Risk Transfer: Cost Reduction in Drag for Medicare

Health policy titans debating the actuarial links between clinical quality and cost trends.  Presidential candidates struggling to find the best "gotcha" Medicare catch-phrase to use against their enemies.  Progressives and conservatives in a death struggle over competing health care "visions" on the role of government.  Yet, underlying all this political cacophony is general agreement that Washington DC's involvement in health care and insurance has increased.

Except, says the contrarian Disease Management Care Blog, that is less true than meets the eye.  That's because in one key respect, Federal involvement in the world's largest health insurance program called "Medicare" is actually decreasing. When viewed through the lens of "risk transfer," there is an ironic bipartisan agreement that Medicare's risk needs to be transferred elsewhere.  The only real disagreement is over where it should go.

Recall that insurance is a contractual agreement to exchange "risk" for "money" a.k.a. risk transfer.  Insurance is a business that accepts and "pools" risk and uses collected premiums to cover the cost of the bad events among the individual contract (or "policy") holders. As the likelihood or cost of a bad event (such as, for example, a death of a wage earner, a home fire, car accident or unexpected illness) increases, the cost of the insurance premium to fix, or "make whole" the bad event also increases. 

Basically, the Medicare program accepts risk in exchange for FICA taxes.  Medicare's problem is that both the likelihood and the cost of illness among its beneficiaries are turning out to be unaffordable.  Unable to increase the premium or decrease the cost of illness, the logical response is to transfer that risk (along with its inadequate levels of money) someplace else.

There are three non-exclusive risk-transfer options that the DMCB thinks are being proposed and implemented at the Federal level:

1) Transferring the risk (and the money) to commercial insurers.  That program is called "Medicare Advantage," which has been criticized for transferring proportionately too much money for the level of risk.  Another approach is Medicare's "secondary payer status," forcing any other active health insurers to pay first when there is any overlap in coverage.

2) Transferring the risk (and the money) back to the Medicare beneficiaries.  Variations of this is a recurring theme among the U.S. House of Representatives Republicans, who are proposing variations of a "defined benefit"  approach, such as capped premium support payments and vouchers.  If the cost of illness exceeds a threshold, it's ultimately up to the person, not the government to eat the economic difference.

3) Transferring the risk (and the money) to the heath care providers. This underlies the logic of the upside gainshare, ACO "two sided" risk, bundling and no-pay for "never events" as well as "hospital acquired conditions."  The logic here is that providers can mitigate risk through greater efficiency or quality.  By tying in the money that goes with it, Medicare is gambling that the physicians can ultimately "bend the curve."  The amazing thing here is that, following the debacle of HMO capitation in the 1990s, the providers are willing to go along with it.

All three options have the political advantage of obscuring cost reductions for Medicare with the patina of efficiency consumerism, quality and value.  The DMCB thinks of it as savings in drag.

What happens if Medicare's risk transfer legerdemain isn't successful?  Our politicians will have to return to the politically unpalatable options of increasing the premium or reducing costs through unilateral fee schedule reductions.  Alternatively, they can outsource that repugnant task that to a newly created animal called the Independent Payment Advisory Board.

That road could (see slide 36) ultimately lead to a single payer.  More on that in an upcoming post.

Putting the CAT Scanner Cancer Risk Into Perspective: Risk Compared to What?

After listening to and reading the news media's alarmist interpretations of an Archives of Internal Medicine report on the risk of 'yikes cancer!' from computerized (axial) tomography (CT or CAT) scans, the ever curious Disease Management Care Blog decided to take a look at the basis for all the fuss. Amy Berrington de Gonzales and colleagues used radiation and cancer data to project the future national incidence of additional cancers that can be expected from today's CAT scanners. The bottom line can be found in this table from the paper. Basically, the highest calculated risk of 'yikes cancer!' runs from a high of 7 per thousand CAT scans (or 0.7%) to 1 per ten thousand CAT scans (or 0.01%). There is a companion article that indicates that the dose of radiation may be even higher, suggesting there will be even more cancer down the road.

That all may sound awful, but the DMCB likes to keep the following factors in mind:

1) The lifetime risk of cancer from exposure to the radiation from a CAT scan is highest among children, but that pales compared to the more immediate risks that come from sedating children to keep them from moving. In the real world of clinical practice, this risk threshold is already considered so high that the added 0.7% risk from the radiation is not much of an additional consideration. Informed guardians already have a powerful incentive to keep CAT scan use to an absolute minimum - and they generally do.

2) As for the grown-ups, stumbling over the assessment of the real magnitude of risk may be a function of innumeracy, but the DMCB believes it can be managed by asking 'compared to what?' So while a CAT scan can present a lifetime risk of hundreths to tenths of one percent:

- having high blood pressure gives 40 year old men an additional 4-5% (one in twenty) chance developing heart failure over a lifetime. The DMCB says know your blood pressure, get it treated if it's high and then smile when they take your picture with the CAT scan.

- the risk of a brain hemorrhage from taking daily aspirin is 2 per 10,000 per year which is on the same order of magnitude every year as the lifetime cancer risk from a CAT scan. The DMCB doesn't recall any alarmist media reports on this risk. Hmmm. Once again, take two aspirins and smile for the scanner in the morning.

- the risk of a dying in a car accident during an average life time, according to this report, is 1 in 83, which works out to be more than 1% for every man, woman and child in the U.S. today. The DMCB recommends wearing a seat belt when you drive to your CAT scan appointment.

3) One way of gauging the risk of any health threat that is reported in the media is to examine the number of reports: the greater their number, stridency and alarm, the lower the real risk. The hubub over CAT scans suggests this inverse relationship is holding up.

4) Yet, Radiology Benefit Managers (or 'RBMs') have long regarded the cancer risk associated with CAT scans (example here) as one factor in controlling their utilization. If unecessary CAT scans are 'denied' for coverage by managed care oganizations, is that now no longer evidence of 'evil' managed care putting profits before patients? Were they right all along?

The DMCB has always been less worried about the radiation dose from CAT scanners than their far more dangerous tendency to find abnormalities in the absence of disease. Asymptomatic spots, densities, signals and other lesions of dubious significance have always bedeviled physicians and their patients, leading to ever more sophisticated imaging studies. If patient fear over CAT scan radiation exposure leads to fewer scans with fewer of their inevitable false positives, that may ultimately be a good thing.

Insurance Risk & Performance Risk: An Alternative Payment Mechanism, Compliments of the Robert Wood Johnson Foundation

While readers and the Obama Adminstration search for better ways to pay for health care, the Disease Management Care Blog remains skeptical of basic fee-for-service, capitation (or 'global fees') and pay for performance. Like the Three Little Bears, the first is too hot (it promotes overuse), the second is too cold (it’s a disincentive for care) and the third is just… unproven.

That’s why it continues to like the ‘episode of care’ (EOC) reimbursement approach of having a single fee for a medical event that covers all subsequent services over a set period of time. An example would be paying for the hip surgery, the hospitalization(s), the physical therapy and all the follow-up physician visits with a single check. Think of EOC as capitation (which is supposed to cover all unrelated medical services over a period of time, typically a month) for a single condition, not a single patient.

There’s a very understandable discussion of the topic in a Robert Wood Johnson Foundation supported report from the Network for Regional Healthcare Improvement. The DMCB likes their notion of distinguishing between ‘insurance risk” (based on the numbers and types of diseases that occur in a population) and ‘performance risk’ (based on what is done to mitigate those diseases, which is a function of the numbers and types of treatments that are applied). The folks at the Network argue that while there is overlap of performance risk with insurance risk, the performance risk still can be measured, monetized and transferred to the providers.

Think of it in terms of automobile insurance. Insurers can sell policies that are designed to make you 'whole' if you have an accident (that’s the risk). They assess your age, gender, past driving record and zip code, which helps define much of the risk being priced. They collect your check (the premium) along with checks from thousands of others. However, they also have to pay attention to how much it would cost to fix the various cars in their book of business, such as parts and labor. In the Network world, the car insurers would collect your premium but then could cut a deal with the car repairmen and their body shops by negotiating a standard fee (which is functionally now an insurance premium that they ironically have to pay) to manage all the repairs that occur over the course of a year. Since the repairs occupy the lion’s share of the costs most of the money goes to the repairmen in the ‘network.’

Ahhh, but you correctly point out that diabetes and high blood pressure don’t work that way, because there is no end - like there is with hip surgery or a car that comes out of the repair shop. No problem, says the Network folks: there are cars out there that don’t necessarily need new fenders but burn oil, have threadbare tires, worn seats and cracked windshields. Since the drivers can’t give them up (they don’t like the alternative), their car needs lots of extra maintenance. For those cars, the insurers can still cut a global performance risk transfer deal with monthly payments that don’t end. That’s called ‘condition-specific capitation.’

The DMCB likes the concept because its ultimately rewards the services that are being provided for the condition at an EOC level. The payment is better targeted. That being said, it has several caveats to keep in mind while we think about paying providers a set monthly fee for a specific condition:

1) if the payment is to be done at a provider level, it requires a significant amount of physician/hospital/provider coordination, with good information transfer and ironclad hand offs. That is more likely to be present in smaller integrated healthcare systems but is tough to achieve in usual community settings with independent providers. Electronic health records are necessary but are not sufficient to pull this off.

2) while many physicians ‘get it’ and would do all the things necessary to mange their risk with higher levels of efficiency, using technology appropriately and keeping their costs to a minimum, many fine physicians would struggle in trying to adapt to this kind of arrangement. Between ‘here’ and ‘there,’ much work would need to be done. This is not a turnkey payment solution. A new fee schedule is necessary but not sufficient.

3) which is why performance risk transfer may be better thought of in terms of disease management. The DMCB thinks DM organizations understand risk (that's their pedigree) and can partner with providers in the absence of formally integrated networks to maximize coordination. Disease management, however, is also necessary but not sufficient to pull this off.

4) this would be a bear to administer, which makes it unsuited for large “mainframe” insurers like Medicare. Accordingly, the DMCB doesn’t expect this to see the light of day in national healthcare one-size-fits-all reform efforts. However, among the smaller innovative, nimble employer-based self-insured entities out there, the DMCB hopes to see this given a shot. Who knows, if a competing government sponsored insurance entity is created, the private insurers may be able to compete by offering novel payment systems like this.

5) finally, while the payment process described above transfers risk, it’s not ‘insurance’ from the point of the State or Federal regulators. Ultimately, only one party can be responsible for managing the total risk of health insurance for an individual policy holder who pays a single premium. In other words, if a hospital fails to meet its contractual obligation to provide timely and medically necessary care which, in turn, results in additional cost to the policy holder, the party ultimately responsible for fixing it is still the health insurer, not the provider.

Risk Beats Retail for Disease Management, Medical Home, P4P and The Electronic Medical Record

The Disease Management Care Blog has been thinking about the two healthcare R’s, risk and retail, from the consumer’s and insurer’s point of view.

Quantifying and pricing ‘risk’ is what insurance is all about. This intellectual achievement is described in Peter Bernstein’s marvelously written Against the Gods, the Remarkable Story of Risk. He describes how the business of risk really took off in England in the 1600s when shippers were willing to pay for financial protection against the unlikely but real possibility that a storm or other mishap would result in the expensive loss of a ship and all its cargo. The individuals willing to write their signature under the contract terms (hence ‘under’ ‘writers’) were shrewd businessmen who made a tidy profit by ‘pooling’ the payments (otherwise known as premiums) for multiple risks across many shippers for a defined period of time. Pooling (known as the law of large numbers) made the average risk narrowly quantifiable (‘X’ ships were known to go down every Y months) and bearable (a larger number of shippers paying the underwriter resulted in a ‘pool’ of money that was available pay for the one or two ships that would go down and later be used as props for the movie the Little Mermaid).

In the 400 years since then, ship insurance has been expanded to cover practically anything, including the risk of becoming sick. This is a great concept for the average consumer because the premium we pay today is far more tolerable than the individual risk of bankruptcy from a tumor, ulcer or a heart attack. As an aside, it's a powerful social good that warrants the oversight of a Department of Insurance in every state in the U.S.

On the other hand, consumers may find notions of retail to be much simpler. In this system, the insurance middleman is cut out. Goods and services can be packaged according to the market-based laws of supply and demand. Shippers and patients can shop and barter for the best service at the best price. Yet, when illness strikes, the market fails: quality and price are not transparent and there often is no time to make a good decision.

Of course, it’s not all black and white. Risk and retail are often mixed. Insurers feel they are being victimized by retail-style demand, while valuable retail health care is getting tired of being viewed as a cost by the insurer. While the insurer wants the best value for their dollar once they have to pay on a loss, it is still a loss.

Why is this important? In the opinion of the DMCB, the policy debate over many of the thorny issues surrounding the delivery of primary care services for the chronically ill have muddled the notions of risk and retail. On one side, altering the delivery of primary care with disease management support, chronic care model re-engineering, pay for performance or information technology is supposed to “pay for itself” thanks to the downstream mitigation of risk. On the retail side, these same services are supposed to increase efficiency (lower the price) or quality (at the same or higher price) or both, leading to a great deal for the consumer (and, by the way, a retail-based resuscitation of primary care). But it doesn’t stop there: most supporters of each of these primary care initiatives argue their merits lay in both risk and retail. As will be seen below, the difference is in what gets emphasized.

Unfortunately, the DMCB agrees the peer-review evidence of risk mitigation is inconclusive, making many underwriters leery about reducing the premium they charge consumers for the risk just because there’s disease management, a medical home, P4P or an electronic medical record (EMR). That leeriness is particularly understandable, because when you think about it, underwriters aren’t being asked to reduce the premium, they’re being asked to keep the premium at the same level and give a percent of it to support disease management, a medical home, P4P or an EMR.

The DMCB points out that while the ‘return on investment’ evidence may be less than perfect, the case for retail has gone mostly ignored and is probably weaker. ‘Concierge medicine’ and ‘minute clinics’ outside of chronic illness may provide some lessons, but to my knowledge no one has tested whether consumers with diabetes or heart failure would be willing to personally pay for remote coaching, the medical home, a lower A1c or an EMR-enabled practice. In the experience of the DMCB, patients want it if someone else pays for it.

So why so much blog? Because the DMCB believes the risk-retail distinction is less important in chronic illness care because the United States has already decided that risk beats retail. Like it or not, the vehicle for managing the expense of chronic illness care is not a financing issue, it’s predominantly an insurance issue:

• Pooling risk remains a powerful if imperfect methodology for making all illness care affordable, including chronic illness.

• It’s unlikely that much of a ‘retail space’ will be carved out just for chronic illness.

• The three remaining candidates for U.S. president and their allies in Congress are emphasizing insurance-based mechanisms in their campaign promises.

In the opinion of the DMCB, this landscape is why disease management has achieved much of its success. While there are retail virtues, its supporters have emphasized its ability to mitigate risk in the commercial insurance sector by: 1) presenting credible if imperfect evidence of risk reduction not only in the peer review literature but in its proprietary business dealings, and 2) has been willing to accept financial performance guarantees (a.k.a risk transfer), in essence rebating a portion of its fees if too many 'ships' end up in the emergency room. In contrast, supporters of the medical home, P4P or EMR have even less to show in terms of proof of savings and are unlikely to even give the money back if things don’t work out.

This has important lessons for supporters of the medical home, physician P4P and office-based EMR. Among the reasons why uptake has not been as widespread as disease management is because in an insurance dominated world where risk beats retail, their emphasis on the ‘retail’ dimensions represents cost to the insurer and hasn’t gained much retail support from consumers.

The DMCB asks if they will be come to terms with the world of insurance and the risk-based 'coin of the realm?' Be able to demonstrate a beneficial impact on insuring the risk of persons with chronic illness? Provide performance guarantees? They will, once the struggle to define their role begins with understanding the role of risk in caring for persons with chronic illness.