Bitter Pills: In Defense of The CMS Move Toward Meaningful Billing

Last week in a bold and historic move (and a bitter pill indeed for many healthcare executives finding themselves explaining the obscure financial inner workings of a hospital) the HHS/CMS released the national charge master data. The release seems to be a response to Steven Brill’s epic TIME magazine article on medical billing: Bitter Pill: How Medical Bills are Killing Us.

In the article (well worth the read), Brill documented the stories of seven mostly underinsured people who were devastated by medical bills. Brill laid the blame squarely on the chargemaster, a list of obscure, non-public charges the hospitals kept, and only the naive and uninsured have to pay. The chargemaster is now open.

I wrote about Brill’s article when it came out in Lighting the Dark Corners of Health Information. Back then I wrote:

“…a good place to start would just be mandating more price transparency. In an era of ever-more high deductible care, how much can providers continue to obscure pricing? People are going to want to know. The lack of transparency seems unsustainable, and forcing price transparency seems like a great place to start.”

So, of course, I was happy to see this move by CMS. The data includes charges, which are not the prices that payers or most others seem to pay, but it is a good place to start.

What’s in the data? There are over 160,000 rows of data that include the hospital name, address, ID number, the number of discharges for the disease-related group (DRG), the “charge,” and the payment amount Medicare pays for each of the most common 100 procedures or DRGs. There’s now a flurry of debate over what the “charge” means, and we’ll get to that in a moment, but it’s good we’re actually having the debate over a real number.

It has been a week since the release of the data and, as of Friday, there were already 100,000 downloads. I downloaded a copy and am looking into data points to associate with the charges. We may not know what they mean or what causes or rational they have, but we can begin to see correlations.

As you might expect, there are obvious correlations to geography, for instance. The reviews are starting to come in on what can be gleaned from the data, but we’re already starting to see some interesting visualizations of the data. Florida and California seems to have higher charges than most areas.

Of course, there’s been quite a few who are not all that happy about the release.

In A Health Care Blog post by David Dranove, Distinguished Professor of Health Industry Management at Northwestern University’s Kellogg Graduate School of Management, calls the release “useless” and “meaningless.”

David Harlow calls it an inundation of “junk” and “too much of nothing.”

Paul Levy, former CEO of “a large Boston hospital” calls it “useless noise” just to name a few.

From what I’ve read, the chorus downplaying the significance of this release believe that the data:

  1. Is meaningless.
  2. Is nothing new.
  3. Doesn’t affect anybody, or, at least, not very many people.
  4. Will be used to further rig the system by seeing what competitors charge.
  5. Is inaccurate.
  6. Is not used for anything.
  7. Is not enough, it’s just data, not information.

The Information is Meaningless

Google dictionary defines meaningless as: having no “significance” and having “no purpose or reason.”

The numbers are not random. The hospitals have a way (however historic) at arriving at a number or amount for various procedures and a reason for sustaining and altering them. It’s just that nobody seems to know what it is, or those that do won’t admit it. It’s cloaked in darkness.

The data CMS released may not be considered complete “transparency,” but I’ll take translucent over opaque any day. These things have a way of getting cleaned up once it’s known the data will be public.

According to Jeffrey Young in Huffington Post, “The primary beneficiaries of such a change (deciding to bring charges down) would be the roughly 49 million Americans who are uninsured, as they are virtually the only people who can be asked to pay outright what the hospital is charging.”

Hospital trade associations agreed with this part.

So the numbers do mean something to the hospitals and even more to the uninsured patients. If the numbers are meaningless, why are they consistently so high? Is it to make the insurance companies, who pay somewhere between the charge and what Medicare pays, feel like they are getting a great deal? Is it a future point of negotiation?

Levy says that “very few” base the insurance payments on a percentage of chargemaster charges, but we have no idea what “very few” is. The release of this kind of data can push the ball further in the direction of finding out.

Sometimes we go from opaque to translucent before we get to transparent. Even if we just go from opaque to translucent, it’s an improvement over absolute darkness and can often be the first step towards clarity.

The Data is Not News, It’s Nothing New

The Medicare payment prices were publicly available, but not pieced together, and the charge data is new. You can argue to the charge’s veracity and meaning (now that it’s public), but at least now we can have the discussion.

Even so, 100,000 downloads is new. OK, maybe someone could have taken the time to piece together the same information with thousands of hours’ worth of work, or it could have been crowdsourced, but this sure helps speed up the process.

The fact that the government has provided the data has a few benefits:

  1. It raises awareness
  2. Provides more assurance that the data is real, that it’s trusted.
  3. It’s free.

You or I or the guy or girl next door could have pieced together some of this info, but it means a lot more when the gov’t does it. As the government is already collecting the information, it’s a huge benefit to society to release it. Why not? What purpose does it serve if it isn’t made public?

This Doesn’t Affect Anybody

This is the real shocker to me.

Everyone who’s taking the contrarian view seems to believe that real lives aren’t affected when uninsured or underinsured people receive medical bills. Many take the position that it only affects a very small amount of people. But, all the same, why would affecting even a small number of people with exorbitant charges be OK?

There were a few more than 1.2 million bankruptcy filings in the U.S. 2012. Estimates are that more than 60% are due to medical bills. For the uninsured, there couldn’t be anything much more significant or meaningful than the dollar amount that drives them into bankruptcy. They may not be able to pay, but that depends on the procedure.

Many uninsured are middle class, hard working and honest — if they could scrape together the resources to go across town to have it done for less, shouldn’t they have that option?

There’s a reddit post with more than 1,000 comments on this topic. This is a hot topic. People want to know.

They’ll Use it to Rig The System

Matthew Holt, in comments on Dranove’s article on The Health Care Blog, half-joked that five analysts at each hospital made up the 100,000 downloads, using the data to ensure that they would be charging the highest possible rate given their geography and competitors’ charges. Sad, but if you’ve worked in health care long enough, plausible.

But if the information really is meaningless, why bother?

Now that the data is public, it may be more difficult to adjust charges indiscriminantly.

It’s Inaccurate!

Among the benefits of a public strategy (you could also say “social” – getting more people involved) is increasing accuracy. Linus’ law states that “given enough eyeballs, all bugs are shallow,” meaning if enough people look at a problem (or an inaccuracy) the easier the solution becomes.

Opening the data to feedback makes the invisible visible and allows the opportunity determine accuracy. If the data is inaccurate, it allows the inaccuracies to be found and fixed and creates an incentive for hospitals to correct them. If the data is not open, there’s a lot less incentive to change it, and that’s really at the heart of this data release.

This puts the hospitals under increased scrutiny. The public now has a basis from which to ask questions and, by mixing the charge data with other data, come up with more questions that we can answer ourselves. We can begin to assess the meaning, importance and veracity of the information, which was impossible before.

Nobody Uses This Data For Anything

Aside from uninsured patients who might see these numbers on their bill, these numbers seem to be a starting point for negotiation for some private insurers. So, there’s a whole lot of incentive for the hospital to prevent reduction of the charges.

Again, if the data is unused, and had no meaning, there shouldn’t be any reason NOT to change it.

It’s Data, Not Information

The great thing about publicly releasing data is that the public now can turn it into information with context; to mix it with other data and find out what it means. Of course, this is impossible without the bold first step of releasing the initial data.


Here’s a simple question to ask someone who says the charge master data doesn’t matter: Great, then why don’t the hospitals just cut them? Why maintain them?

Until now, they haven’t had a reason to eliminate them. Let’s see what the reasons are now that the information is open.

Momentum for change in healthcare is stronger than perhaps any other industry. A trillion dollars a year is at stake and nobody’s going to let go of their share easily. Improvements won’t occur without a sustained effort by many people to keep prices public, so let’s continue to press for more. Amazing things happen when people have the information they need to make an informed decision.

I’ll repeat the quote I used in my “Dark Corners” piece. According to Nobel Prize Winner George Ackerloff (via Nate Silver):

“…in a market plagued by asymmetries of information, the quality of goods will decrease and the market will come to be dominated by crooked sellers and gullible or desperate buyers.”

That’s no way to run a health care system.

A data release is not just about the data, it’s a social event, it connects people and gives context to the data. In this respect, this event may have been one of the biggest health care announcements in a long time. People will talk, analyses will be done, and in the end we’ll all have a better idea about how this system works and what we can do to fix it.

Meaning is in the eye of the analyst. So let’s use this as a starting point to move toward meaningful billing. Let the analysis, and the release of data, continue.

…Hat tip to Michael Planchart for the idea for the title on Meaningful Billing. Thanks for Vince Kuraitis for suggesting the Paul Levy post and comments.

leonardMore from Leonard Kish


The following two tabs change content below.
Leonard is Principal and Co-Founder at VivaPhi, an agency that solves multi-disciplinary business problems involving data science, software, biomedical science, behavioral science, health care, product design, community development, marketing, consumer engagement and organizational design. He has been quoted in Forbes and other top-tier publications for thought leadership on patient and consumer engagement. In addition to his role at VivaPhi, he is Chair of the Marketing and Communications Group for the Collaborative Health Consortium. Prior to VivaPhi he held the position of Vice President of Operations at Capitis Healthcare International as well as executive positions with several startups. He started his career as a software requirements analyst on Qwest Communication’s highest priority IT project while earning a triad of advanced degrees from the University of Colorado. These included an MBA, a Master’s of Science in Information Systems and a Master’s in Biomedical Sciences (Thesis on System Dynamics in Parkinson’s Disease). Leonard earned a Bachelor’s in Zoology from Miami University in Oxford, Ohio. He’s interested in how systems evolve, and how to help them evolve, in a variety of unique contexts. Connect with Leonard: @leonardkish, LinkedIn and Google+

, , , ,

  • Pingback: In Defense of The CMS Move Toward Meaningful Bi...()

  • Michael Goodson

    The reference to the “national charge master data” is misleading as no such national database exists to the best of my knowledge. Each individual provider facility (hospital) has its own unique charge-master which contains charges that may or may not be related in some consistent way to the actual cost of the supplies or resources required to provide healthcare services.

    Should hospitals be able to markup their costs in any way they see fit? If a healthcare supply item cost the facility $50, is a 50% markup to $75.00 reasonable? Is a 5000% markup to $2500 reasonable? Isn’t the outrage over the differences in procedural costs the result of not knowing whether pricing is reasonable or not? We probably wouldn’t care if they were selling something we didn’t have to buy.

    If inconsistently applied or unreasonable markups are the means of keeping operating margins from moving too close to the red zone, don’t we have to ask ourselves why some hospitals provide high quality services less expensively than others? If quality outcomes are provided, do large discrepancies in charges suggest costly inefficiencies in patterns or care? Do they suggest excessive profits for shareholders? The data is not meaningless and, yes, pricing transparency seems like a great place to start.

    Those who believe that the cost of healthcare to individuals, insurance companies or the government should be whatever the market will bear, will have no motive to determine what their true costs are to deliver healthcare services and will continue to charge amounts that are difficult to correlate with the component variables of labor, supplies and overhead.

    Those who believe that the cost of healthcare to individuals, insurance companies or the government should more closely track the costs incurred to efficiently and effectively provide high quality clinical outcomes will be challenged to determine what that cost should be. Few hospitals know that answer.

    Finding the true costs of delivering high quality healthcare services often involves pulling data from a variety of different in-house computer systems that normally don’t exchange information and compiling it into a picture of the cost of treating specific illnesses by physician. This usually requires expensive data marts or data warehouses and a multidisciplinary team to examine the data – an expensive and tedious processes.

    Perhaps it’s just easier to make large mark-ups on supplies and services until facilities achieve the operating margins they require to stay in the black. After all, not all of their patients are Medicare patients. As long as clinical outcomes are high quality, does it really matter if these quality outcomes are obtained efficiently and cost effectively? Private insurers can pass the additional cost on in higher premiums. Self-payers can simply pay a higher price.

  • Leonard Kish

    CMS’s data release includes the charges that hospitals report to CMS, as well as the payment CMS provides. The charges should correspond to the provider’s chargemaster.

    Indeed, its hard to have price-pressure if nobody knows the price. Maybe markups are a way to stay in the black. Now that more data is public, we can begin to ask these questions.

  • distinguished home care services

    My brother recommended I may like this blog. He was totally right. This publish truly made my day. You can not believe just how a lot time I had spent for this information! Thank you!

  • Pingback: Hospital Chargemasters and Open Data from CMS — The conversation continues()