Death from coronary heart disease (CHD) has declined greatly over the past 30 years but it remains a major cause of death and disability in the United States as well as a major source of medical spending. In an interview with PHSI, University of Michigan assistant professor of internal medicine Allison Rosen talks about why she and Harvard University colleague David Cutler are looking at the value of spending on heart disease and other chronic conditions to better understand how much this spending buys. With partial support from PHSI, Rosen and colleagues recently examined trends in the costs and health benefits of treatment for CHD. Their findings were reported in the January/February 2007 issue of Health Affairs.

PHSI: It's clear that spending for coronary heart disease has been worthwhile but you point to how difficult it is to link health spending to health improvements. Can you explain?

AR: In many of these studies we're looking at the degree to which spending has increased and the extent to which health has improved; but these two trends are often explored independently and then superimposed with the assumption that the spending led to the health outcomes (or vice versa). While this can tell us something about the value of spending overall, this masks a lot of variability in the value of the specific underlying services – it's likely that a number of services are very high value but a number are also likely to be of low or unknown value. Some of the work that David Cutler and I are pursuing relates to better identifying the specific services that provide the greatest health improvements relative to their costs.

PHSI: Why is understanding that link, that causality so important?

AR: It is tremendously important. While, in aggregate, we get excellent value for our spending, we must be able to better identify which of the underlying services are most (or least) valuable if we want to provide policymakers with clear targets for prioritizing allocation of increasingly limited resources.

PHSI: What do you mean by "value"? Why is it important to think about when considering health financing and policy?

AR: I would define value as the health benefits achieved for the money we spend. As resources become increasingly constrained and large gaps in both quality of care and access to care persist, something has to give. Our best policy option is to prioritize providing services that are worth the cost – spending more where that is justified and less where we are not achieving appropriate returns.

PHSI: You talk about how more research is needed to better "maximize" the productivity of our health care spending. What do you mean by that?

AR: We need to understand exactly where the value lies and whether our payment policies are creating the right incentives for value. For example, after a heart attack we know that beta blockers, ACE inhibitors, and statins provide substantial value. We've done a good job at improving care for those who have had heart attacks but we can do much better. Only 50 to 55 percent of heart attack patients over the long term are taking these life saving and quite valuable medications. In turn, if we want to maximize the productivity – or value – of our heart disease spending, we need to know how the value of these therapies compares to that of others.

So while we've gotten excellent value for secondary prevention of heart disease we could do better at identifying and targeting money to make sure the most valuable things are done. For example, if we know that beta blockers are really valuable in patients with heart disease, we should target our spending to make sure they are used more. There are a whole slew of tools at our disposable to improve use, and currently we are not using those tools well. Payment policies often work in opposition to the provision of many high value therapies. We reward providers for increasing beta blocker use following heart attacks yet we penalize patients by increasing their co-pay for the very beta blocker we're incentivizing physicians to prescribe.

PHSI: So, should the federal government provide coverage for all of those who need anti-hypertensive medications?

AR: That's not an easy question to answer. While my work with David Cutler focuses on figuring out the value of the health care system; much of my other research is related to redesigning benefits to improve the use of high value therapies. This entails targeting resources very specifically to their best uses. If you ask me whether Medicare should provide coverage for beta blockers for heart attack patients, my answer is, 'yes, that's a wise use of resources.' In fact, some of my past research asked this very question about ACE inhibitor coverage for diabetics, showing that Medicare would save hundreds of thousands of lives and billions of dollars. A study in this same issue of Health Affairs by Niteesh Choudry reported similar findings for high value heart attack therapies.

Here at the University of Michigan Center for Value Based Insurance Design we're examining what happens when large employers lower copayments in a very targeted way for therapies that are highly valued but underused. Right now, we have several ongoing studies with large employers, including a controlled trial in which University employees with diabetes have received targeted copayment reductions for such things as ACE inhibitors and eye exams. We hope to show that it will improve compliance to those therapies most important for preventing downstream complications.

PHSI: Can you talk about how then we would better understand what is driving health improvements so we can target resources more effectively?

AR: If we want to target resources to improve value, we must be able to measure value in a consistent manner. Looking at the value of the health system disease-by-disease is better than looking at the health system broadly. After all, this is the way we develop therapies and treatments.

PHSI: You and David Cutler are working on a project related to the National Health Expenditure Accounts. What are you learning from that work?

AR: We're trying to expand the National Health Expenditure Accounts (NHEA), which track health care spending in the United States, to add a measure of population health in addition to the spending that's already tracked. Currently the NHEA tracks the flow of funds into and out of health care, providing information on the payers and the services paid for. While national health policy is frequently predicated on these estimates, the NHA provides no information on the true output of the health care system – health. By adding a measure of population health and tracking this alongside spending, we hope to say something about health improvement relative to spending.