WAGTalk: Victor Rodwin, “What’s the Best Health System & How Would We Know?”

By Adem Lewis / in , , , , , , , , , , , , , , /

Good morning. I was cleaning out old
books over the weekend, and I came across one of my favorite authors. I couldn’t resist.
Anybody here read Dr. Seuss? And inside there was an insert —
this piece of paper which was a book that came out- one of his poems that
came out in his obituary, actually. I thought it was relevant to the work that
I’m about to present. [reading from poem] “My uncle ordered popovers from the restaurant’s bill of fare.” You know popovers, those of you who don’t
come from New York? “And, when they were “served, he regarded them with a
penetrating stare. Then he spoke great “words of wisdom, as he sat there on that
chair: ‘To eat these things,’ said my uncle, “‘you must exercise great care. You may
swallow down what’s solid, but you must “‘spit out the air!’ And as you partake of
the world’s bill of fare, that’s darned “good advice to follow. Do a lot of spitting out
the hot air. And be careful what you swallow.” The field of health policy
and management, as many other fields today: we’re drowning in data. None of us
can escape the bombardment of information all around us. We encounter
expressions that are politically incorrect to question. Evidence-based medicine. Anybody in favor
of non-evidence-based medicine? Evidence-based management.
Evidence-based policy. Accountable Care Organizations. Anybody
in favor of non Accountable Care Organizations? Healthy cities.
Anybody like unhealthy cities? Age-friendly cities.
Anybody in favor of non-age-friendly cities? So, my advice to you today is:
indeed, be careful what you swallow. Now I’m going to illustrate this proposition
by discussing the challenge of evaluating health care systems
around the world. The World Health Organization — it’s based in Geneva — introduced the idea
of measuring health system performance, and we talk a lot about that at the
Wagner School. And their report actually ranked
about 190 healthcare systems around the world. Since then, the Commonwealth Fund, right here in New York City, which has focused on 11 very wealthy healthcare systems in wealthy
nations, also ranks 11 health systems of wealthy nations and asks, rhetorically,
the title of my talk. Why is the US not the best? I’m going to speak to you about:
how would we know which system is the best? And it’s a subject that covers many aspects about what I teach, and about some of my research. One way of knowing, and it’s a very
important way of knowing these days, is simply based on beliefs and values.
Indeed, many Americans would agree with former President Bush that,
“Americans have the best health care system in the world,”
and that’s an article of faith. I submit to you that most Brits and most Canadians and most of the French would also have that same faith in their own health care system,
and I have evidence to back it up. But beyond belief, how would we know which
health care system is the best? Well, some people say it’s all a matter of spending.
If you look at the United States in 1980, as a percent of GDP, we spent more than
any other country of the OECD countries, and that gap has only increased. We now
spend more money than any country in the World, so some people would infer that we
have the best health care system; we’re resource-intensive, we have everything
here. People come from all over the world. Now, the question is: is spending more,
necessarily, good? Why do we spend more? Is it a price effect; do we spend more? Most
of the research suggests that we spend more because we have the highest prices.
Is having the highest prices good? It’s very good for people who work in health
care. Those of you who will get jobs in health care will have higher salaries. Doctors have higher salaries; administrators have
higher salaries. It’s a booming profession. Overall, the evidence does suggest
that we spend more because we have high prices. But also, beyond prices,
there’s some work at OECD, which looks — and I admit, it’s very hard I’m racing
through this — but beyond prices, we generally have more resources — not in
every dimension, but we’re right up there with numbers of physicians;
although not the most, we have a tremendous number of MRIs per capita; a
lot of technologies; a large share of the population employed in healthcare. So,
we’re a pretty fat octagon, if you look at those graphs published by the OECD.
We have a lot. But is that the way to measure a healthcare system?
Maybe there are more appropriate or other criteria one should take in mind
for assessing healthcare systems. Maybe we should focus on survival.
Anybody here think we should focus on survival — how well does the population survive?
Who’s in favor of that? Hands! Quickly. Okay? Our Dean Sherry Glied, with whom
you’ve just met, has done a very interesting study which suggests that in
1975, the United States spent the least and had some of the lowest survival
rates, and in 2005, thirty years later, spent a great deal more. Much, much higher
expenditure. You see the US there in the corner? — I can’t point to it. But also,
very low survival rates compared to other systems. So, that’s not so good. For
those of you who said survival is important as a measure of the health
system, I submit to you that it’s not. It’s not a very important measure of the
health care system, because the health care system only helps survive at the margin.
There are too many other factors that account for survival. For example, there’s
a very important survey — 2016, from the Commonwealth Fund — which,
when you ask a representative population — [reading sign] Does that say half? Okay —
about whether they experience material hardship and whether they’re stressed about being
able to pay rent/mortgage or buy nutritious meals, we come out the highest.
That is the worst. A large- One-fifth of the population suggests that there’s a
lot of material hardship. Now, that may have much more to do with survival than
the health care system. So, in fact, the problem of judging
health care systems based on survival or based on life expectancy at birth is
that it violates the great equation. Health care is not equal to health. It’s
only one dimension in producing population health. That’s a major part of
the Introduction to Health Policy and Management course. Our own Brian Elbel —
I don’t know if he’s here today — did a study with Elizabeth Bradley from Yale,
in which he compared the percentage of gross national product that we spend on
health care and then looked at what we spend on social services: social care,
old-age pensions and support services for older adults, survivors benefits,
disabilities, sickness cash benefits, family supports, employment programs. And on that
score, if you look at the graph, you see that we spend the least as a percent of
GDP on social care, even though we spend the most on GDP on health care. What other criteria should we look
at to figure out if we are the best? Well, one other way,
and it’s been looked at by members of Congress over and over again,
is to look at just survival from cancer. In this respect, we do very well. We have
some of the highest survival rates of cancer. But is that the right metric to
use to evaluate and determine what system is the best? The critics of this
approach would suggest that the reason we have very high survival rates is
because we have very early detection, and since we detect the disease earlier,
people who have the disease live longer. So, that may not be the perfect metric
for assessing our health care system. So, what do we do? What else should we look
at? Well, here’s a good intuitive way of looking at a question: you ask people,
which the Commonwealth Fund did, are you confident that you will receive the most
effective treatment if you are sick? And then you look at that by income. People
above the median income, people below the median income, and you compare. On this
scale, the United Kingdom and Switzerland have the highest gap between those below
average income and above average income; and our overall confidence levels are not so high in the United States — they’re
much higher in the UK and Switzerland. Maybe that’s the way to judge a health care
system and figure out what’s best. Another issue is access and financial
barriers. If you believe that that’s the key element, the most important thing
about a healthcare system — you want to make sure that people don’t have
financial barriers — then we’re doing very, very badly. Because 43% of people below
the median income expressed a cost related barrier to health care, and 32%
above the median income — that’s people in this room — have financial barriers
to health care. So, how do we put this all together? Well, let’s look at
one more indicator: appointment times. A lot of people suggest that [in] a good health care
system, you shouldn’t have to wait too long for an appointment the same day or
the next day. Look at the data from the Commonwealth Fund. There, we’re
average. It’s worse in Canada; it’s much better in the Netherlands, in New Zealand,
Austria, Sweden, UK. The real challenge in this game is how do you take these
discrete indicators and put them together to make sense and answer the
question of what is best. And I suggest it’s a very difficult task, and you must
beware what you swallow, in reading what’s said about this. The World Health
Organization put together a composite indicator, which combines many discreet
indicators and ranked the US system as #37. That infuriated people in the United States
who were responsible for health care. But, it may be true. But, it may not
be true. When we look at how the system was put together — it was published in the
New England Journal of Medicine, which is a pretty reputable journal —
and people took the study very seriously, which is why you should watch
what you swallow. When you look more carefully at the study, it’s based on
three criteria. First: overall or average health. And I said we shouldn’t judge a
health system just on overall or average health, we shouldn’t even weight it very
highly because health systems are only one factor contributing to overall
health. The second criterion was responsiveness; is a system responsive
to the consumers? And the third criterion was
fairness in the way it’s financed. Well, they weighted the disability adjusted life
expectancy — the measure of health — fifty percent. It’s a very biased way of looking at it. So, I submit there are many problems. Let’s look at how they looked
at, even, responsiveness. It’s a brilliant, conceptual way to break down
responsiveness. Respect for dignity, confidentiality, autonomy, prompt
attention, quality of amenities, access to social support. But the problem here is
that the survey they did was a convenience sample. If anybody did this in one
of the Wagner classes, they’d be flunked out. They cross-examined,
or they interviewed, one thousand people working for WHO around
the world. And then they generalized about what different countries
think. And it was published by the World Health Organization; that’s another
reason you should be careful what you swallow. The Commonwealth Fund looks at
eleven countries, and makes some of the same mistakes. They have some brilliant
data, we send many of our graduates there, and they’ve done some very important
work. But again, they have four criteria. They look at effective care, safe care,
coordinated care, patient-centered care under the rubric of quality of care; then
they look at access; then they look at efficiency; then they look at equity; then
then they look at healthy lives. And they don’t tell you how they weight
these things, and then they come up with ranks. There’s 67 indicators, 5 criteria;
it’s enough to make you dizzy. But, at the end, you get a rank.
But what does the rank mean? I have a colleague in England at the London
School of Economics who wrote a provocative paper on what I’m trying to say.
He called it “The folly of cross-country ranking exercises.” Let’s
just assume that these numbers are right; that, along the criterion of equity,
responsiveness, cost-containment, and efficiency, England ranks 9-4-7-7.
How would you create a composite score? You’d add them up, you’d
divide by four, and you’d get a score. Suppose the French look at this
and say, “We want to weight this differently.” Well, I mean, you would
have to decide what the relative importance is of each one of these
criteria. The English may weight equity very highly, they may weight
responsiveness a little less highly. The French may look at this and rate
responsiveness much more highly and equity still highly, but cost-containment
less highly. So, depending on how you weight the scores, which is
often hidden, you get a very, very different result. Let me summarize, and then I’m through. How should we rank health system
performance? How should we think about it? Should we think about it in terms of
population health? Healthy populations have better health systems; they’re
healthy because of their health systems? I’d say no. Here’s a few other ways of
thinking about it: I’d love to think about it in terms of a health system
which reduces the frequency of adverse events; I’d like to have a health system
where there are no medical errors. I know in the United States, 100,000 people
die a year due to medical errors, but I can’t compare it to any other
country because no other country has a study comparable to that; so we have no
data, so we eliminate that. What about the probability of survival when you’re
gravely ill? I would love to be able to rank a health system based on if I’m in
the country and I’m gravely ill, my probability of survival is high. No data
to compare health systems on that criteria. Waiting times: that’s very
important for some people. We have a tremendous number of data on that, but
how do we make sense of that data? And who do we ask? Do we ask all adults, do we
ask chronically ill adults, do we ask nurses, do we ask doctors, do we ask
everybody, do we ask people who work-? And suppose you’re waiting for something
which is no good to begin with — then it really doesn’t matter. But we do have
some data there, and the Commonwealth Fund has produced a great deal and we
can study that. That has to do with the perceptions of the health system too, not
just on waiting times but on how it works on any number of criteria. Well, I’ve spoken to you about certain
inadequacies, certain potentials. I now conclude on my two favorite indicators, and that would be, first, something
called amenable mortality. Rather than ranking health systems based on
life expectancy at birth, look at the people who die who shouldn’t die if the
health system is actually performing well. That’s a very important indicator, and there
are studies that are looking at that today. On that score, we do very
badly, I’m sorry to say. And we’ve- Our progress has been very slow. The final
indicator I present to you is one that my colleague John Billings has devised
many years ago, and that’s called ‘access to primary care’ based on what’s called
‘avoidable hospitalizations’ for things you should never be hospitalized for if you’re
well taken care of by the primary care system. For, example bacterial pneumonia
congestive heart failure, asthma, cellulitis. You shouldn’t end up in the hospital for these things. So, we can measure that in different
countries. And I’ve done some international work which suggests, in
that respect, that the United States comes out very badly. We have the highest
rate of avoidable hospitalizations compared to England, Germany, and France,
but we don’t have data on other countries. I’m trying to give you some sense
of that. These last two indicators, I think, are very important in answering
the question of what is best, in conjunction with a lot of the survey
data that’s available and in conjunction with explicit statement of what your
values are and what the relative weights of different criteria should be. But
everything I’ve said so far assumes that the nation state is the relevant unit
of analysis. For those of you in urban planning, you know that’s not always true.
Much of my research has challenged that assumption and examines avoidable
mortality, avoidable hospitalization, and health care systems in big cities. And
that’s what I’m working on for the next ten years. Thanks very much for your attention.

Leave a Reply

Your email address will not be published. Required fields are marked *