Health Affairs, 25, no. 2 (2006): w48-w50 (Published online 7 February 2006) doi: 10.1377/hlthaff.25.w48 © 2006 by Project HOPE	New Online   Senate Health Reform Bill   Rewarding Providers   Public Option Policy Brief   Health Reform & Abortion   Delivery System Reform

This Article Abstract Reprint (PDF) Submit a response to this article Alert me when this article is cited Alert me when Comments are posted Alert me if a correction is posted Services E-mail this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to Citation Manager Reprints & Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Cutler, D. M. Search for Related Content PubMed PubMed Citation Articles by Cutler, D. M.

Web Exclusives

PERSPECTIVE

Making Sense Of Medical Technology

Abstract

 
In seeking to explain contradictory reports that increased medical spending is both valuable and wasteful, Jonathan Skinner and colleagues rightly emphasize differences in regional practice patterns in response to constraints. However, it is more accurate to view the situation not as a choice between valuable but inexpensive care and more costly care with worse outcomes—as Skinner and colleagues theorize—but as a choice between responses to regional inefficiencies. Given that we do not know what will best improve the efficiency of medical care in a given area, experiments with financial incentives to both providers and individual patients should receive high priority.

As Skinner and colleagues note, the data and methods differ in the two types of studies. Studies showing that medical spending is valuable generally consider time-series evidence on medical care costs and outcomes. Over time, countries spend more on medical care, and health improves. Since enough of the improvement in health appears to result from medical advances, spending more is worth it. The studies suggesting low value, in contrast, are usually cross-sectional: Regions of the country that spend more on medical care do not have superior health outcomes. If anything, outcomes are worse. The same finding is true in analysis of international spending differences. Thus, more care cannot be worth much.

Comparing two methods. Why do the time series and cross-sections differ so much? The difference could result from several factors. One possibility, which Skinner and colleagues note and reject, is flat-of-the-curve medicine: The benefits to more-intensive care are exhausted at the level of spending in low-spending regions, so additional resources in high-spending areas bring little in the way of health improvement. This explanation is superficially attractive; it implies that all is efficient but that some regions choose to do more than others. But it is surely wrong. The relationship between inputs and outcomes would have to be very flat indeed for spending in one area at half the rate in another to span entirely marginal care.

A second theory is that spending and technology are not as closely related at the area level as they are in the time series. Over time, the vast bulk of the increase in medical spending is a result of increased diffusion of such services. At the area level, however, Skinner and colleagues note that greater medical spending is not correlated with higher use of intensive surgical services. Rather, it is associated with increased physician visits and tests. So, one might expect less relation between spending and outcomes at the area level. Still, the lack of any relation between physician visits or tests and outcomes requires some explanation.

A third theory is that the productivity of the same service differs across areas.³ Usually, this is told as a story in which areas that do more high-tech interventions are better at it and thus have better outcomes. The opposite could also be true. Areas where many angio-plasties are performed might also be areas where many physicians perform them; thus, a good number of physicians have not moved down the learning curve. Skinner and colleagues could use their data to see the relation between doing more and the concentration of procedures as a way of testing this.

What this latter theory brings up is a general issue about regional practice patterns. It is where Skinner and colleagues place their emphasis, and it is surely right. Some regions have chosen to provide more valuable but inexpensive care, while others provide more-costly care, with lower outcomes. The authors suggest that each of these regions is constrained in its own way: Some have chosen one production function, while others have chosen a different one. We might prefer one to the other, but there is no way to make the transition between the two.

This theory is broadly correct but wrong in parts. Skinner and colleagues let inefficient areas off the hook too easily. The correct way to view the situation is with a single production function. Some regions are closer to the production frontier, and others are further away. The ones closer to the frontier get more output per dollar, while the others get less. Everyone is imperfect, although some are closer to perfection than others. Areas that spend more might even have worse care.

Improving efficiency. If one accepts that all areas are inefficient to some extent, there are several questions that one is led to ask. The first is how to make inefficient regions more efficient. The issue of efficiency improvement is fundamental to any health care reform. Without explicit rationing, it is the only hope we have of saving money in medicine. Skinner and colleagues mention a few ideas, although they do not draw them out: restructure hospital resources; improve physician treatment patterns; help with coordination across specialists; preach about the benefits of low-cost, highly effective care. As they note, there are no guarantees: Just forcing regions that spend more to cut back does not mean that the care withdrawn would be the least valuable care.

The issue of efficiency is important because the dollar values are immense. In Skinner and colleagues’ analysis, areas in the highest quartile of physicians per patient had cost increases that exceeded those in areas in the lowest quartile by more than $3,000 (see their Exhibit 5). If lack of coordination across physicians is the basic reason for area differences, one could hire a case manager for under $3,000 per case, save money, and improve outcomes.

In the post–managed care era, there are two fundamental strategies that might improve efficiency. The first is through patients: Provide financial incentives to individuals to choose lower-cost, higher-quality care. The financial incentives come in the form of higher cost sharing: deductibles well above traditional levels, possibly with a subsidy for preventive care. Proponents claim that making care more of a market will lead to efficiency gains like in other markets. The second strategy is via providers: Pay providers more for higher-quality care and less for lower-valued care. When providers have incentives to do the right thing, the argument goes, care will improve. Information is a key part of both strategies; neither people nor physicians can make rational choices without knowledge about what is appropriate and what is not.

Skinner and colleagues are generally supportive of the second strategy. I am as well.⁴ Thus, my sympathies lie heavily along the lines the paper suggests. But the truth is that we do not really know what will most improve the efficiency of medical care in inefficient areas. Given the financial stakes, experimenting with both ideas is a high priority.

Suppose that all areas could be brought to the efficient frontier. What would the medical world look like then? In the world of Skinner and colleagues, it is happy, but not ecstatically so. Survival would certainly improve greatly. Compared to an average one-year survival improvement of 9.8 per 100 acute myocardial infarction (AMI) patients, the improvement in high-quality areas was 12.5 per 100 AMI patients, a 25 percent improvement. But cost growth would moderate only a bit. Relative to an average cost increase of $12,400 per MI, the cost increase in efficient regions is about $10,000 per MI. The bulk of the cost problem would thus still be there.

In Skinner and colleagues’ analysis, this is about as good as can be done, since they assume that areas with low spending growth define the frontier of best possible outcomes. This question is really deserving of a closer look, though. Is there overused care in low-spending areas? Are preventive measures used enough? There is a real need for someone to answer the question: If we took everything we know and applied it to health care, how much beyond the best current practice could we be?

Changes in AMI care and cost. Almost as an aside, Skinner and colleagues note that while costs for AMI care increased throughout the 1986–2002 time period, outcome growth slowed after 1996. They do not know what to make of this. Nor do I. The quality of heart attack treatment seems certain to have improved over this time period. After a bit of a learning curve, physicians learned how to implant stents with lower risk of restenosis. Low-tech care diffused as well; the rate of beta-blocker use in MI patients in commercial plans rose from 63 percent in 1996 to 94 percent in 2002, although the Medicare increase over a shorter time period (1998–99 compared with 2000–01) was smaller.⁵ Some of the change might be a population that is sicker in some ways.⁶ Another possible explanation is that some care was harmful, which we will only learn with time. It may also be that advances in therapy help several years down the road, not in the first year. In any case, some more exploration is warranted.

Overall, Skinner and colleagues take a big step in understanding the multiple impacts of medical advance. Like any provocative paper, they also leave us with a host of questions.

Editor's Notes

 
David Cutler (dcutler{at}harvard.edu) is the Otto Eckstein Professor of Applied Economics and dean for the social sciences at Harvard University in Cambridge, Massachusetts.

The author is grateful for research support from the National Institute on Aging.

NOTES

Top NOTES

 

1. Regarding the first argument, see D.M. Cutler and M. McClellan, "Is Technological Change in Medicine Worth It?" Health Affairs 20, no. 5 (2001): 11–29[Abstract/Free Full Text]; M.E. Chernew, R.A. Hirth, and D.M. Cutler, "Increased Spending on Health Care: How Much Can the United States Afford?" Health Affairs 22, no. 4 (2003): 15–25[Abstract/Free Full Text]; and S. Glied and S.E. Little, "The Uninsured and the Benefits of Medical Progress," Health Affairs 22, no. 4 (2003): 210–219.[Abstract/Free Full Text]Regarding the second, see K. Baicker and A. Chandra, "Medicare Spending, the Physician Workforce, and Beneficiaries’ Quality of Care," Health Affairs 23 (2004): w184–w197 (published online 7 April 2004; 10.1377/hlthaff.w4.184).
2. J.S. Skinner, D.O. Staiger, and E.S. Fisher, "Is Technological Change in Medicine Always Worth It? The Case of Acute Myocardial Infarction," Health Affairs 25 (2006): w34–w47 (published online 7 February 2006; 10.1377/hlthaff.25.w34).[Abstract/Free Full Text]
3. A. Chandra and D. Staiger, "Testing a Roy Model with Spillovers: Treatment of Heart Attacks," NBER Working Paper no. 10811 (Cambridge, Mass.: National Bureau of Economic Research, October 2004).
4. D.M. Cutler, Your Money or Your Life: Strong Medicine for America’s Health Care System (New York: Oxford University Press, 2004).
5. National Committee for Quality Assurance, The State of Managed Care Quality (Washington: NCQA, 2005).Regarding the Medicare increase, see S.F. Jencks, E.D. Huff, and T. Cuerdon, "Change in the Quality of Care Delivered to Medicare Beneficiaries, 1998–1999 to 2000–2001," Journal of the American Medical Association 289, no. 3 (2003): 305–312.[Abstract/Free Full Text]
6. A.S. Ash et al., "Using Claims Data to Examine Mortality Trends following Hospitalization for Heart Attack in Medicare," Health Services Research 38, no. 5 (2003): 1253–1262.[CrossRef][Web of Science][Medline]

                      What's this?