Are More Expensive Hospitals Better?

With the cost of pretty much everything on the rise lately, I’ve been thinking a lot about how price tags and quality relate to each other — or sometimes don’t.

Recently, my podcasting colleague and former teacher, economist Steve Levitt, told me a story that he’d told on Freakonomics Radio way back in 2010, called “Do More Expensive Wines Taste Better?” He gave me the short version:

Steve LEVITT: So, Bapu, I was a Junior Fellow in the Harvard Society of Fellows, which is an amazing opportunity. It’s three years as a young researcher — no obligations, except to show up to fancy dinners and to a weekly wine tasting. That’s your total obligation. And they have this enormous budget for wine. I don’t care about wine. I can’t tell a good wine from a bad wine. So, being an economist, I said, “Hey, why don’t we do a blind wine tasting once just to be a lot of fun?” And they said, “Oh, that would be fun — why don’t we go ahead and do that!” What they didn’t know is that I had a trick up my sleeve. The wine cellar that we kept had all the very fancy wines. And I had our wine steward pick out some excellent examples of some particular varietal. And then on the way to the wine tasting, I bought the cheapest bottle of wine I could find at the liquor store of that same kind of grape.

He put that cheap wine in one decanter. He put a fancy wine in another decanter. And then he put a second fancy wine into two of the other decanters — these were now identical. So, altogether —

LEVITT: Four decanters, three wines. Okay. And then I had people taste each of those four decanters and make comments about how good it was.

The Harvard Fellows sniffed and sipped. I imagine they savored, they swirled. They contemplated a whole range of appropriate adjectives. And then they wrote down their reviews and ratings for each of the four decanters.

LEVITT: And so, then I — I quickly went in the other room and I crunched the numbers. And it turned out that these wine snobs — not only did they not have preferences for the expensive wine, but they had no idea when they were drinking the same wine twice that it was the same wine! Because they rated it very differently — as differently as they rated the different bottles of wine. So, I triumphantly emerged from the room, and I reported my results to the people around. And it went very badly.

Well, some folks took it in stride, but not most. One fancy academic insisted afterwards that, actually, he had a cold.

LEVITT: I had really — I had not anticipated that in this group of academics the search for truth was not paramount when it came to wine.

From the Freakonomics Radio Network, this is Freakonomics, M.D.

I’m Bapu Jena. I’m an economist and I’m also a medical doctor. Each episode, I dissect an interesting question at the sweet spot between health and economics. Today is actually not about wine, but something else. When it comes to our health care, do prices tell us anything about quality?

Ateev MEHTROTRA: The No. 1 financial concern for Americans is not their education, food, housing. It is healthcare.

And, if you’re a healthcare economist searching for the Holy Grail, where the price of care is transparent and people pick their health care based, in part, on that information, you might want to be careful what you wish for.

Zack COOPER: You know, one of the really scary parts for me about this paper was I didn’t want this paper to be the excuse used by every hospital with high prices.

Back to my conversation with Steve Levitt.

Bapu JENA: The reason I wanted to hear that story is because it makes me think of a problem that we have in healthcare. So, let me ask you, I’m sure you go on — you shop on Amazon — and when I go on Amazon, let’s say I’m trying to buy, like, some Tupperware — I look at the reviews, I look at the prices. And, you know, one thing you’re probably struck by, just tons of information about the quality of the products that you’re going to buy. There’s been this movement to try and tell consumers — in this case patients — what the prices are of the different hospitals that they may go to, or the different types of healthcare services they may get. Like, if you knew the precise quality of every hospital, or every single place that could give you a flu vaccine, would you be more inclined to use price in those settings?

LEVITT: If I’m going to get the flu vaccine? Sure. Maybe I’d rather have a flu vaccine for $7.99 instead of for $12.99.

JENA: Your story made me think, like, if you tell people the prices of wine bottles or healthcare services, are they going to come to similar conclusions?

MEHROTRA: What is the rule — you always choose the second cheapest? Is that the rule? I always hear that, that people will choose the second cheapest wine.

That’s another colleague of mine — this time from my day job.

MEHROTRA: My name is Ateev Mehrotra. I’m a professor of healthcare policy and medicine at Harvard Medical School.

Ateev studies the relationship between price and quality in healthcare services.

MEHROTRA: The No. 1 financial concern for Americans is not their education, food, housing. It is healthcare. And so, it’s not unreasonable to think that there would be a lot of interest in this kind of information, so that they could choose where to go.

To shed light on the question, Ateev has worked on a lot of studies about the price transparency tools that currently exist for healthcare. He’s interested in how Americans interact with them when shopping for their care.

So, sure — on Amazon you can see, for example, if the iPhone case that you’re thinking about buying is being offered at a competitive price. You can figure out if other buyers have found the same case sturdy and attractive. If you’re looking on Amazon Marketplace, you can see if the seller is known to ship quickly and reliably. In other words, at the end, you have all the information you need to hit the purchase button with confidence. But should we expect the same model to work with healthcare? If we give people the information, will they act on price?

TRUMP: The consumer is very powerful, and this is going to make them more powerful.

This is a Presidential press event that Donald Trump did back in November of 2019.

TRUMP: So, I signed, as you know, an Executive Order — historic. And we’re requiring price transparency in healthcare, forcing companies to compete for your business. Our goal was to give patients the knowledge they need about the real price of healthcare services. They’ll be able to check them, compare them, go to different locations — so they can shop for the highest quality care at the lowest cost.

On the day the order was fully rolled out, Trump tweeted: “Enjoy all the extra money you will have.” Ateev Mehrotra understands where the enthusiasm came from.

MEHROTRA: This idea of consumer-directed health care has been part of the health policy debate for decades now.

One reason is especially important.

MEHROTRA: There is tremendous, tremendous price variation within communities. For an M.R.I. of the lower leg or a colonoscopy or any procedure, you can sometimes see 2-, 3-, 4-fold differences in the price from one hospital to the other. So, there is the potential to save a tremendous amount of money by choosing selectively.

Ateev and his colleagues have studied the impact of these sorts of tools, which inform patients about the prices of different types of care they may need.

MEHROTRA: Many of these tools are introduced by employers. And so, we will compare them to a set of employers that look otherwise similar, but did not introduce the tools. We’ll compare all the employees of the intervention group versus the employees of the other group — the control group. And we’ll look over time to see the changes in where they get care and, for those who were offered the tool, how they use the tools.

To make the approach as narrow and controlled as possible, Ateev looks at what he calls “shoppable services” — like radiology, or lab tests — medical procedures that many people believe are basically the same wherever you go.

MEHROTRA: First, it’s ideally something in healthcare where the average American doesn’t really perceive a quality difference. So, if someone’s going to operate on you, you’re really worried about the quality differences. But if you go for a basic laboratory tests — a complete blood count or C.B.C., as we call it in healthcare — nah, it doesn’t really matter that much.

LEVITT: We’re talking about something like an M.R.I., something non-invasive. Something where a machine gives out a reading and then I can take that reading and go find the best person to read the M.R.I. anywhere?

Steve Levitt again.

LEVITT: Sure. I’d rather have that M.R.I. done more cheaply than more expensively.

There’s more to Ateev’s research, though.

MEHROTRA: The second part is that ideally that service is less than the deductible. For a lot of healthcare — say a knee surgery — it’s way above the deductible. So, I’m going to have to pay the full deductible or whatever’s left on my deductible for the year, no matter what. And then all the price differences are going to be borne by the health plan for me. And so, I don’t really care.

Studies show that people will choose to save money on straightforward things, like buying generic drugs. That’s because typically there are clear savings — with no sacrifices or risks. But Ateev and his colleagues have done a lot of work over the last eight years that reveals something different.

MEHROTRA: In the studies that we have done on the impact of price transparency tools, in net we have not seen any substantive impact on both where patients get care and overall spending.

Let me say that again because it’s important. Giving people information about the prices of medical care doesn’t seem to lead them to “shop with their feet” — to find the cheapest options for care. So, why aren’t people using these tools?

MEHROTRA: It’s the million-dollar question. Why isn’t this really working?

No one — including Ateev — is sure yet. But his research has identified three clear takeaways. First, most people don’t know that price transparency tools exist. But even if they do:

MEHROTRA: Depending on the study, up to maybe 12% of patients who are offered such a tool used it. And after a couple of months, they forgot about the tool.

But here’s the biggest surprise: When folks do use the price transparency tools — on average, those people do not end up choosing the less expensive options.

So, in other words, even when people have good information about the different prices they may be charged for a given lab, radiology exam, or procedure, they often don’t pick the place with the lowest price. Maybe that wouldn’t come as a surprise to you when we’re talking about a specialized procedure, like a colonoscopy. Because — well, you know — who wants a bargain-basement colonoscopy? But this lack of shopping based on price is also true for things that are far less specialized and that are much more uniform, like lab tests or imaging.

It’s hard to know exactly why that is. But one thing is: when we’re getting medical care, most of us are primarily concerned with quality. We’re far more likely to follow the referrals from our doctors, or other people we trust, even if that means paying more.

And the reason that price comparison tools might not steer people to picking the lowest price option is because those same tools don’t really help us assess quality, at least in a way that’s convincing to people. I mean, look — it’s a lot easier to find reliable reviews of iPhone cases or Tupperware than to figure out which lab or imaging facility will do the best job for us.

MEHROTRA: Most of these price transparency tools do not have quality information that is very useful for the for the patient. I don’t think that price transparency initiatives have a lot of legs in the long term.

So, in the absence of good information on quality, maybe some people look at price as a signifier of quality. In that case, they might even choose to spend more. After all —

MEHROTRA: Almost everywhere in your life, there is some relationship between price and quality.

LEVITT: If I buy a car and one car costs $50,000 and another car costs $550, I can almost guarantee you that the $50,000 car is going to be much better.

Coming up, let’s look at one of those bigger ticket items. Are higher-priced hospitals any better?

COOPER: I think it’s very, very telling that most of the stuff we buy, we’ve got better information on price and quality — than we do for healthcare.

We’ll talk about some new evidence — after the break. This is Freakonomics, M.D.

*          *          *

I’m Bapu Jena, and earlier in the episode you heard me chatting with Steve Levitt about the relationship between price and quality.

JENA: So, let me ask you one question: do you have a prediction of whether or not higher price hospitals would be better in terms of measurable outcomes? Because there’s a paper that looks at this and I’m curious to see if you can predict what they found.

LEVITT: I would not expect that price would be a very good indicator of quality. That would be my hunch.

This research paper I asked Levitt about was released just this past February by the National Bureau of Economic Research. It’s called “Do Higher-Priced Hospitals Deliver Higher-Quality Care?”

COOPER: Potentially the only paper on the topic, I think? Yeah. A copious literature!

That’s Zack Cooper. He’s a health economist at Yale and a coauthor of the study.

COOPER: The hospital sector is an enormous share of the economy: 6%. So, it’s huge amounts of money — really important. And we’ve seen a problem with hospital prices in the U.S. They’ve been going up quite a bit over time. They vary a lot within and across regions — even for things where we don’t think quality varies, like M.R.I. scans. So, the goal was, broadly, to know whether patients get better outcomes when they go to higher priced hospitals.

“Better outcomes” is a simple term — but it’s complex to measure, if you think about it. A good outcome could mean you recover more quickly, more completely, with fewer long-term side effects. Or it could mean that you had a really good experience with the care that was provided to you. Or both. People vary in what they value more. So, Zack and his coauthors zeroed in on the most basic element that everyone cares about: Did the patient leave the hospital alive?

But they faced another problem, too, in how they designed their study: patients who are treated at higher price hospitals might be different from those treated at lower price hospitals.

COOPER: Patients aren’t, sort of, randomly assigned to hospitals. Their choices are real world choices — where we might be worried that, for example, sicker patients go to better and higher–priced facilities —

— or that wealthier or higher-educated people end up at higher-priced hospitals. This could bias their results. To get around this problem, they focused on emergency room visits. First of all, ER trips are more likely to be medically urgent — there’s a possibility the patient might actually die. Second of all, the patient is less likely to have selected the hospital themself because there’s usually an ambulance that picked up the patient and brought them in. Zack calls this the “secret sauce.”

COOPER: Ambulance companies are largely randomly assigned to calls. It just happens to be which car is near you at the moment.

That means that, even if an ambulance company has preferences for which facilities they deliver to, it’s still random which ambulance company is going to have a vehicle closest to a person-in-need at any moment. So, the choice of destination hospital is randomly selected by how all the ambulances, from all the companies, are distributed around town.

Zack and his colleagues knew the outcomes of patients brought to the emergency room and admitted to the hospital — who lived, who died — and they knew the hospital preferences of the ambulances that picked those patients up. Their goal was to figure out a hospital’s mortality rate based on only those patients who were brought to that hospital randomly. And then to use that mortality rate as a signal of a hospital’s quality. More patients dying in a particular hospital meant lower quality, and vice versa.

At the end of all this, what Zack and his friends discovered about price and quality was closely connected to another element: market competition.

The US government regulates the prices charged for Medicare and Medicaid patients. But for the privately-insured, hospitals negotiate prices with insurers — and, as a result, these prices are arguably more market-driven. In most areas of commerce, we rely on business competition to generate competitive prices. But for the last 30 years, the hospital sector has been acting a little differently.

COOPER: We’ve seen waves of mergers that have left many markets highly concentrated. A region that’s highly concentrated — a purely, you know, monopoly market — the hospital only has one competitor. In an unconcentrated market, think of a hospital that’s surrounded by other similar hospitals and no single provider has a dominant market position.

Zack and his coauthors took every hospital in the country, one-by-one, and they looked at the area surrounding each.

COOPER: And we said, “We’re going to travel a distance of 30 minutes travel time. And we’re going to see how many competitors and what the market shares are of different hospitals within those 30-minute travel times around each hospital.”

There’s now a live document where you can look up the measure of consolidation for every hospital in the country. We’ve linked to that document on our show page at freakonomics.com.

COOPER: I think part of the reason we put this data out there is so that folks can go have a look. And then have a think about, you know, what that means about the type of care they’re getting — and the price they’re paying. The hospitals didn’t love that.

The measure of market consolidation is known as the Herfindahl-Hirschman Index — let’s call it H.H.I. A market with an H.H.I. of less than 1,500 is considered a competitive marketplace — multiple sellers are vying for your business. An H.H.I. above 2,500 is considered to be concentrated. It turns out that almost 70% of the hospitals in the US are located in highly concentrated markets with an H.H.I. greater than 4,000. And lots of hospitals in smaller cities can have an H.H.I. as high as 10,000, which implies a 100% market share.

So, what does this mean? It means that hospitals in these regions don’t face any local competition and their prices may reflect that.

COOPER: That’s just a pure monopoly.

So, after all that, what did Zack find?

COOPER: The first surprise was that going to these higher–priced hospitals in competitive markets really did lead to better outcomes. And the scale of that was pretty dramatic — you were 47% more likely to survive that visit. That’s an enormous amount. I mean, that’s — you know — if I had a, a device or if I had a drug that lowered hospital mortality by 47%, right? You’d be giving me a Nobel Prize. So, that was surprise one.

In competitive markets — where there were lots of hospitals competing with one another — price was a signal of quality. Higher priced hospitals had lower mortality.

Surprise two was that in competitive markets, the higher prices charged by the most expensive hospitals were also cost–effective.

COOPER: You know, we did the sort of awful thing that economists do where you, you measure the additional spending per life saved. And there’s just a link between price and quality that I think many of us, frankly, didn’t expect to see.

But the third surprise?

COOPER: In hospitals that weren’t facing competition, going to these high–priced facilities really jacked up your spending — so, a 52% increase in spending. And just didn’t lead to any difference in outcomes. And, given that half the country is in a market where their hospital doesn’t face competition, it tells you that there’s something going on that we should be very, very worried about.

Zack and his colleagues found that in markets that were not very competitive, higher priced hospitals led to greater hospital spending without any measurable reduction in mortality. So, according to Zack’s approach, in those markets, price was not a signal of quality. In more competitive markets, it was.

This might be because in less competitive markets, hospitals can sometimes have greater market power, which allows them to charge higher prices to insurers and patients. In those markets, price may simply be a reflection of a hospital’s ability to charge higher prices than it is a measure of quality.

COOPER: You know, one of the really scary parts for me about this paper — and I actually thought a lot about it before we released it — was I — I didn’t want this paper to be the excuse used by every hospital with high prices. And so, I think it’s important to say, like — this paper does not show that raising hospital’s prices is going to lead to more quality. The paper doesn’t show that cutting hospital’s prices is going to harm quality. What it really, I think, highlights is that — when hospitals face competition — we end up with a price that is related to quality.

The first lecture that I give to my undergraduate students at Harvard at the beginning of the semester is about quality of health care — how to define it, how to measure it. To do that, you need to understand what quality means in general.

Take your smartphone, for example. You want it to work well for your needs — have good reception, good sound quality, good internet access. You want your phone to be effective. You want it to work fast and be up-to-date — in other words, timely. You also don’t want it to blow up on a plane, so, it needs to be safe. And the apps I have on my phone are different from the ones you have on yours. Our phones are personalized. It’s also helpful if the phone’s affordable. And if only the wealthiest, most-educated people in society had access to phones, that might pose an issue of equity.

Turns out that those same measures of quality apply to healthcare. The Institute of Medicine, in one of its landmark reports called “Crossing the Quality Chasm,” defined high quality care as care that’s “effective, safe, efficient or affordable, timely, personalized, and equitable.”

So, now that we have a framework for thinking about quality, the next question I ask my students is how do you measure it? The standard way is to think about three things: the structure of care (whether a hospital has an I.C.U., for example), the process by which care is provided (so, how long does it take for antibiotics to be given to patients who have a pneumonia?), and, of course, outcomes (do patients live or die).

With larger and better data, economists and health policy researchers have been able to improve our measurement of quality, so much so that the federal government and insurers actually use these measures to try and assess the quality of hospitals and doctors. But these methods aren’t perfect and a lot of people, including some who’ve been in this show, are working to improve them.

So, with all that said, how exactly do our two economists, Steve Levitt and Zack Cooper, shop for a hospital, if they need one?

LEVITT: I’m not a person who has a lot of experience, personally, with going to hospitals. I’ll tell you one thing I would not do — would be to try to save money on a hospital.

COOPER: I would look for a hospital that does a lot of what I’m getting. And I would probably ask a handful of physicians who I trusted about where they would go. But I can tell you where I wouldn’t be going. I wouldn’t be going to any hospital that – that is shielded from competition.

There you have it. That’s it for today’s show. If you enjoyed this episode, check out an earlier one called “How Can You Choose the Best Doctor?”

Coming up next week on Freakonomics, M.D. —

What do tap dancing and surgery have in common?

Katherine MOORE: You don’t have to think about it. You just do it.

Justin DIMICK: It really is like time melts away.

So, both require a high level of focus. But what happens when this extreme concentration is broken, particularly in the operating room? And what does this mean for patients?

DIMICK: I thought you were going to ask me, “Should we have a policy that surgeons don’t operate on their birthdays?”

Our next episode will get into the effects of multitasking and distractions in the world of medicine. Thanks to all of you, for listening, writing in, and supporting the show. If you can, leave a review for Freakonomics, M.D. wherever you get your podcasts. It really helps us out.

Freakonomics, M.D. is part of the Freakonomics Radio Network, which also includes Freakonomics Radio, No Stupid Questions, and People I (Mostly) Admire. All our shows are produced by Stitcher and Renbud Radio. You can find us on Twitter and Instagram at @drbapupod. This episode was produced by Sarah Lilley and mixed by Eleanor Osborne. Our senior producer is Julie Kanfer. Our staff also includes Alison Craiglow, Greg Rippin, Gabriel Roth, Rebecca Lee Douglas, Morgan Levey, Zack Lapinski, Mary Diduch, Ryan Kelley, Jasmin Klinger, Lyric Bowditch, Jacob Clemente, Alina Kulman, and Stephen Dubner. Our music was composed by Luis Guerra. To find a transcript, links to research, and a newsletter sign-up, go to Freakonomics.com. If you like this show, or any other show in the Freakonomics Radio Network, please recommend it to your family and friends. That’s the best way to support the podcasts you love. As always, thanks for listening.

*          *          *

LEVITT: Never go to a hospital if you have any choice. Only when you become unconscious and other people take you to the hospital. Other than that, don’t ever set foot in a hospital.