model to determine the number of workers necessary, and using an integer programming model to schedule specific employees: http://zimmer.csufresno.edu/~sasanr/Teaching-Material/MIS/DS...

Queuing theory is fascinating and applicable to many real-life situations. For class, I built this queue simulator [1] that suggests the number of cashiers necessary for a restaurant (e.g. a Panera or Starbucks) based on the number of customers per hour, the average service time, and the goal customer checkout time by using an M/M/c [2] model:

http://queue.philipithomas.com/

[1] It's technically not a simulation because it solves the problem analytically, but here is a simulation-based approach I also wrote: https://github.com/philipithomas/M_M_1

[1] http://en.wikipedia.org/wiki/M/M/c_queue

The scariest part of the article is that he's cutting 10% off of the wait time, and that's 40 minutes.

This means that normal wait times were around 7 hours!

I worked in Suzhou for a month and I can tell you they are terrible at things like this. When I was there they had just began working on a subway, but they were doing construction all at once throughout the city. This meant that two lanes were closed on nearly every block just for construction. A ten minute bus ride would take an hour in the horrible traffic.

The researcher is from Suzhou, this article has nothing to do with Suzhou, however.

Oh, and you should have just walked.

Reply on news.ycombinator.com