Hacker News new | ask | show | jobs
by jeanlucneptune 4581 days ago
Assuming this data is from the Medicare (insurance program for those >65) dataset released by the Centers for Medicare and Medicaid services earlier this year (http://go.cms.gov/1bowKJA) I have a couple of points to make:

1.) Not sure if you grabbed "charges" or "payments", but "charges" are what the providers (hospitals) billed to Medicare, and "payments" are what Medicare actually paid the providers. "Payments" would be the best indicator of "price".

2.) Regardless of whether you are using "charges" or "payments", it's pretty much irrelevant for the average person. Why? Because the patient never pays the full amount (except for the uninsured...see below).

3.) While you don't pay the full amount in most cases, you will have to pay something. However, what you actually pay varies WIDELY based on your health insurer (the "payor") and the design of the health insurance plan.

4.) Also note that every "payor" negotiates a different price, so the Medicare price does not equal the Commercial Insurance price which does not equal the Medicaid price which does not equal the price the uninsured person pays.

5.) Finally, the diagnostic/procedural terminology in this dataset is impenetrable to most people. For example, does anyone on HN know what "transient ischemia" is? And if you do, do you understand that you can have transient ischemia in many parts of your body (FTR, I'm an MD)?

Although I appreciate the effort and the clean visualization, I don't think this data is particularly useful to an individual trying to make health care decisions for reasons 1, 2, 3, 4, and 5 (and probably others that I'm missing).

The one place where this data may be useful on an individual level is in the case of the uninsured. When you have no insurance the hospital basically makes up a price (taken from something called the "chargemaster") that is way more than what commercial insurance would pay, and WAY WAY more than what Medicare would pay. Having access to this data might help an uninsured person negotiate a lower price when the hospital comes after you with a giant bill. Medicare is a pretty solid standard to compare to - i.e. "you're charging me 5X, whereas you would only charge Medicare X".

On a societal level I think this data is also extremely valuable because now you can start to analyze pricing disparities across procedure types and geographies, which is really helpful. You can also put pressure on providers who are gouging individuals paying sticker price (i.e. the uninsured), which is important given that medical bills drive the majority of personal bankruptcies.

We ran a competition on this topic recently at Health 2.0 and you can learn more about the issue and the datasets on our site. You can also view other visualizations of this very important data:

http://www.health2con.com/devchallenge/rwjf-hospital-price-t...

Transparency in health pricing is a very important topic and something you'll hear a lot more about in the future. Happy to discuss further with anyone who wants to learn more (@jeanlucneptune, jeanlucneptune@gmail.com).

2 comments

Nice! Me and my friend made this app back in a few weeks ago, and here is the link.http://pricemaps.betterdoctor.com/#/

1. We used the 'Average Payment' for the actual number shown on the site. 2. Even though it's irrelevant, but it actually goes back to you by your co-pay and other expenses to you. Not 100% sure though. The point in our app is that, we want to compare the price vs quality in the US top hospitals. They actually vary a lot. 3,4 I agree. Co-pay is what you pay mostly. 5. We cleaned up the procedure names manually, and matched the procedure related specialties to the hospital specialty ranking made by US News, and it gives much better perspective to the users.

Nice to see people doing more health related stuff out there!

I like the betterdoctor price maps! Too bad you didn't have a chance to enter our competition, because you might have been a winner. By point:

1.) Good choice 2.) You need to consider co-pays as well as allowable charges, deductibles, co-insurance (if applicable), and benefit maximums. 2b.) The quality data out there sucks (hard to measure quality in HC) and is hard to find. 3.) Cool 4.) Ibid 5.) Clever

HC now is like the internet in the early 90s. Huge opportunities that will hopefully attract talented developers, technologists, and entrepreneurs.

Thanks!

Yes, we would love to take our app to the next stage by providing more comparison functionalities. One of the reason why we skipped quite a bit of the metric is that we don't want to make the users to be too much confused with choices.

This is one of the problems with the current health care system. There are too many choices, and there is no simple way to do any sort of quick decision like renting care, etc.

totally agree! my friend an me started in August on a similar mission http://www.pricepain.com - even tried YCombinator this winter but did not make it. If you checkout our /why section we collected a lot of recent articles and links on the subject
Nice! Looks pretty cool.

How did you get these data? Is there any public source for this? We are using US News Ranking & CMS charge data.

Best, developers behind the BetterDoctor PriceMaps

"My friends and I" not "me and my friends"

Yes, call me a Grammar Nazi Troll, but I see a shocking amount of this most fundamental error all over forums where otherwise intelligent, educated people discuss things.

You sound ignorant when you make the same mistakes as my 12-yo daughter. I can tell you make it because you are typing as you speak, which means you are in public, in meetings, and some day in presentations using bad English.

People will judge you by how you speak. I'm not being a troll, I'm helping you have a better future.

The rule is to use the same pronoun you'd use as if you didn't have two parties involved, right? "My friends and I did X" is right because "I did X" is, and "Me and my friends did X" is wrong because "Me did X" is obviously wrong.
Sorry guys, not a native here. I am sorry if I upset you with my poor English.

I learnt my lesson :)

It's not something we are "upset" about, and considering the level of English you used throughout the rest of your post and response, I'm not sure "not a native" is a valid excuse.

I suspect if you aren't a native, the reason you say "my friends and me" is because it's a bad habit you've picked up from natives, like typing "gonna". I think anyone who can make a sentence such as,

"We cleaned up the procedure names manually, and matched the procedure related specialties to the hospital specialty ranking made by US News, and it gives much better perspective to the users"

is far beyond making English 101 mistakes confusing objects and subjects of sentences. Actually, in 15 years of on and off English teaching, I have run into very few non-natives who have that issue, since they had to actually study it. I find you to be somewhat unique.

In any case, there are plenty of natives who consistently say write and say this, so hopefully a few of them will take it to heart.

Thanks :) I fix it then :)
I like the idea and the effort.

One thing I hate about it, though:

Again, third party javascripts are loaded in the background (Google, as usual, for example).

Come on people, you know this tracks everybody around the entire Web, stop using those scripts (is it so hard to use local js scripts? For example, find local "Web analytics" here: https://prism-break.org/).

What you're asking your users to do here is: telling Google (and the NSA, etc.) about your health problems. Not smart. At all.