Many people don’t know that medical prices vary greatly. This is one way of looking at that, with an interactive aspect that lets you feel the differences.
What is the PriceMap?
Did you know that medical prices vary widely? This interactive map is based on prices paid by the U.S. government to hospitals under the Medicare program, for people who are over 65 or disabled.
Medicare pays well below market price, so this would not necessarily be the cost that you would get walking in the door. That said, these numbers carry some correlation to market prices. Insurance companies and providers often use Medicare prices as a negotiating point, and this is the most complete government dataset we could find.
How should this data be used? However you want to use it. Some people recommend that patients ask prices in advance; others recommend that patients negotiate with providers, especially out-of-network ones, on the basis of Medicare data. The data also give insight into costs and how they vary. On another level, though, it’s hard to draw concrete conclusions from these numbers, because they don’t clearly ask and answer certain questions (see below).
Find something interesting? Drop us a note at info [at] clearheathcosts [dot] com.
Where did you get this information?
It comes from a government database of prices paid to providers of these services for Medicare patients in the facilities listed. It’s housed on the site for Medicare data, repository of at least 900 datasets.
This database displays the average payment by Medicare (the federal program for people over 65 and the disabled) for each medical condition or surgical procedure at a hospital. It was created March 8, 2010, and updated April 26, 2011, according to the site.
The government also makes other databases of prices available.
While the prices are not exactly market prices – Medicare pays a modest percentage of what hospitals, doctors and other providers charge – the Medicare rate bears some correlation to the market rate. It is often used as a point of discussion or departure when insurance companies negotiate with providers for reimbursement rates, and is the closest thing to a regulated price that exists in this confusing marketplace.
Did you modify the data?
We removed procedures showing a <$11 price tag — these numbers seemed to be erroneous and distracting. This does highlight serious data sanitation problems or a lack of proper communication about the data elements.
We’ve also lumped together all of the data with coding showing complications, no complications and major complications together. (The coding in this case comes under the Diagnosis Related Group or D.R.G., a system devised by the government to group together cases expected to have similar hospital resource use.) We did this largely because a total expected cost of going to a facility for a procedure is predicted by how likely they are to charge you for complications. We understand that not everyone will agree with our methodology — of course some facilities will have patients with more complications — but we wanted to standardize somehow, and this is the means we chose. We’d be happy to think of different ways of doing this.
If you have information regarding this please contact us.
Why are you showing me this?
Because it’s graphic representation of how chaotic pricing is in the health-care marketplace. We’re working to bring any kind of transparency we can to the market. While these numbers don’t offer a clear consumer guide (it’s not like looking up the cheapest car in the area), they give you a sense of scale – can prices really vary that much? — and it does so in a user-friendly way.
No, really, why are you showing me this?
Because we worked with the data platform folks over at Lucky Sort to build this.
Why would I want to know what Medicare pays to get a myocardial infarction treated?
We used this database for this particular interactive map because it was the most complete and easy-to-access publicly available data set we could find. The engine is designed so we can easily swap in other kinds of data.
Again, we didn’t design this as a shopping guide, but as something you could look at as a way of thinking about prices in the health-care marketplace.
So if I wanted to have my surgery in Utah instead of New York, I would save a bunch of money?
Well, you could think of it that way. We’re not suggesting that you choose to have your surgery in Utah because we don’t give medical advice.
Why are the numbers so random? In some places the same procedure in the same city can be a few hundred dollars in one place and in the six figures at another place. What’s going on here?
We don’t understand all the variations ourselves. Public data is often messy. We’re not certain the people in the government who collected these numbers understand the variations.
We think some of this is due to collection errors, some to input errors, and some to simple misunderstandings. Data collection on this scale is challenging at best, and we have great sympathy for anyone trying to wrestle this many numbers to the ground.
We should also note that the National Academy of Sciences issued a report in June 2011 saying that Medicare payments are based on inaccurate, unreliable data. The Medicare price report said the system of paying doctors and hospitals is flawed, and a new plan for reimbursement is expected.
Explain to me again how Medicare prices relate to the regular market rate. I’m confused.
One good way to understand this issue is to ask the naïve question: why such a huge price variation?
Isn’t it true that these numbers might mean nothing? I mean, if they’re not risk-standardized, or adjusted for age, or whatever.
We can’t know all the underlying factors. We used this database because this is some of the best data that’s available in the world of health care; since the system’s so complicated, and since the providers and insurers and others don’t necessarily benefit from openness, the government data is what we’ve got.
It is true, though, that it’s not risk-standardized that we know of (excluding the most expensive cases and adjusting the data for characteristics of the population). A friend who’s a doctor at Mount Sinai in New York City points out that income is a factor; poor patients tend to cost more. “Hospital X might seem to be more expensive because it has a sicker population with less preventive care and more complications perhaps,” she said.
Age is a factor, too. And the nature of the facility: a referral center, or a big teaching hospital, will have more tough cases than a small community hospital, which will not treat tough cases and will instead send them on to a bigger facility. “If you arrive and are complicated, you’ll get transferred,” she said. “Looking at cost, you almost can’t compare the two, since they’re doing such different things.”
Our friend at Mount Sinai also points out that the outcomes depend on secondary diagnoses. If a person with high blood pressure also has diabetes, that makes the whole picture more complicated.
And, she notes, there are more definitions of success than the label “least expensive.” In other words, she said: “You don’t care about price, but you do care about cost-effectiveness. How much are you getting for your money.”
Beyond that, she points out, it’s a bit hard with such a big overarching diagnosis to say what’s really at work here. If we had been able to compare the price of an x-ray at each hospital, that would indeed be quite illuminating.
But here, you have specific hospitals (one of the reasons we chose this database) and big complicated diagnoses. Each patient getting treatment in a hospital has received different procedures. (This must be true, because Medicare pays what Medicare pays for a procedure.)
Therefore, the question becomes more like this: are patients at More Expensive Hospital A getting more treatment than those at Less Expensive Hospital B because 1) patients at Hospital A are sicker, or 2) doctors are delivering more care at Hospital A because that’s the culture of the hospital, or 3) Hospital A has an overuse problem, or 4) none of the above?
That would be hard to know.
Are you planning to do more of this?
We’d like to. We did this first as a sample of what one might do to visualize the pricing variations, with a database that was close to hand. If you have any suggestions, please let us know at info (at) clearhealthcosts.com.
How can I help?
If you know of any databases that we can wrangle like this, let us know.
If you have any information about pricing disparities that you’d like to share, let us know.
If you have any other suggestions, let us know.
We’re at info (at) clearhealthcosts.com.