Wednesday, January 08, 2014

The data are wrong. Our patients are sicker.

People in health care don't like it when numbers emerge that are uncomfortable.  Take these, issued today by the Massachusetts Health Policy Commission in its latest report on the drivers of the high cost of care in our state.

Variation, particularly when not correlated to quality of outcome, is particularly troublesome for some incumbents.  Academic medical centers often have their answer, but as the HPC explains, it doesn't hold water:

One oft-cited theory for the cause of this variation is that certain types of hospitals, such as those that teach physician residents and fellows, must incur additional expenses to support their mission. However, the difference in median expenses per discharge between teaching hospitals and all hospitals ($1,030) was less than the difference between individual teaching hospitals ($3,107 between the 75th percentile and 25th percentile teaching hospitals). Moreover, there were a number of teaching hospitals that incurred fewer expenses per discharge than the statewide all-hospital median of approximately $9,000 per discharge.

So perhaps the high cost ones will now revert to the usual squawking: "This isn't fair. The data are wrong.  Our patients are sicker."  Except here, the data are the best that could be available--all the claims for all the hospitals and all the payers in the state--even adjusted for wages.  And the acuity of patients across the spectrum of academic medical centers does not vary widely--but, just in case, the numbers are case-mix adjusted.

This report is a good step forward.  Now, if the HPC were to just put names under each column, instead of leaving them unmarked, it could take a major step forward in two of its own policy recommendations:

--Fostering a value-based market in which payers and providers openly compete to provide services and in which consumers and employers have the appropriate information and incentives to make high-value choices for their care and coverage options; and

--Enhancing transparency and data availability necessary for providers, payers, purchasers, and policymakers to successfully implement reforms and evaluate performance over time.


Anonymous said...

i have to agree; what is the point of issuing such a report if there are no actionable items in it? It would be a lot simpler to just put names on the graphs than go through the usual governmental responses of holding hearings, etc. In fact, it seems to me that has already been done in your state; hearings at which you testified, no?

nonlocal MD

Anonymous said...

Great post Paul. Needs to be in all States!

Barry Carol said...

Paul –

Is there any data available that would tell us how many employees per licensed bed work within each of these teaching hospitals and what their average occupancy rate is? Employees per licensed bed could serve as at least a rough or crude indicator of costs. There could be very significant differences though among hospitals in how many employees they may have working off site in hospital owned imaging centers, rehab facilities, physician practices and the like.

Paul Levy said...


I don't know, but my theory is that the dependent variable here is the cost. The independent variable is the market power of the hospital. Those that are able to extract more money from insurance companies, spend that money. That's the correlation I'd like to see.

Anonymous said...

My guess is that the highest cost facilities simply have the highest ratio of high cost specialists.

Dartmouth Atlas at the micro level.

Paul Levy said...

Maybe in other regions, but I don't think that's the case in MA, and especially in Boston. The academic centers here are rife with specialists.

Anonymous said...

Are there really 16 "major" teaching hospitals in Massachusetts? Some teaching hospitals are essentially at the level of big community hospitals with residents, and others are not tertiary but "quartenary" care. The idea that the acuity is the same when comparing BWH and Faulkner, MGH and the VA, Mt Auburn and BIDMC, seems like a stretch.

Now, if the variation doesn't sort that way, then that's a better argument. If Faulkner, for instance, costs more than BIDMC, then that's bogus. But I'd assume that BWH costs more than Faulkner--it _should_. And not sure how they're adjusting, but for instance, if you're going by ICD9 code to adjust the case mix, then "leukemia" at Mt Auburn does not equal "leukemia" at BWH to take one glaringly obvious example.

Here's a list I found randomly, I think Cambridge might be included in a list of 16. Looking at the list, I think you'd be hard-pressed to argue that case mix is the same; so the question is partly, how good is the case mix adjustment? Some of these hospitals refer the more complicated people _within_ diagnoses to other hospitals on this list. (To take the most obvious example, the medically complex patient with chest pain who shows up in the Faulkner ED gets shipped to BWH; the simple patient who shows up in the BWH ED literally gets shipped down and admitted to the Faulkner. Complex surgical cases get transferred from Mt Auburn to BIDMC, etc etc.)

Beth Israel Deaconess Medical Center
Boston Medical Center
Brigham and Women's Hospital
Children's Hospital Boston
Faulkner Hospital
Massachusetts General Hospital
Tufts - New England Medical Center
St. Elizabeth's Medical Center
VA Boston Healthcare
Lahey Hospital & Medical Center
Mount Auburn Hospital
Berkshire Medical Center
Baystate Medical Center
St. Vincent Hospital - Worcester Medical Center
UMass Memorial Health Care

Paul Levy said...

Thanks for helping to reinforce my case! Both about the inevitability of people questioning the data--and the value of labeling the chart.

Barry Carol said...

I read recently that Boston has 15 teaching hospitals while the city’s population is only about 650,000 people. That’s quite a large cluster of healthcare expertise. I don’t know what percentage of the revenue these hospitals generate is from patients coming in from other countries and other regions of the U.S. but I doubt that it’s very high.

Teaching hospitals have inherently higher costs because of their education and research missions. However, there must be a huge amount of routine care getting performed in these hospitals that community hospitals could do perfectly well for substantially lower cost.

To the extent that the powerful hospital systems charge just as much for a service, test or procedure at their community hospitals as at the teaching hospital and insist on all or nothing contracting with insurers, those are issues that can and should be addressed by regulators. Insurers should also be able to contract with a powerful hospital for some services like organ transplants, sophisticated surgeries and cancer treatment but not for other services that can be performed more cost-effectively elsewhere.

I’ve long wanted to see price and quality transparency in healthcare and would certainly welcome information on costs with each hospital identified by name.

Paul Levy said...

There are no regulators on this matter in Massachusetts, Barry. Any such approach to regulation of actual hospital rates has been off the table for years. As I noted in 2011:

"The fastest way to raise hackles among Massachusetts hospitals, doctors, insurance companies, and even businesses is to suggest that a state rate-setting body would do a better job in setting payment rates between insurance companies and providers than the marketplace."

I suggested an approach back in 2009--assuming the state was considering a move to capitation:

"It seems to me that a logical conclusion that follows from these principles and recommendations is that the capitated rates to be collected by the providers should be set by state regulation rather than by the current negotiations and market forces at play in the state.

"So, it seems to me that the job of the state in this new order is to establish risk-adjusted annual health budgets for the different risk categories of the public; to weight them by geographic cost-of-living factors; and to apply the appropriate adders for medical education and the other factors mentioned above. This would then become the rate schedule to be used by every insurer with every primary care doctor in the state, regardless of what provider network that doctor was affiliated with."

And in 2012:

"The current pricing regime is characterized by secret negotiations between the dominant insurance company and the dominant provider group, which has led to persistent and pervasive over-pricing of that system's services. This argument is not solely about the cost of care at, and prices paid to, one academic medical center: It is about the price paid for care in the entire system (physicians and hospitals) of which that AMC is but a part. There is a need for independent oversight over parties that clearly are not interested in solving the problem."

Anonymous said...

If you are curious about case mix indices for some of our academic medical centers

2010 for all patients (not just medicare patients) This was Massachusetts data that the state stopped making available for free (I think some of the big medical centers were embarrassed by the data as shown by Paul)

A 2010 chart shown at cost hearings (and published here by Paul) showed that Tufts Med while have a similar case mix had costs more than 30% less than the partners hospitals (MGH and Brigham)

MGH 1.3929
Tufts Med 1.3794
Lahey 1.3761
Brigham 1.3297
Childrens 1.3312
UMass 1.25
BID 1.2224
Baystate 1.1618
St Eliz 1.1123
St Vincents 1.0686
BMC 1.0336

Note this came from the CHIA site before they stopped giving this data away.


from 2012 CMS data [NOTE: just for medicare patients] the case mix indices.

Brigham 2.066698
Tufts Med 1.931213
MGH 1.882797
Baystate 1.772911
BID 1.7632291
Lahey 1.679040
UMass 1.644376
St Eliz 1.577283
BMC 1.557713
St Vincents 1.484072

By the way from past views of cost data I think Baystate or Cambridge Health Alliance are the two lowest cost academic medical center/teaching hospitals. Baystate is closer to an academic medical center in terms of case mix. Its overall case mix has increased from 2010 per change in CMS upward recently. Cambridge Health Alliance is more of a community teaching hospital.

Anonymous said...

If health care costs were regulated it would be politically difficult to reduce medical spending.

In the U.S. we spend roughly 18% of GDP on healthcare. For all OECD countries the average is roughly 9%, Virtually all of those countries have universal care (we still don't) and many (most) OECD countries have better health outcomes and healthier longevity.

So we spend twice as much for an inferior product!

As we have all seen many times over the last few years, it is virtually impossible to get our national legislature to agree on a reduction in the percentage increase in spending (grow 4% per year not 5% per year)....let alone a reduction in spending (spending goes from X to 95% of X).

For those who recognize the benefit of government spending to help those in need, at the rate healthcare spending has been growing, it will crowd all governmental discretionary expenditures in the next generation unless a new path is taken. (It is unclear if the reduction in the growth of heatlh care spending last few years is a result of a slow economic recovery from the recession or something else)

If government spending on healthcare could be reduced to something like other OECD countries - say down to 12%, a 33% reduction from present levels that would vastly improve the health of our economy, increase job creation, turn around governmental finances at all levels (local, state, and fed) from deficit to surplus.

That can not be done by regulators. There would be too much political pressure to not cut funding to those who "care for the sick".

The only way our political system will achieve a reduction in health care spending is by injecting real competition.

As we have discussed before. That requires transparency of quality and costs from health providers and insurers to health consumers and mechanism to allow those consumers to choose high quality, low cost options in the market. Those options are available as we see in the chart shown above and in many charts Paul has shown for years from CHIA etc.

There are high quality low cost community hospitals and academic medical centers in Massachusetts. Consumers don't have an incentive to choose them.(So a high percentage of patients get care at the most expensive academic medical centers even for routine matters)

The health care payment system for consumers must be adjusted to make choosing high quality low cost providers - logical and financially necessary.

In the long run - if competition can be made to work it is a far better outcome than increased regulation.

Al Lewis said...

Comment copied from The Health Care Blog, which reprinted this post:

Wait a sec, you mean “Our patients are sicker” isn’t part of the Hippocratic Oath?

One time — just once in 30 years — I did hear a hospital CEO (Lake Forest Hospital in Illinois) tell me while giving me a hospital tour that his patients were nowhere near as sick as other people’s patients, and just after he said this we turned the corner onto a different unit, where we almost got run over by a patient in his hospital gown dribbling a basketball down the corridor.

Massachusetts has some of the country’s best data and it sounds like they put it to good use in this analysis, albeit blinded. Call me a conspiracy theorist but I suspect differences in cost are due to Partners’ antitrust-worthy market share.

Saurabh Jha said...

From THCB:

“So perhaps the high cost ones will now revert to the usual squawking: “This isn’t fair. The data are wrong. Our patients are sicker.”

Predict your opponent’s argument and you diminish its impact! Ahh that timeless arguing technique!

On a serious, and plausibly educational note (educational for me), how good are statistical techniques of risk adjusting?

I ask also because on another thread on Medicaid’s patients outcomes versus uninsured many stated that such comparisons cannot be made because risk-adjusted apples are still apples and can’t be compared to risk-adjusted oranges which are still oranges.

COI: no skin in this game.

Anonymous said...

The case mix indices shown above deny the argument, "our patients are sicker" at least in relation to Tufts Medical Center and Lahey Hospital and for medicare patients, Baystate.

The Partners Hospitals - MGH and Brigham & Womens - cost a great deal more than all three. Others studies done in recent years by the similar groups (like the Health Policy Commission or its predecessor's) here in Mass show quality among the top half and more of the states acute care hospitals is all broadly similar including for these academic medical centers.

Patients do not care that Partners hospitals (and Childrens) are much more expensive, since in most cases that added cost doesn't affect their wallets.

For patients to actively choose high quality low cost hospitals and medical groups, they need to benefit financially enough to overcome the "magnetism of Harvard".

[Note: both Partners hospitals are Harvard affiliates, as is Childrens which is even more expensive though much smaller and specialized and the next most successful hospital group after Partners is arguably the of BID, the other Harvard academic affiliate.]

Lynn Nicholas said...

Paul, while I certainly agree that there is waste in the healthcare system, I want to point out that there are also myriad efforts being made in Massachusetts hospitals to combat this. The data the HPC used to ‘extrapolate’ the cost trends for its report do not take those efficiency and improvement efforts into account. My blog post on MHA’s website goes in to more detail and gives some examples. Our current system has its high and low points and can be further streamlined and improved. But the waste now in the system was built over decades, and it’s going to take time and attention to tease it out. That process is already underway, and it should be acknowledged, even as we continue our efforts to make still greater improvements.

Anonymous said...

As for 'efforts to make still greater improvements', an analogy can be drawn to the current re-focus in clinical care from rewarding process improvements with incentives, to rewarding improvements in outcomes. We already know that our measuring gauges are not perfect and never will be (or never enough for some people) - but the focus on cost efficiency and waste needs to switch from 'efforts' to 'results'. I think the report indicates inadequate results. Good intentions are not enough, and 'wait for it' isn't either, anymore.
How about 'no excuse'?


Anonymous said...

As I posted in another article, there is a boston academic medical center that knows how to get cost efficiency.

Tufts Medical Center, working in concert with Lahey Hospital is the "most cost efficient academic medical center in the country" according to data from the University Health System Consortium. UHC is made up of 120 academic medical centers and their affiliated hospitals.

Tufts Medical Center & Lahey are much less expensive than their peers in academic medicine in the Boston area especially when adjusted for case mix.

Patients in Eastern Massachusetts have good high quality, low cost options for tertiary and quaternary care.

But the Partners hospitals - MGH and Brigham & Womens - and Childrens continue to take market share at much higher cost to patients, insurers and tax payers.

BID is somewhere in between. Not as expensive as Partners and Childrens (which is the most expensive of all), but more expensive than its other local competitors especially when adjusted for case mix.

There are those who focus on and get results.

But that is not enough without the proper financial incentives for patients and cost and quality transparency.