Monday, November 23, 2015

‘Growth Clouds’ On The Horizon For Health Spending?

Blog_Oakman

I was honored to work with the National Health Expenditures (NHE) Team in the CMS Office of the Actuary throughout 1995-2012, and I am honored again to have the opportunity to provide some thoughts on long-range spending trends in the U.S. Many health policy experts have referred to these accounts as the "gold standard" for comprehensive and authoritative information about the cost of health care in the United States, and the NHE Team's articles on both the historical and projected NHE accounts routinely appear at the top of Health Affairs' "most frequently read" list.

The current focus on the long history of the NHE accounts creates a golden opportunity to reflect on health care costs, how we have come to this point, and where we are likely headed. (I'm tempted to paraphrase this subject as "Where are we going, and why are we in this handbasket?")

Today we're in the midst of a monumental effort to develop and assess alternative means of providing and paying for health care. Accountable care organizations, value-based care (and insurance design), medical homes, competition, bundled payments, integration across service providers, prescription drug and treatment adherence, preventive and wellness services, electronic health records, narrow provider networks, consumer-driven health plans, telemedicine, and a host of other initiatives are all being tested; this experimentation has encouraged much optimism that the quality of care can be significantly improved while also restraining cost growth.

The ultimate success (or failure) of these innovations in restraining cost growth will be reflected in the impartial data on national health expenditures. I'm optimistic that at least some of these efforts will prove successful — but I'm also mindful that few of them would affect the primary drivers of health care cost growth. A series of one-time reductions in the level of health care costs would be welcome, of course, but they would not represent a long-term reduction in the health care cost growth rate. Moreover, a look at one of the key drivers of growth—namely new medical technology—suggests that there are major "growth clouds" on the horizon.

The NHE Team

The roster of economists, statisticians, actuaries, and other staff who created and have prepared the NHE accounts reads like a Who's Who among health care technical experts in the U.S. Dorothy Rice and Barbara Cooper initiated the formal accounts, with early help from the legendary Helen Lazenby. Katie Levit presided over the historical accounts for many years, while Mark Freeland, Dan Waldo, and Sally Burner periodically produced 10-year projections.

These days, Steve Heffler, Aaron Catlin, and John Poisal ably lead the National Health Statistics Group. Cathy Cowan, Anne Martin, Micah Hartman, Lekha Whittle, Ben Washington, Mary Carol Barron, Dave Lassman, Bridget Dickensheets, Joe Benson, Heidi Oumarou, and Nathan Espinosa produce the historical accounts, and Sean Keehan, Sheila Smith, Andrea Sisko, Gigi Cuckler, Andrew Madison, C.J. Wolfe, Devin Stone, and Joe Lizonitz estimate health spending for the next 10 years.

Collectively, these individuals have produced a comprehensive body of knowledge about health expenditures in the U.S. that is invaluable, and all the more so because it is objective and unbiased — no easy feat in today's excessively partisan atmosphere. We owe these experts, past and present, our sincerest gratitude and support.

Reducing The Level Of Costs Versus The Cost Growth Rate

Almost all of the health care initiatives listed above have the potential to improve the quality of care. Many of them could also reduce the level of health care expenditures. For example, use of a nurse hotline can help prevent inappropriate emergency room visits and avoid unnecessary hospital readmissions; together, these steps can reduce health care expenditures by roughly 5 percent. Their impact on long-term health care cost growth is negligible, however, unless you expect ER misuse and unnecessary readmissions to get steadily worse over time.

Reducing the long-term growth rate for health care costs is much more difficult. The Affordable Care Act's reduction in annual Medicare payment rate updates for most categories of providers, by the increase in economy-wide multifactor productivity, definitely lowers the program's expenditure growth rate. Whether this provision can continue to do so for years to come, without jeopardizing access to care for beneficiaries, is doubtful. (An excellent discussion of this question is available in Chapter 4 of the report of the 2010-2011 Technical Review Panel on the Medicare Trustees Report.)

There are relatively few guaranteed ways to reduce the growth rate, and some of them are highly controversial — think "defined contribution," for example. Other approaches, such as value-based payment innovations, may have the potential to do so but are still under development and testing.

The Drivers Of Health Care Cost Growth

To assess how to reduce long-term health care cost growth, it's helpful to start with an understanding of the factors that cause such costs to grow in the first place. In their definitive article, Sheila Smith, Joe Newhouse, and Mark Freeland disaggregated past health care cost growth into five primary factors, with the following results:

Foster-table1

Although changing demographics have moderately affected past health care cost growth, the retirement and aging of the baby boom generation will have a much larger impact for the next several decades. Some of us have been warning about this demographic shift for over 40 years, but it is now upon us, and relatively little can be done at this late stage to address the higher cost associated with an older population. Raising the age of eligibility for Medicare benefits may or may not be a sound idea for public policy, but doing so would only shift costs from Medicare to individuals, employers, and other public programs (for example, the Federal subsidies for individual health insurance policies in the ACA marketplaces). The impact on aggregate health spending would be minimal. End-of-life counseling continues to be a good idea from the standpoint of both quality of life and cost, but it remains controversial.

The proportion of people with health insurance will increase significantly as a result of the Affordable Care Act, with availability of insurance regardless of health status, subsidized coverage in the new marketplaces, and the expansion of Medicaid. These increases in coverage will add to health care cost growth, but not on a long-term basis. Once the transition is completed over the next few years, the insured proportion is unlikely to change substantially. There may be some tendency, however, for individuals and employers to scale back on their coverage, both in response to costs generally and to the ACA's "Cadillac tax" on generous employer coverage in particular (if this provision survives). Overall, I expect that changes in the prevalence and generosity of health insurance will not have a sizable impact on health care cost growth in the long term.

The NHE accounts have documented a long history of medical price growth that exceeds economy-wide price growth, with a persistent differential of at least a 0.5 percentage point. Much of this difference can be attributed to the higher labor share associated with health care services: wages tend to increase at a faster rate than prices for manufactured goods and other inputs to production. As a result, businesses with high labor shares have greater input price growth than those with small labor shares. (Growth in prices for health care services can be lower than associated input price increases if providers can increase their productivity. Strong evidence suggests, however, that health care productivity gains are far lower than for the economy at large.)

Over time, physician assistants and nurses can take over some of the duties previously performed by physicians. This shift would not reduce price growth for a fixed market basket of inputs, but it would change the mix of inputs in a way that would reduce the cost of providing care. It is not clear, however, that such a change could be made year after year, indefinitely, without ultimately affecting the quality of care.

The two largest contributors to historical health care cost growth have been rising incomes and new medical technology. The Smith-Newhouse-Freeland article discusses these factors in detail, so I will merely summarize that (i) the richer we are as a nation, the more health care we desire and can afford, and (ii) to date, most medical technology has been "cost increasing." The first of these factors is not necessarily a problem: as our incomes increase, we may well prefer to devote a larger share to more and better health care, in lieu of, say, greater consumption of iPhones, Camaros, and dinners out. A problem would occur, however, if our expenditures for health care began reducing other desired spending. (This effect is already quite apparent in the Federal Budget and most State Budgets, although it reflects all of the health care growth factors—demographics, insurance, excess medical price inflation, and technology—in addition to income gains. The serious imbalance between Federal outlays and revenues is also a factor.)

The impact of rising incomes on health care expenditures is largely voluntary and, to a significant degree, self-regulating; absent sufficient income growth, we're not going to choose to spend an ever-increasing proportion on health care. In other words, it's unlikely that the future U.S. population will be walking around naked, homeless, and hungry but with outstanding health care.

That leaves us with the impact of medical technology, which deserves a closer look.

Medical Technology

Over the years, we have all benefited tremendously from improvements and innovations in medical technology—new treatments, devices, and drugs—and we will continue to do so. Historically, virtually all new technology has been approved and adopted in the U.S., even though not all of these developments have been beneficial or cost effective. Examples are not hard to find:

  • The drug Nexium is an über-example, having replaced its nearly identical predecessor Prilosec when the latter's patent expired.
  • Proton beam therapy is more effective than traditional radiation treatments for certain uncommon forms of cancer, such as tumors of the eye — but it has not been shown to be more effective for other forms. Nonetheless, there is a nationwide race to build new proton beam facilities, which offer an attractive return on investment from treating prostate cancer. (To date, studies are inconclusive as to whether proton beam therapy reduces adverse side effects in such cases.) There will soon be three such facilities in the Washington, DC – Baltimore area, when a single facility would probably accommodate the medical needs. Moreover, with a facility cost of about $200 million, the price of treatment is three to six times that of traditional forms.
  • Hospitals and other providers often compete with each other based on their medical technology. Remember when each hospital needed its own MRI machine? And, later, a more patient-friendly "open air" MRI? And Da Vinci robotic surgical systems? The acquisition of expensive new technology often leads to greater-than-justified use, to recoup costs.

Some new technology is highly effective but extraordinarily expensive:

  • The prescription drug Sovaldi, for curing hepatitis C, is a good example: available evidence suggests that its cost of production for an 84-day course of treatment is between $150 and $250, compared to its retail price of $84,000. A vice president of the company producing Sovaldi, Gregg Alton, stated: "I think we can get it into an affordable range for [foreign markets]. It'll be from the high hundreds to low thousands for these types of markets."
  • Several important new drugs have been introduced in the last few years for the treatment of multiple sclerosis. Prices range from $45,000 to over $100,000 per year.
  • Gene therapy, which uses a virus to replace an abnormal gene with a healthy one, can also be a literal lifesaver. Although this technology is still in the early stages of use, it is likely to prove exceptionally costly. No such treatments have yet been approved in the U.S., but, in other countries, $1,000,000 per one-time treatment is about average.

A third category of medical innovation involves efforts to develop "cost-reducing" technology. Many parts of the world cannot afford U.S. technology and have begun designing and producing lower-cost versions. There are similar efforts underway within the U.S.:

  • Current-generation implantable defibrillators can be inserted subcutaneously, eliminating the need for open-chest surgery and reducing costs significantly.
  • "Personal sedation stations" can reduce the need for more expensive care by anesthesiologists.
  • A few years ago, in response to competition from other countries, a major U.S. corporation developed an MRI machine that cost about one-fourth of the usual amount and required far less training to use. It was aggressively marketed in developing nations — and also in the U.S., where large physician practices were the primary target consumer. Over-utilization of hospital-based MRIs had already been an issue, and the broader ownership of these devices has probably exacerbated the problem.

It can be challenging to distinguish between valuable, cost-effective new technology versus other innovations that add no value or are excessively priced relative to outcomes. Attempting to do so usually becomes an exercise in Quality-Adjusted Life Years (QALYs) and other factors that make us uneasy. The effort is also fraught with political controversy, with accusations of "death panels" and the like hurled about. Further, more prudent adoption of new medical technology cannot be implemented piecemeal. Imagine, for example, if private health insurance covers a new treatment but Medicare does not. The resulting lawsuits would quickly overturn the Medicare coverage decision.

Given these circumstances, I suspect that only a national organization, not unlike England's National Institute for Health and Care Excellence (NICE), could lead to a more appropriate adoption of new technology in the U.S. Such an outcome would have both primary and secondary benefits. First, forgoing expensive new drugs and treatments that do not result in better outcomes would reduce costs and help ensure that our national health expenditures are used optimally. A secondary benefit would occur if the more prudent adoption of technology fed back into the research and development process and contributed to a greater focus on cost-reducing technology. The Patient-Centered Outcomes Research Institute is a step in this direction — but only a small one.

Looking Forward

Increasing incomes and new medical technology have historically been the largest factors contributing to rapid health care cost growth. This trend is likely to continue and to cause higher growth rates than we have seen in the 2008-2013 period. More prudent adoption of new technology in the future, focusing on outcomes, comparative value, and cost-effectiveness, would help moderate growth rates, but it is unlikely to cure the problem of health care cost growth that exceeds economic growth. We still want—and will get—effective new drugs, devices, and treatments for our loved ones and ourselves. An improving standard of living will help us afford more and better health care, but we should anticipate that government, business, and family budgets will continue to face increasing pressures from health care cost growth.

As these trends play out over time, the ongoing availability of the historical and projected national health expenditure accounts will be invaluable as an objective guide to researchers and policy makers alike. Although projections are notoriously uncertain, I predict that, by the 100th anniversary of the NHE accounts, we will have resolved the issue of how to balance our collective demands for health care with our collective ability to pay for it!



from Health Affairs Blog http://ift.tt/1PKtq0H

No comments:

Post a Comment