I have been thinking for a while now that one of the problems with the dialog about improving the US healthcare system is that we have done a poor job of articulating what we want to accomplish.
Do we want healthcare with the best possible outcomes? Or do we want healthcare that represents the optimal allocation of some amount of money that as a society we consider affordable (I think of this as the most cost-effective healthcare)?
This post is about how to frame these questions so we can have a rational discussion and debate. Here is my illustration of costs and quality in healthcare.
- Having a mammogram every decade represents a modest level of diagnostic intensity. Having one every year is a much higher level of intensity, while never having one at all is a lower level of intensity.
- Similarly, Cholesterol measurements annually is a higher level of intensity than having them measured once a decade.
- And if you have cardiovascular problems, there is an escalating scale of therapeutic intensity that starts at medication, and rises through angioplasty and/or stenting to perhaps coronary bypass, or even an artificial heart.
As you would expect, the top red curve shows that costs rise more or less in lock-step with the intensity of intervention. The blue curve is more interesting. It shows the outcomes of the interventions. These might be measured in some unit such as increase in Quality Adjusted Life years (QUALYS) for example. Initially the Outcomes increase (ie improve) as clinicians do more stuff to the patient. But notice that eventually this increase flattens out and then starts to decline.
What explains this effect? Well, initial medical interventions are typically extremely effective at improving outcomes. If Type I diabetics take regular insulin shots, it means the difference between life and death, as do many surgical interventions. If you have occasional mammograms, they are likely to catch cancers earlier than they would otherwise be caught, leading to better outcomes.
But as the intensity of therapy/diagnosis increases we see diminishing returns, followed by adverse effects of the intense interventions.
- For example, one mammogram every few years is generally considered a good thing. But if you had one every day, there would be almost no benefit from earlier disease detection, while you would be exposed to a much greater x-ray dose – thus potentially increasing the chances of cancer, and thus reducing the overall outcome (actually decreasing the QUALYs).
- There are many other areas where over treating can lead to adverse outcomes. Another example might be prophylactic stenting of relatively un-diseased coronary arteries.
Optimal outcome healthcare
All of this illustrates a pretty simple concept. Namely, that there is a level of therapeutic / diagnostic intensity that gives the optimal outcomes. Both too much and too little intervention by clinicians lead to reduced quality of outcome (as measured for example in QUALYs). And if you think of this from the perspective of an individual you are probably thinking “That is the level of medical attention I want, thank you very much”.
The trouble is that from the perspective of society as a whole, that may well not be the right way to think about things, and I would argue that it is not at all the way governments and other societies are thinking about the delivery of healthcare. And this is important, because I think it is at the heart of our current debate about the future of healthcare in the USA.
Cost effective health
If you look at the lower section of the graphs above, you see a curve in red called “Costs per unit of increased quality”. This would be a measure such as “Cost per increased QUALY” or the cost of gaining an extra year of life (adjusted for quality of that life). And what this graph illustrates is that for the initial low intensity interventions, each extra QUALY is pretty inexpensive, but then as you add ever greater intensity of intervention it gets more and more expensive to buy an extra year of life.
This is not a controversial point of view. Think of the heroic, end-of-life efforts required to gain a few more days. Or think of the high costs of an artificial heart, or some of the newer biotech therapies, to gain a few extra months/years of life. Now none of this means these interventions are bad. it just means they are expensive per year of life gained, and that other interventions (e.g. bed nets in malaria-prone regions) might have more bang for the buck.
Health economists routinely think in terms of a threshold for $$/QUALY. In other words, a society (or an insurer or government) decides that it doesn’t want to pay for interventions that cost more than $xx per QUALY, because they think the money would be better spent elsewhere.
This leads you to the second of the vertical green lines on the graph labelled as “Optimal cost effectiveness”. This line corresponds to the optimal amount of interventional intensity one would wish to employ in order to get the best outcomes possible – consistent with staying below some threshold of acceptable $$/QUALY.
Optimal outcome vs cost effectiveness
The first very important point to make is that the line representing optimal cost effectiveness and the line representing optimal outcome are not necessarily in the same place. This is intuitively obvious if you pause for thought, although the knee jerk reaction of many is that these two measures should be the same.
The two lines reflect very different analyses and so it would be sheer coincidence if they overlapped. And the medical literature contains many scholarly analyses of interventions that are medically effective but have very high costs per QUALY.
And there are panels (e.g. NICE in the UK) that routinely conduct analyses of new therapies or diagnostic techniques to evaluate various factors including cost effectiveness ($$/QUALY). Some of the new techniques are deemed not to be cost effective. This is not at all to say those therapies are not good. Just that they are expensive.
What should we want?
So what should we want? Optimal outcomes or optimal cost effectiveness?
It seems to me that for any individual, one would ideally like the optimal outcome. And if someone else is paying for it, then I totally want that. If I have to pay for it myself, I will think long and hard about just where I want to draw the line between perfection of result and huge overwhelming cost.
But if I am thinking on behalf of society as a whole, or thinking about how to best spend my tax dollars, I rapidly decide that we cannot afford a system in which all consumers of healthcare get optimal outcome care. (This could be the subject of a whole other post, but suffice it to say we don’t have that system today and nor do other countries and I think this is a defensible assumption). So from a societal point of view I rapidly gravitate to the idea that we should have some type of system that focuses on optimal cost effectiveness.
In other words, we want to decide how much we can afford as a nation, and then allocate that in a cost effective way to the interventions that will best increase the overall health of the nation. This is the goal of an organization like NICE as I understand it.
Is there a role for individual choice?
If we have a healthcare system in which the government pays for healthcare, then it seems as if cost effectiveness is the logical metric. But what about an individual who might wish to spend her own money to get some level of quality (optimal outcome) that exceeds what society can afford for all its citizens? Should this be possible?
You can certainly see examples of other countries that have a two-layered healthcare system. They have a government run system that focuses on optimally spending whatever amount the country is comfortable spending on healthcare, and then they have an additional layer in which individuals can, if they wish, obtain higher intensity interventions at their own cost. This is one approach to creating a system in which individuals can decide whether they care about cost effectiveness or optimal quality.
But there are other countries in which it is deemed unacceptable that rich people should be able to get better care, and thus one is “not allowed” to get care that goes beyond some sanctioned, cost-effective “norm”.
What should the US have? Well, that is for the populace as a whole to decide. But I at least find it helpful to think in terms of these two very different metrics as I think about the various options that lie ahead.