Slashdot Mirror
Algorithm Contest Aims To Predict Health Problems
databuff writes "The April 4 launch of the $3 million Heritage Health Prize has been announced by the Heritage Provider Network, a network of doctors. The competition challenges data hackers to build algorithms that predict who will go to the hospital in the next year, so that preventative action can be taken. An algorithm might find that somebody with diabetes, hypertension and high cholesterol is a 90 per cent risk for hospitalization. Knowing this, it might be cheaper for an HMO to enroll them in an exercise program now rather than pay the likely hospital bill. The competition takes the same approach as the $1 million Netflix Prize, but solves a far more significant problem."
Is that now synonymous with programmers?
Just label them as a 'High Risk Candidate' and jack up their premiums 2-3x so they can no longer afford healthcare by the point at which they need service :P
then print "Not good"
If Google really cared they would fix Android Chrome to reflow text, instead of discriminating
-Tm
Support TBI Research: http://www.raisinhope.org
... I am sparing my mad skills for the 'predict the next to commit a crime' contest.
That's likely what will happen, but not necessarily the only result.
To lower company premiums, Safeway bribes employees to quit smoking and/or lose weight:
http://online.wsj.com/article/SB124476804026308603.html
There's no -1 for "I don't get it."
Might be a good thing for our health-care system, but I doubt that it would be implemented - too much red tape around privacy and the data is not centralized, so it could be hard to have access to this data.
Dear Obama Health Care Customer,
We have identified you as a high risk patient. We have taken the liberty of enrolling you in an exercise program which will start immediately. Please read the included forms and documentation pertaining to your health. If you decide to not follow this program then you can expect an increase in your personal health tax.
Thank you for your patronage.
The Netflix Prize was sponsored by the company who gained the most from it. In the same vein, this should then be sponsored by an HMO.
Damn_registrars has no butt-hole. Damn_registrars has no use for a butt-hole.
The purpose of insurance is cover random catastrophic expenses, not to let a person cost-shift known expenses to his or her neighbors. If your body is breaking down because you spend decades being fat and never did anything about it then foot the bill yourself. If you can't afford the cost of fixing a problem caused by your own lifestyle decisions then tough shit.
People don't want to end up in the hospital. Yet they do nothing in-spite of all the knowledge and information available. Obesity is increasing. You don't want to end up in the hospital? Keep your BMI < 22 (for average person! you know who you are!)
If your BPI is %gt; 25, you lead a sedentary life, you eat crap, then it is a good chance you will end up in the hospital or worse, ahead of your time.
"Knowing this, it might be cheaper for an HMO to enroll them in an exercise program" OR DROP THEIR COVERAGE!
Contest judge,"Why I'm perfectly healthy, why would you pick me?"
Fat Tony: *punch to the face of contest judge*
Contest judge,"Well I see your point, but that isn't exactly going to send me to the hospital."
Fat Tony: *draws a gun*
Contest Judge,"Ok ok, I'll go to the hospital, here's your money."
Fat Tony,"Who says I want your money?"
God spoke to me.
In a way, doctors are trained to ignore teh science. They start with someone who already has a problem, and treat the symptoms as best they can. Science has determined many of the causes, but they are not profitable for the oligarchy, so they train our doctors to sell us pills for the symptoms.
somebody with diabetes, hypertension and high cholesterol is a 90 per cent risk for hospitalization. Knowing this, it might be cheaper for an HMO to enroll them in an exercise program now rather than pay the likely hospital bill.
The Lipid Peroxidation chain reaction is a large part of what causes the diabetes, hypertension and [oxidized] cholesterol problems.
Lipid peroxidation is such a huge problem today because the western world switched its main sources of dietary fats from animals (mostly saturated butter/lard) to seeds (mostly polyunsaturated corn/soy/rapeseed/linseed)... And even the animals we eat aren't as healthy as they once were because now they subsists on seeds instead of grass/insects/etc.
This contest is a waste of time.
Learn the rules so you know how to break them properly.
www.teslabox.com
The competition challenges data hackers to build algorithms that predict who will go to the hospital in the next year, so that they can be dropped.
fixed.
Disconnect your television. Do your own research. Draw your own conclusions. They're probably lying. Don't be a sheep.
Why don't they just build a lottery machine to approximate it and sell another lottery game at convenience stores? Let people pay to run around going to all of the stores to find the algorithm which most correctly approximates the data.
the NPG electrode was replaced with carbon blac
Probability of visiting a hospital in the next year P
P = R
Where R = 1 if
- you ride a motorbike
- you're one of (bullfighter, boxer, lumberjack, cheerleader)
- you consume more than 30 hamburgers per week
- you consume more than 1 bottle of whiskey a day or equivalent
Now where's my 3 million?
"Health Insurance company offers $3 million dollar bounty for technical-sounding excuse to raise premiums"
Just measure carbohydrate intake. That correlates with all of the "diseases of civilization", like obesity, diabetes, heart disease, cancer, alzheimer's and other chronic diseases.
Nice that you linked to a (semi-) reputable source for the definition of lipid peroxidation. Can you do the same for the dietary and medical claims?
"In a way, doctors are trained to ignore teh science. They start with someone who already has a problem, and treat the symptoms as best they can. Science has determined many of the causes, but they are not profitable for the oligarchy, so they train our doctors to sell us pills for the symptoms"
that is completely false. While may Dr.s are not scientists, they still prefer to cure someone. It's a lie perpetrated by people whose own 'belief' aren't born out scientifically. Since they are so attached to them they invoke conspiracy that are nonsense.
"The Lipid Peroxidation [wikipedia.org] chain reaction is a large part of what causes the diabetes, hypertension and [oxidized] cholesterol problems.":
That is complete nonsense.
speaking of doctors and science:
http://www.sciencebasedmedicine.org/
The Kruger Dunning explains most post on
I like how everyone and their uncle suddenly runs a competition with a grand prize for the lucky one who meets the goal.
Need a new company logo? You could hire a graphic designer. Or you run a design competition, first prize an iPad. Much cheaper than paying someone, and the blogs will pick it up for free advertisement. Then you shell out the shinny gadget and look charitable.
Anyone interested in a basement-cleanup-competition?
Nice that you linked to a (semi-) reputable source for the definition of lipid peroxidation. Can you do the same for the dietary and medical claims?
Suitable Fats, Unsuitable Fats: Issues in Nutrition has a nice list of references. I'd start there.
Learn the rules so you know how to break them properly.
www.teslabox.com
While may Dr.s are not scientists, they still prefer to cure someone.
Of course, but their training is deficient. If you spend years learning the minutia of pharmacology and surgery, it's easy to "miss the forest for the trees". If your doctor's training only gives a cursory overview of the role of nutrition, AND the conventional wisdom about certain nutritional concepts is wrong (e.g. saturated fats were vilified so A.D.M. can make billions selling seed oils), your doctor is going to be biased for the things that he spent the majority of his training learning about.
That is complete nonsense.
Are you defending the mass consumption of rancid oils? I think you are. All "vegetable" oils are deodorized as a part of the production process.
It's hard to exclude oxygen from a biological system. Which leaves antioxidants, and it takes a lot of antioxidants to deal with the massive quantities of polyunsaturated oils in a person's weekly servings of A.D.M. soybean-oil-based Kraft salad dressing.
Your tone was pretty rude, so I'm going to have to refer you to my response to the anonymous poster, who politely asked for a substantiating link for the health claims.
Learn the rules so you know how to break them properly.
www.teslabox.com
It won't be long until they start euthanizing people before they become too expensive to treat.
Population - Population that will die = Chances of them making an algorithm that predicts Health Problems Correctly for Individuals.
Be seeing you...
Feh, I had some of that soybean-oil (and corn syrup)-based salad dressing tonight, and it was disgusting. Give me buttermilk+mayo+herbs over the 40 ingredients in this crap they provide under the branding "natural" something or other ranch dressing.
Smart individual buyers join a professional or other organization that has their own group.
IEEE makes you wait a year to get into their health insurance group. It's worth joining while you have a job so you have backup group insurance available.
John McAfee 'It was like that time I hired that Bangkok prostitute; to do my taxes, while I fucked my accountant'
Give me buttermilk+mayo+herbs over the 40 ingredients in this crap they provide under the branding "natural" something or other ranch dressing.
Yeah, except most mayo's are made with the same Soybean oil... I made mayo with olive oil once, but it was a little work.
Thanks for writing though. I definitely agree with your sentiment about "natural" branding for certain products. :)
Learn the rules so you know how to break them properly.
www.teslabox.com
I want to provide a "fuller" picture. I am a very fat and have a chronic illness. I am also an extremely capable data scientist. I consider myself a contributing member of society. If I stay healthy, I could plausibly develop algorithms that positively impact this health prize. However, if the American health system continues its downward trajectory, it won't matter in the grave.
"There are two problems with the healthcare reform law," said Jonathan Gluck, a senior executive at HPN. "We pay for quantity, not quality. The more services provided, the more the provider gets paid."
The second is that we can't count to two.
'The tyrant will always find pretext for his tyranny.' - Aesop's Fables
but Cayce had a strategy to flush out the gall bladder every so often. It involved eating raw apples and only raw apples for up to three days. On the last day of the apple diet, you'd take 3 tablespoons to 1/2 cup of extra virgin olive oil.
When the body subsists on a fat-free diet for long enough, the gall bladder stocks up on bile. When a large quantity of oil suddenly appears in the small intestine, the gallbladder's contents are dumped out to help break down the oil. If one has any stones,
Even though you don't have the gall bladder anymore, you'd probably still benefit from the pectin (fiber) and malic acid in the raw apples. With that said, some people don't do well with large quantities of apples, so talk to your doctor to see if this would be appropriate...
I guess I'm a "reform hater" too, but only because I think "reform" should make the medical system more effective. See my other post in this story, the western approach to medicine is completely broken.
Learn the rules so you know how to break them properly.
www.teslabox.com
By my algorithm you'd be forced to get another doctor that wasn't a waster or pay more. Okay, maybe it's simple, but it is effective which I can prove with personal experience.
Here's the algorithm:
If ((you can pass your doctor on the street) && (they don't recognize you)){
you = in trouble;
}else{
you = good;
}
Here's the anecdotal proof:
A few years ago life seemed to be really coming together. Happily married, just bought my favorite house in the area, was coding like a rock star and arguably the healthiest person in the office. Didn't smoke, drink or eat restaurant food. Also, was riding my bike 100 k / day round trip to work and getting really into carving out a perfect life putting one good day in after the next.
Like you, my doctor was a tosser but I figured, I'm fine.
A cold snap before Thanksgiving caught me off guard and got a slight chill. No big deal. But spent the long holiday in bed and went to see my doctor on Monday. He hardly looked at me, not tests just said I had strep gave me a mega dose of antibiotics and a note for my boss.
Swell. My neck it did. The pills did nothing. Called and to get another appointment. But just got more pills. I was feeling like crap. Felt like my doctor's nurse wasn't even relaying my messages. It was futile to rely on him so I went into the walk in clinic and saw a physician's assistant. He walked in and saw right away what was what. I saw it register in his eye the second he walked in the exam room after which he just blurted out he could see I had lymphoma. He had to back pedal a bit. Protocol and all. But arranged to have me see a doctor that afternoon.
Went home for lunch and they called from the hospital to see if I could come in sooner. So that could be another calculation, if you wait less time to be seen your fucked. Scratch that, they'll use that as an excuse to extend waiting.
Soon I was diagnosed Stage 4B. They don't say terminal any more. And why should they, you've gotta enough to deal with surgeries. Well they could have done em all at once because my surgeon could see what needed to be done but that's not approved by my apparently top notch insurance.
It just seemed so unreal. One day I'm making back of the envelope calculations from the numbers out of my Garmin that if I kept at it and focused I could be competitive enough to ride the pursuit in the next Olympics in London and then it turns out I just had an unnaturally high red blood cell count that any pro cyclist would covet because I had cancer. Did working out help me? I'd like to think so. Like to think maybe my training didn't get me to the Olympics but allowed me to survive. But maybe it did me in. My fat smoking sister didn't get cancer.
Another data you could use but would require surveillence to gather would be the number of doctors, nurses, cleaning ladies and orderlies who sublty as John Cleese wink wink nudge nudge advise you to not use more than one dose of Fentalyn patch as two would kill you, that's an indication that you might soon cease to be.
Another good data point they might want to focus on is the coincidence of spending a month in isolation because you've got like zero white blood cells and corresponding occurences for getting written up for tasks you failed to complete while on disability. Getting fired for taking too many sick days would be a potential data point to calculate in the algorithm as well.
Here's the thing. What ever happened to duty? I must have been focusing on doing my job and missed it but at some point the word lost meaning and worse is now the new euphamism for #2. I mean what kind of business model is it that puts the burden of performance on the customer? I had never heard of fentalyn before, but suspect that so many people know what it is is because they are as high as the arrogant shites who would abrocate responsibility to fluff up their bottom line long tail be damned.
So maybe my anecd
1: Get doctors to start to build an expert system for diagnosing patients &co using things like speech recognission to assist in recording absolutely everything that is said in the consultation, all questions and answers, symptoms, checks and results, all crys of pain etc...
2: Along side this do random, regular blood etc.. screens to get other hard data.
3: Also take some pictures and that kind of thing.
4: possibly contact IBM for their jepodoy winning kit too.
5: Combine all the expertise and amendements to the data that created the expert system, along with medication given and taken, diet details, recreational drug usage, blood screens, exercise, weight, stool samples whatever etc... across all patients all doctors and possibly the world.
building an expert system, that records absolutely everything, not just the things that usually get recorded. and allows doctors to put on what they did to delta the reported symptoms and the computer suggested treatment and diagnosis to further improve the computer assistance and expert system.
7: Data mine the hell out of it.
8: You now have yourself, real time or as near real time, ready to hand, scientific, empirical medical data and research like the best drugs trial in the world with sample sizes of millions and the knowledge and experteese of all the Mds and without screening people like autistics out of the equation.
Oh, as a bonus you get to get all those horrid side effects reported properly and get to find out exactly what the illicit drugs do to you.... and no lies or fabrication on medical records.
There you go, that's my algorythm.. where's the dollar. I'm gona spend it on a gramma and spelling nazi to check my posts.
thank God the internet isn't a human right.
Is this comparable to Netflix in terms of privacy, too? Because the second iteration of the Netflix Prize was postponed after a lawsuit charged that it breached customer privacy.
Instead of making it cheaper for a person to ignore his health, make it more expensive. Have high deductibles so that he's better off taking preventive measures than having to be hospitalized and paying a high deductible. Or I suppose you could just make it so hospital visits were guaranteed 100% fatal...
I did not RTFA, but it sounds like they are looking for conditional probabilities using well established models traversing some observed and some unknown. But since they are asking for an algorithm, I will suggest that they use Viterbi. There, please send me my check now; Viterbi already has all the Qualcomm money he needs to fill his swimming pool with gold coins.
"In a way, doctors are trained to ignore teh science."
As a person who works for an academic medical center (I'm a programmer working on a well known EMR. I do large amounts of data mining for research among other things), I take offense to this statement. Doctors are knee deep in the science, and many are helping to change the way the sciences (Aetiology, biology, computing sciences) interact to produce better outcomes for patients.
"They start with someone who already has a problem, and treat the symptoms as best they can. "
And this comes from the fact that most people do not go to the doctor until they are already sick. So, the best a doctor can do is treat the symptoms of their issue, such as obesity, until the patient changes the lifestyle that is causing the problem such as smoking/drinking/eating too much. At that time the illness would be gone and treatment can stop. If people would go in for regular check ups than doctors could do more to preempt the onset of disease. And if patients would actually listen to their doctors advice and stop doing those things that are causing disease...
" Science has determined many of the causes, but they are not profitable for the oligarchy, so they train our doctors to sell us pills for the symptoms."
Again, science has determined many of the causes, such as obesity, drug use, alcohol abuse, all things that only the person with those disease causing habits can stop without using "pills for the symptoms". As far as I know it's still illegal for a doctor to tie a patient down until they go through DTs.
And as far as profitability, last I checked hospitals are still businesses and need income to pay their bills.