Reducing Over-treatment in the Era of Health Care Reform

I'd like to welcome you to today's Medical
Center Hour, which is combined with the Department
of Medicine's Grand Rounds in a program entitled, Reducing
the Overtreatment in the Era of Health Care Reform. I'm Marcia Day Childress from
the Center for Biomedical Ethics and Humanities. We produce the
Medical Center Hour, and again this week
we're doing this in tandem, happily, with
the Department of Medicine. Health care information can
confuse doctors and patients alike. What are the risks
and benefits of tests like mammograms, aggressive
blood pressure control, of EKGs or lung cancer
screening or heart stents? When patients can't accurately
answer these questions, they can find it
difficult to have sensible conversations with
their doctors about their care. And lack of comprehensible
medical information not only interferes with
shared decision-making, but can also lead to
overscreening and overtreatment with deleterious
consequences for patients as well as for the health
care delivery system and for medical system reform. Indeed, today
overtreatment as a result of flawed decision-making
drives a lot of our system's low-value health care costs. In this program today, internist
Andy Lazris and scientist Erik Rifkin join us to assess
this challenging situation. They will then present
as one solution a novel decision
aid that they have developed called the
Benefit-Risk Characterization Theater. They offer their
decision-strategy tool as a way to convey health
information more simply, factually, and in a
non-numerical format so as to explain the
risks and benefits of some common
medical interventions. They'll demonstrate
this approach, and they'll discuss
the potential for it to improve health care
delivery and improve patient– and I dare say–
physician satisfaction, and to reduce overtreatment. Andrew Lazris, who is
on my immediate right, is co-founder of, and a partner
in, Personal Physician Care, a primary care internal medicine
practice in Columbia, Maryland. Erik Rifkin, on my far right, is
an associate research scientist with the Department of
Environmental Health and Engineering in
Johns Hopkins University School of Public Health. He also heads a Baltimore-based
environmental consulting firm, Rifkin and Associates. Together in 2015,
Andy and Erik wrote the book, Interpreting
Health Benefits and Risks: A Practical Guide to Facilitate
Doctor-Patient Communication. Dr. Lazris is also author
of Curing Medicare: A doctor's view on how
our health care system is failing older Americans. It was published just last year,
and you'll find copies of it available for sale
in the bookstore just outside the upstairs
door to the auditorium. Before we start, let me say
that both speakers have signed disclosure forms, and
neither had disclosures related to this program. So I'd like to welcome Andy
Lazris and Erik Rifkin, and in an hour in which they
will tell us some things, and you'll get to
talk to them as well. Welcome. ANDREW LAZRIS: Thank you. [APPLAUSE] Thank you. It's great to be here. I did my residency here. How long ago is it now? AUDIENCE: 23. ANDREW LAZRIS: 23. Long time ago. Several of the people
that trained me are here. Jerry Donowitz is up there,
Gene Corbett, Dan Becker. They all look exactly the same. As I get older,
they look younger. And Erik and I have been working
together for several years when we wrote this book. Erik had written a
previous book where he introduced some
of these theaters, and it really changed the way
I thought about health care and medicine, and I adopted
those into my practice. And since we've been
doing this, we've been approached by a
lot of different groups, from insurance companies
like AHIP, which is a major collaboration of
insurance companies, to AARP, to various academic centers. We're going out
west to do a keynote at a self-insured institution. So a lot of people
are very interested now in this era of
health care reform and coming up with
some kind of system that will enable much more
sensible decision-making. So we just wanted to
share that with you. So the challenge
a lot of us know about– the
Institute of Medicine estimates that every year
we waste $750 billion in medical care that is useless. And just to give you a
sense of what that means, the entire Pentagon budget
is about $550 billion. So just the amount of
waste in medical care is bigger than the
Pentagon budget. 40% to 50% of interventions
that we do, do not help people. They're considered low value. And there's a lot of strong
pressure on the community to change this,
but very few ideas have come out to say this
is what we need to do. This cartoon explains what
part of the problem is. A man with an arrow in
his head, and the doctor is ordering a bunch
of tests to figure out what's wrong with him. And that's often what
we do in medicine. If anyone has not read the
article by Gawande in the New Yorker called Overkill– he is
the author of Being Mortal– it's worth it. You can get it online. And he defines
what low-value care is and gives a lot
of evidence of how much low-value care there is. Low-value care is care that
either does not help people or actually can harm people. And a lot of what we do
falls into that category. A lot of that is derived from
what we do in the hospital. We have a plethora
of specialists in this country that are
responsible for a lot of undervalued care, a lot
of high-technology care, and our country is moving
more in that direction. And as we're thinking about
why health care costs so much, we're not looking
at any of that. What are the barriers to
actually doing something about this? The insurance payment system
is one of the biggest barriers. Our insurance
payment to doctors is derived by a small committee
that meets secretly once a year called
the RUC Committee, and that determines what doctors
get paid for what they do. So a doctor gets paid $1,000
or $2,000 to put in a stent, and another doctor gets paid
a couple hundred dollars to talk about everything
you can do to avoid a stent. That formula was derived
from this committee. Patients also don't have
access to accurate data. The data they get is from
drug ads, from newspapers, from their doctors. The doctors themselves
often don't get data from anything that's beyond
those sources either. And my patients– I'm just a
plain old, private practice doctor. I have no academic affiliations. I see patients five days
a week, and my patients are struggling to find
information all the time. There's also
inherently, in anything we do in medicine,
a lack of certainty. A lot of times as
doctors, we like to go in and be very
declarative and say, "This is what you need to do." You need your mammogram,
you need your colonoscopy. We got to get your
blood pressure down. But even in the most
basic interventions that we explain to our patients,
there's not a lot of certainty. For one patient that might be
useful, and for another patient that's not. There's a lot of variability
from patient to patient. Often when there is
uncertainty, it's a knee-jerk response of doctors
just to do a lot of stuff. So it's a lot easier for me
to order a bunch of tests than to explain
to patients why I shouldn't be ordering the test. It's much quicker for me. So when this patient asks why I
need three CT for a for a cold, the doctor says, "Just do it." My first practice I was
in out of residency, we owned our own lab, and
the head doctor of the group told me, he said look,
order labs on everyone. They'll think you're
doing a lot of stuff for them, that you'll get
them out of the room quicker, and we'll make money
on the blood work. And he really believed. So I think we can convince
ourselves to think this way. Quality indicators. We are now being
judged by what's called quality and value. But quality is a very
1984-esque concept. Quality means we follow a
bunch of very set parameters that we have to insist
all our patients do. So for instance, if I talk to
a patient about a mammogram and explain the pros
and cons of a mammogram, and the patient decides
not to get the mammogram, I'm going to get an F on
my quality indicators. To talk to a
patient and may have them make their own
decision will hurt me. So it's much more sensible for
me to just tell the patient, as this cartoon does, please
don't waste the doctor's time and questions. Just go and tell
them what to do. There are some
groups– like there's a group in Boston called Care
That Matters that has looked at several of these
quality indicators, especially the
ones in Blue Cross, and we're now looking
at ones in Medicare. And the amount of
them that actually have strong scientific data
behind them are minimal. If you look at that
site, Care That Matters, they have a red light
for quality indicators that cause harm, yellow light
for ones that are neutral, and green light for
ones that actually have some good evidence. And there's one green
light indicator. But yet that's how
we're getting paid. If anyone hasn't seen this
book, it's a worthwhile read. Michael Lewis is the man who
wrote Moneyball and The Big Short. And he's talking about
two Israeli psychologists who talked about
how we think and why we're not rational
creatures in our thinking. It's really
important when you're talking to patients, or as a
patient when you're hearing information, to know our flaws. So first of all, we have a very
poor understanding of math. When people throw a lot
of numbers at patients, the patients' eyes
tend to glaze over, and it's very difficult
to accept a big number. When I talk to
patients about numbers, the bigger the
number is, the more they're just going to
start staring at the wall. So when we talk about
risks and benefits, using numbers can
sometimes be an impediment. Anecdotes trump statistics. We know that when people demand
antibiotics for a common cold, and we give them all the
evidence that antibiotics don't work for a cold,
they'll say, yeah, well, it works for me– or my aunt gets
antibiotics all the time and it works for them– because
statistics are meaningless. In this book, they
talk about a study they did where they tried to
track arthritis and weather– the correlation between
arthritis and weather among a big cohort of patients. And there was no correlation. It was a scattergram. The weather had nothing
to do with arthritis. So they brought that
information back to specific people in
the study and said, look, there was no correlation
between arthritis and weather. And they all said,
yeah, that's not true. On Saturday it was raining
and I really hurt a lot. So they point to that one
day, and that means more than the actual facts
even though they were involved in the study. Single events carry weight. We know that in a non-medical
way that the country's been growing jobs– about
200,000 jobs a month– but when 600 jobs are
saved at Carrier Plant, that seems to have a lot more
impact than all those 200,000 jobs, because those big
numbers just get lost. But if you can point
to one single event, that seems to be more important. And people are also willing
to gamble if the gambling is to prevent a bad outcome. So often when things
are presented– and you'll notice this in
medical presentations– they're presented that we
can prevent you from having a stroke if we do this. We can prevent you from
falling down or breaking a hip if we put you on this drug. It's always in the negative. The example they
use in the book is if you're given a choice of
either having $500 taken away or a 50% chance of either
having no money taken away or $1,000 taken
away, most people will take the second option. They would rather take a chance
than just have $500 taken away. And so it's really
important when you talk about any kind
of medical intervention that you do something that
is non-numerical, that is very familiar,
that it's something that people can relate to. And that's one of the
ideas that we had when we came up with the theater. So there's a lot
of misinformation. We all know it. We're getting pounded by
very inaccurate information that's in the newspaper,
that's in pharmaceutical ads, and that, to us, is one of
the biggest barriers of why we cannot have health care
reform that's meaningful. ERIK RIFKIN: Thanks for coming. Andy and I always dress
a bit differently, and I think it has to do with
our backgrounds a little bit, but I'm not entirely sure. So Andy mentioned a
number of barriers, and how do we overcome
those barriers? And I thought I would quote a
line from a recently new Nobel laureate, and it's the answer
is blowin' in the wind. And that particular phrase
has a lot of meanings, but in this particular
context, we all know about the need for
sharing decision-making, and we all know
about the importance of empowering patients. And we all know about
why these things are not happening in the value
of the decision aids, but the progress
has been glacial. And people are trying,
but they're not quite getting there. So one of the
reasons they're not getting there is because
doctors in general perhaps don't understand risks
and benefits as well as they should. A recent article in JAMA– it was in August of 2016– basically said that
the– it was done at the University of
Maryland School of Medicine, and over 79% of
the doctors weren't really sure about the
benefits and risks, and they overestimated the
benefits of intervention. And most respondents
weren't competent about their knowledge. And as you can see
in the last point, studies show that patients
consistently overestimate benefit and underestimate harm. This is generically the case. And while this was done at
the University of Maryland, the chances are relatively
good that similar numbers could be found at medical
schools throughout the country. Perhaps this is
the one exception, but it's been our experience
that that, indeed, is the case. So what really are the problems,
and how can they be addressed? So one of the things
that should be noted– and it should be given a
great deal of importance– is the concept of relative
and absolute risks. And risks are always given to
you by pharmaceutical companies and then reported to the
newspaper in relative terms. And we will explain that in a
moment with a video we have, but doctors should always
know the difference between relative
and absolute values. And in our view, if you're
communicating with patients, it is absolutely necessary to
use absolute values rather than relative values. The reality is that
while high-value care has been adopted as a
professional responsibility, the story of how this is
implemented in practice is yet to unfold. So what do we need
specifically to move ahead? What is the path
forward, and how can we change things, rather
than by continuing to tell you that
we're not really there and the process has been slow? The information using
the decision aid– or using the tools–
should be presented to patients that does not use
numbers or math in any way. It's virtually impossible to go
in concerned about your health and being asked to interpret
values and statistics that most people
aren't aware of. And secondly, it's
always good to tell people what the
benefits and risks are of medication or of surgery
or of screening tests, or other types of
medical intervention. The format must be clear and
it must be straightforward. And we should always, as
I mentioned a moment ago, use absolute values, and the
denominator should be 1,000. So when you're talking
about benefits and risks, most people can see 1
in 1,000 or 10 in 1,000. That's an important point. Now all these points that
I'm making are not my idea. All these points have been
referenced in extensive review articles and corporate
studies, and these are the things they say are
necessary for a decision aid that can really make
us get to the next level– for a universal decision aid. But unfortunately,
that hasn't happened. And it hasn't happened because
what has been missing in all of the decision
aids that you see– although some of them are
sophisticated and very well put together– is the lack of familiarity. And by a lack of familiarity,
I mean a patient has to see, and you as a doctor
has to see, how this can impact them immediately. And how do you get that? You don't get it by having
rectangles and squares and you show dots in those
rectangles and squares, because people can't really
relate to what that means. It reminds me of
a Far Side cartoon when the owner is saying, "Blah,
blah, blah, Ginger, blah, blah, blah." And so if a patient
goes into a hospital, and he thinks he may have
heart disease or cancer, what he hears is, "Heart
disease, blah, blah, blah." We have to have something that
gets back to a more real level. So we thought that
a theater seating chart would be the way to go. Everyone's been to a theater. Everyone's been to a
concert, and everyone's been to a ball game. And we can sort of
place ourselves directly in that theater, and we
can see how that will relate to what we're doing. So we developed a
theater of 1,000 people. It's a hypothetical
theater, but we thought it would work well
in discussing health benefits and risks, and we call it a
benefit-risk characterization theater. As you can see– I think you can see that– that's a theater with
exactly 1,000 seats, and these seats
represent individuals. Now these are absolute numbers. We talk about a
whole population. And so if 2 people benefit, then
there'll be 2 of those seats will be blackened in. For example, if you look at
the theater on your left, there are 3 blackened seats. So let's say, for example,
that those are the benefits from taking a cancer drug. You can also see on
the other side, which represent the risks, that
there are about a third of those seats are blackened in. So in that sense,
the drug is going to be helpful to 3
people out of 1,000, but it won't be helpful
for 300 or so out of 1,000 because of the adverse effects. And you and your patient can
sit down and talk about this– a unique, qualitative
information. But with this
information, a patient could see intuitively,
immediately whether the risks are better than the benefit. And you can turn it around
the other way, where the slide on your left has
3 blackened seats and those are the risks, whereas
the benefits look extensive on the other
side of about 300 of those seats blackened in. So we think this is a very
intuitive way to move forward, and we think it's a way
that doctors can communicate effectively to facilitate
communication between patients and doctors. Could I ask–
we're going to talk about relative and
absolute risks in a moment, but we have a video which
does a much better job than I could do it, so if someone
could turn that on, I'd appreciate it. You can see how relative risk
can distort the real benefits and the risks of something
like having a mammogram because they only use
as the denominator a very, very small number,
because that number is compared to a group that has had a
another very small number. And it's 4 versus 5, which is
20%, but it's still 1 in 1,000. And patients should
only see that 1 in 1,000 because epidemiologists
doing cohort studies should be using
relative risk and risk ratios, but not
doctors when they're communicating with patients. So what I'd like to do in the
next few minutes– that I have, at least– is talk about
atrial fibrillation. And the question is, how
do patients determine if Coumadin is right for them? And the reason I'm
using this as an example is back in November of
2015, I had a heart attack. And I was told that
I had Afib, and I was told I needed to take Coumadin. But I asked a few questions. I asked them, could
you please give me the relative and absolute
values, but preferably the absolute values, of
about how much benefit will be available to me? And I got a blank stare. I said, well, I was just
wondering if you could tell me the difference between people
who take Coumadin and don't take Coumadin and get the
absolute risk reduction, so that I would know whether
the number was meaningful. And I still got a blank stare. I was told that
Coumadin is necessary if you want to prevent stroke. And I said– and I'm
not being glib here, although it sounds like it– the answer was you know
how serious a stroke is. And I told him not
as much as you did, but I really need an answer
to some of these questions because I coauthored
a book recently, and I have chapters on
this stuff, and come on. [LAUGHTER] So it didn't go too far, and I
was told by a number of people independently that I really
needed to be on Coumadin. So the final question I
asked was, I'm already on Plavix and aspirin. I drink half a
bottle of wine a day, and you're suggesting
another blood thinner. What would be the risk
from that, and they said, I don't think you understood us. We told you already
that having a stroke is really not a good thing. So I was in touch with Andy. Andy has been my
personal doctor as well, and he was giving me
advice as I was groggily going through these sessions. But the standard is that if
you have to treat patients with Afib using anticoagulants
such as Coumadin or warfarin, and these medicines
can be dangerous, and even though
they're dangerous, it's best to go ahead
and take them anyhow because in the long run,
it's in your best interest. So we wrote an
article in JAMA called A Grateful But Not Passive
Patient, and it referred to me and Andy and I coauthored it. And we talked about some
of the things that happened and how I felt and
why they happened, and we talked about the
Coumadin experience as well. And although this
is all interesting because I was lying on that
bed thinking, my goodness, I've written a book about it. Should I listen to
these guys or not? I don't really think I'm sure
about what they're saying. I don't think they're talking
to me about uncertainty. But what the doctors usually
say, and what they said to me– I'm looking down at
this because this– I have to read this [INAUDIBLE]. "If you have Afib, which can
increase your risk of stroke, with Coumadin your risk of
stroke is reduced by 50% and, therefore, the
benefits are significant." But those are relative numbers. They are meaningless numbers
in talking to someone like me in that state. And I said to them, you know,
you're using relative numbers. And again, there
was no response. And I said, there are downsides
of potential serious bleeds, and your blood has to be
monitored periodically, but the benefits clearly
outweigh the risks? I brought up the question
again and it was ignored again. So what we did– our book that takes some of
this information, but Andy and I went back to see what
was really going on. So here's a theater that
shows 6 blackened seats, and it talks about the
benefits of Coumadin. And in this theater of
1,000 people with Afib who take Coumadin,
approximately 6 will avoid disabling stroke
compared with 1,000 people who take aspirin in a given year. So I was looking at that and,
OK, what about the other side? What about the risks? The risks of Coumadin
has 12 seats in it, and these seats are designated
the risks associated with bleeds, and you can see
here that 6 of those bleeds will be in the brain
and cause death. So the option was
essentially dying of getting a stroke or a
bleed were about the same. I really didn't know what to
do, and then we thought, well, what the hell? Let's think about
your situation– your specific situation. You're taking Plavix and
you're taking aspirin. Let's see if we can
find some data which compares that plus Coumadin. And that's the information. The information shows another
theater of 1,000 people, and 70 more of them
will suffer major bleeds if they take Coumadin
with Plavix and aspirin. And that was recommended
to me by every surgeon, every individual in the
hospital at that time. So that's 7 out of
100– that's 7%. And I made the decision
in collaboration with Andy that I didn't want
to take that risk, so I didn't take any Coumadin. But it would be nice if all
of this kind of information were available to
patients all the time. It's not hard to
do these theaters. It's not hard to prepare
them and to communicate the information so that
both you and the patient understand exactly
with the issues are. And more specifically,
what's really important here is that this decision– it's a patient-oriented
decision. It's not a medical decision
or a scientific decision or an analytical decision. It's how risk-averse you are. And we're all
different, and we all have a right to be as
risk-averse or not risk-averse as we wish to be. And that should be
the focus of what we do or don't do as well
as doctors are giving us the right information. And with that, I'm going to
turn this baton back to Andy. ANDREW LAZRIS: So so far
the decision about Coumadin is working. We'll see how it goes. What Erik didn't
tell you is he left the hospital on potentially
five or six medicines that he was told he had to
take for the rest of his life, maybe more, including
hydrostatins, two other medicines that lowered
his blood pressure to 90. He couldn't get up the steps. He was dizzy. He was told he needed a stent
put in one of his arteries that had nothing to do
with his heart attack. He chose not to do that. He's now on two medicines–? Aspirin and low-dose statin and
no stent, so we're hoping we're right. I want to talk about the
use of these theaters in things that might be
related to health care reform. Interestingly, I'm
using EKG as an example because the insurance companies
were really focused on that when we talked to them. They felt like, especially
for self-insured people, we're constantly
asking for EKGs. And for that matter, in
my community where I work, cardiologists are
ordering stress tests pretty much every year
on asymptomatic people. And my patients
come back and talk about how thorough the
cardiologist is for doing this. But my patients frequently
ask as part of their physical exam– whenever that might be– if they should get an EKG. They've always had one. So I've used these theaters
to dissuade people. So the information about EKG– you can even look this up in
the US Preventive Services Task Force, which has a lot
of good information, but patients that I have
don't really understand it– is that if you
take an EKG on top of the clinical information you
already have, it adds no value. It will pick up no information
that you can get from– that you can't get from your
own clinical information, and it will save no lives. And this, by the way, is
true of a stress test. Even a stress test with Valium
in an asymptomatic person has never been
shown to save lives. However, an EKG has an
incredibly high false positive rate. About 800 out of 1,000
EKGs that are abnormal are abnormal in a normal heart. And so when you get an
abnormal EKG, what do you do? Very often a cardiologist
will send them for a stress test that
will also have a similarly high false positive rate. They may then go for
a catheterization. I just watched my father
go through this dance, and my father used the
thorough word again. And the doctor is just– and the doctor was
considering putting a stent in something
that was barely blocked, which is where I
put the brakes on. But this is what happens
when you have a test that is very inaccurate like an EKG. You'll end up getting false
information that will lead you to further testing,
perhaps treatment, and you'll cause harm. And the theater on
the right actually shows that there's 20 serious
consequences that occur out of 1,000 people who get
an EKG without symptoms, and those could be stroke,
heart attack, or death. Major complications. So it's something I have
to explain to my patients. They say, what's the
harm of getting an EKG? There is a harm in getting
tests that actually are very inaccurate, which
are what most tests are, and in using them just
in screening in people who don't have symptoms. And my ability to
use these theaters has been very helpful for
me to dissuade patients even from getting their
annual stress test. What if I had a normal EKG? A lot of people say that I
just want to make sure I'm OK, so I'm going to
have a normal EKG. So this is how many people
who have had a heart attack would've had a normal EKG the
day before their heart attack. So having a normal EKG
doesn't tell you anything. You could get that normal
EKG, jump up and down, and have a heart attack
on your way down. And it's true of
stress tests too. Most heart attacks occur
in places that are not blocked more than about 20%. They're minor blockages. And those won't be picked
up by EKGs or stress test. And that goes to another
point that I always bring with my patients, which
is if something is blocked, a lot of times a
cardiologist will say we're going to put a stent in there. Now Erik had a stent
put in because he was having a heart attack,
and I call that symptomatic. That was the biggest
symptom you can have. A lot of my patients get stents
put in without any symptoms. I had someone recently
who came to me with a cough and a little
chest pain from the cough, and we were doing all
the usual stuff to that, but he saw his cardiologist. The only reason he was
seeing a cardiologist is five or six years ago
in the emergency room he had a murmur
detected, so now he's going back for annual
echocardiograms. The cardiologist said, well,
let's do a stress test. Next thing you know, he had a
stent put in a 70% blockage. He came back to see me
and he was still coughing. He was wondering
what was going on, and now he was on
Plavix and aspirin for at least a year,
which confers a risk. So when I told
you about the fact that most heart attacks occur
in non-blocked arteries, would indicate if you
fixed the blocked artery, you may not prevent
a heart attack, and this is what's been
shown in the literature. So in asymptomatic
people, people who get stents put in
arteries that are not the left knee or
proximal LAD, there is actually no benefit
over just taking an aspirin and a
low-dose statin. So you're getting
no benefit in terms of the major outcomes, which
are death and heart attack. However, in people
who get stents, there is definitely some
severe complications. So approximately 20 of
those people in the theater will have major complications
from the stent itself, whether it's a major
bleed, whether it's a complication from the stent
not working, closing off. You could have a much worse
outcome having a stent than not having a stent. And when we talk to insurance
companies about this, my question is, why are
you paying so much money for these stents when, in
fact, it confers no value and actually can cause harm? And that's the way
it's always been. Cardiac revascularization
costs about $50 billion a year. Bloomberg News
estimates $2.4 billion spent on unnecessary
stents, and that is really a lowball figure. There are people who would
give a higher figure than that. We found a couple
studies that looked at how many cardiologists
recommend stents with a discussion. It's about 25% of
cardiologists typically discuss the risks and benefits
of stents before offering them. Most people just said
you need a stent. That was the discussion. And most patients
inherently believe– like what I talked about
earlier– if you could feel something's right, it's
right, so they inherently believe if you open up an
artery, that makes you better. So again, using the
theaters can help people see it a different way. There's a test that Medicare
has recently offered to pay for, which is CT screening
for lung cancer. And it's an expensive program
that has been studied. There's one big study recently
that suggested a 20% reduction in lung cancer death in people
who got this test versus people who didn't. And that was over five years. So you get the CT
scan every year, the study was done
after five years, and you saw how much
the reduction is. But again, that's
a relative benefit. The real benefit is
3.3 out of 1,000 people who had this test compared to
people who didn't avert a lung cancer death in five years. So again, it depends
on what you think. As a 50-year-old
smoker, which is what we're talking
about getting this test, is that enough for you to
get these annual tests? Is 3.3 something you say, yeah,
I could be one of those seats? Some people yes, some people no. It certainly feels a lot smaller
than a 20% risk reduction. These are inaccurate tests. Very inaccurate. And approximately 675 out of
1,000 people over those five years will be told at least once
they have an abnormal CT scan. Some of them will be told
we'll keep an eye on it. You might have
cancer, you might not. Or we'll keep an
eye on it, we'll check you again next year. Others, though, have to
go for other testing. So in this big study and
some subsequent studies, 233 needed additional testing– typically with a PET scan,
sometimes with a biopsy. And there were nine
people in this cohort that were harmed
over the five years from this additional testing– popping the lung,
infections, pneumonia, things that can come from biopsies. But even beyond the
harm, to show people the entirety of the information,
there are people who have said, I don't want to live with
that uncertainty of knowing I might have an
abnormal test and I don't know what to do about it. That level of stress is not
something a lot of my patients often want to accept. And I have to say that I've
had very few patients who were willing to go
through this test after I've shown these theaters. There's another option for
the 50-year-old smoker. If the 50-year-old smoker
simply quit smoking– at age 50 there are
approximately 200 deaths averted in the next 25 years
just from quitting smoking. And how much does
that cost Medicare? And what are the side effects? How many people are
harmed from the process? There should be a couple CEOs
from the tobacco companies in there. They didn't get harmed. But again, to juxtapose
the lung cancer screening with the smoking quitting has
convinced a few of my patients to quit smoking because
they came in gung-ho to have this test, and when they
see this other thing they realize just
how absurd it is for them to continue to smoke. I was telling Erik– we
talked about a lot of people don't understand risk. I have a lot of smokers
who continue to smoke, but they won't go
down to Florida because they're afraid of Zika. Sometimes you have to,
with smokers especially, show them the numbers
compared to something else. So lung cancer screening– Medicare estimates that it's
going to cost about $9 billion a year. Again, not an
insignificant figure. Less than 50% doctors
in all of the states actually did discuss
the risks and benefits. And, by the way, Medicare
mandates that you discuss the risks and benefits. This is the only test
Medicare mandates you do it. They do not give
you how you do it, so a lot of people in
this study discussed it by saying there's a
20% risk reduction and there's some side effects,
which is how a lot of people do. Again, emphasizing the
benefit and the risk. And, of course, the cost
of smoking cessation is not bad at all. So clinical guidelines I
talked about a little bit, and can you see
the use for that? Again, clinical
guidelines are based on information
that's very generic and doesn't apply to your
specific patient sitting in front of you in the room. I'm a geriatric doctor. I take care of a lot
of very old people. And in my book, Curing Medicare,
I tell a story about someone from the Orange Clinic
who I took care of– and Gene Corbett probably
doesn't remember this because it was a tough story. Old guy on a farm in his 90s. And I went out there,
and his blood pressure was out of control high– 220 over 110,
something like that. And I had just read
the SHEP study, which is a new study that said
systolic blood pressure that's high in the elderly
will cause strokes. And I came back to Gene
and John and I said, we got to treat this guy. And they said, you
know what, no big deal. Just relax. He's OK. He's still farming. But I insist on it. They let me put him on, I
think, a low-dose beta blocker or maybe hydrochlorothiazide. And I was going to check on
him next week, and he died. Now whether he died because he
needed every bit of that blood pressure to get into
his narrow arteries– he needed 200 [INAUDIBLE]. His body was doing him a favor,
and I, the brilliant doctor med resident said no, that
doesn't fit with SHEP. Shows you just how we have to
view everyone individually. Especially in very
old people, I've found that high
blood pressure is crucial to their very survival. If you get their blood
pressure down to normal range, they get confused,
dizzy, they fall, they're tired all the time. One of my favorite patients
who died recently– well in her 90s, exercising
exercised every day– typically had a systolic blood
pressure in the mid 200s. And every time she
saw someone else, they kept trying
to get her down, and she ended up in
the emergency room. Clinical guidelines do not allow
us to have that flexibility. So this Sprint study, which
a lot of people are aware of, suggests that in
older people who are high risk for
heart disease, there's a 20% reduction in death with
aggressive blood pressure control, which is blood
pressure below 120. And Medicare has adopted that as
one of its clinical guidelines. I have to indicate if
my patient's blood– my 99-year-old's blood
pressure is above 120, I have to give a
plan of what I'm going to do to get it below. If it's above 140,
then I actually get dinged as failing
the guideline. So what did the SHEP
study really show? So people with high risk
of coronary disease– and it was a very skewed study
that I talked about a lot in some of my blogs– 3.5 deaths were
averted in a year in people who had aggressive
blood pressure control. In that same group, there are
20 severe, life-threatening side effects, which the
article poo-pooed, if you read the article. And if you look at other
studies that preceded and came after the
Sprint study, they're actually in people who had
aggressive blood pressure controls in other cohorts– not people with severe
high risk coronary disease. There were eight to
ten additional deaths in this group when you had
blood pressure lowered too much. So this was not a
consensus study. The Accord study, the HOPE-3
study, which came after Sprint, and a large VA study if
it were real disease, all showed a higher
death rate in people who had aggressive
blood pressure control. And Sprint had about 3,000
or 4,000 people in it. These studies collectively
have well over 100,000 people, and yet Medicare's guidelines
still insist that we do this. A lot of people know ACP,
American College of Physicians, and American Associated
Family Physicians have changed their tune and
gone against that guideline. But as I'm doing my
quality indicators– as I do them every day– Medicare hasn't changed at all. So using these kind of theaters
to look at all the quality indicators and then
actually individualize them for patients would make
sense, and perhaps our quality indicators would be
a much better judge if it just allows us
to have a conversation. So with this example or
with the mammogram example, as long as we can prove we
had a conversation, then perhaps that alone
is the indicator. The patient can decide
what he or she wants to do. We conducted focus groups. Now, they're our own
biased focus groups, and we provided
food so people were happy [LAUGHTER] so they may
have just done what we said. But we thought they were
relatively reasonable, and people were very
turned on by these theaters because they immediately saw
what risks and benefits were. We showed some drug ads. We gave them some numbers that
were in the newspaper recently, and then we showed
them the theaters and they saw the difference. And they really felt
the familiar setting when we asked them
at the end what was the most important part. They could actually see
themselves in the theater. Some of the theaters in
the Kaiser Health video that we showed we're
a lot more graphic, and if we could do those all the
time, that would be wonderful. One caveat, there's something
called Number Needed to Treat, which a
lot of people use, and we threw that out
there too as a possibility. People were utterly confused
by Number Needed to Treat. They did not know
what that meant. When we said 6 out
of 1,000 people get strokes who were on
Coumadin compared to aspirin, and then we did Number
Needed to Treat– if you treat 120 people and 1
person will avoid a stroke– this was too much. So we really felt
the Number Needed to Treat– too numerical. People don't like that. So the benefits. The benefits we've
discussed already. The biggest thing we have found
is that this allows us to have shared decision-making that
puts the patient in a seat the patient deserves to be in– a patient who can understand
what's actually going on and gets actual
information, and to avoid some of this overtesting
and overtreatment that is listed below,
like the guy going to the perfectly healthy but
worried about cancer window. Or a lot of those
screens where you see they check your carotids,
your abdominal aorta and your legs for $100. I thought this was just
funny so I put it in. They found cholesterol pills
clogging up his arteries. It follows what's going on. So creating sensible
clinical guidelines, incorporating this into
electronic medical records we think would be
very important. So as the guidelines come,
the theater comes right up. And you can go
through it and then you indicate you
had the discussion. Using this to change to a
value-based payment system. So instead of paying doctors for
low-value care lots of money, you pay doctors for
high-value care lots of money, like primary care doctors. We give the highest value care. That's just a fact. It can be used in the media. We talked to places
like the Baltimore Sun, and we said whenever you
publish a health article, put one of these
in because then you can really see what's going on. But that's not going to
make front page news. This Sprint study
makes front page news because it shows a 27%
reduction in death. You see those
three little seats, that's going back to page 20. And we thought creating a
patient-oriented website would be a very good idea. So lots of further study
could be done with this. How much acceptance
does this really have if you get a big
kind of group analysis? Will it actually
reduce overtreatment? There is some evidence
that decision aids do reduce overtreatment. What's the best
way to utilize it? Is it having a website? Is it having
doctors initiate it? How could they be
easily customizable? Because we know, for instance,
someone with a strong family history of breast
cancer will have a different theater than
someone with no family history. So we need to be
able to customize it. And measuring the intensity of
effect, because as Erik said, dying versus having a hangnail
are two different things, and it's important
to distinguish those. So we do have a blog, and
every time a new article is printed we'll throw a
theater up on the blog just to put it in that format. And we do also go through cancer
screenings and other things periodically just to look
at what's newest out there. So questions, I guess? ROSS BUERLEIN: My
name's Ross Buerlein, and I'm with the Chief Residents
for the Department of Medicine, and on behalf of the
Department of Medicine, thank you both for
coming and for your talk. As health care
providers, we seek to form a unique relationship
with our patients during a pretty vulnerable
time in their life, and they often grant us
a bit of omnipotent trust during that time. And I think it's up to
us to form some shared decision-making models and
improve our communication with them during those times. Your models definitely highlight
the importance of that, so thank you. If anybody has
any questions, I'd be happy to bring
you a microphone. AUDIENCE: I can just speak up. I'm Daniel. I'm one of the third-year
medicine residents, and we're doing a lot of
screening colonoscopies next year, of course,
which are of very benefit. But I was more curious about–
there's a website, the NNT. I don't know if you guys
were familiar with that or had any thoughts about it. It's basically a
look at the number needed to treat for a whole
lot of common conditions. I don't know if you
had any– you thought it's useful or not useful. ANDREW LAZRIS: I know
the NNT very well. I know the person who does it. He's part of a group
called the Lab Institute. He works with the Lab Institute,
which I encourage all of you to join because it
deals with this issue. The NNT has terrific
information in it. And, in fact, some
of the information that I obtained when we wrote
the book was from that website and from talking to them. The concern I had with
it was, number one, it looked at
information two ways. One was number needed to
treat, and the other one was similar to more of
an absolute risk benefit. You could click to either one. Very wordy and not
very pictorial. And when I show a lot of
my patients that site, it was just too much for them. I think as a doctor
it was helpful to me. But again, it lost
the familiar setting that I found was most
useful for the patient. ERIK RIFKIN: Just one addition. The number needed to
treat is a reciprocal of absolute risk reduction. They both are
accurate and here's another that the ability
for patients and doctors to communicate using that
technique is a little bit difficult. AUDIENCE: Absolutely. I think it's more of a
provider-oriented website that it is a patient-oriented,
although they were a little more
optimistic about Coumadin than I think you guys were. ANDREW LAZRIS: No, no. Actually a lot of our
doubt is from that site. If you look at the
absolute, it's very similar. And you have to understand that
the difference that we have on ours is our strokes that
actually persist– what are called disabling strokes. If you actually say just strokes
that get better tomorrow, then your numbers are
going to be higher. That's like a lot
of the studies that have looked at bisphosphonates
for spinal compression fractures look at
compression fractures that nobody ever
noticed even exist. You start distinguishing from
painful compression fractures, your numbers are a lot smaller. So we thought it incumbent
upon us to look at things that actually impact patients. AUDIENCE: Andy, when you have
a discussion with a patient and you say there's
relative risk [INAUDIBLE] and they make a choice
for non-intervention because something bad
happens, which might happen, what's your discussion
with the family, with the patient
themselves after that? ANDREW LAZRIS: I'll always,
when I have a discussion say that something
bad can happen. And, for instance, Coumadin
is a big example of that. You might have someone who– I've done enough of this– hasn't happened
yet– that someone might get a stroke when they
decide not to take Coumadin. But I make it very
clear at the beginning that this risk exists,
and that it's up to them which risk they prefer. If we knew which risk we were
actually going to be hit with, that would make the
decision a lot easier. And that's true. Bad outcomes occur. People who have chosen– Erik, we talked
about colonoscopy. We have data on that. I looked at the data and
decided to get a colonoscopy, or looked at the same
data and decided not to. If Erik gets colon cancer,
will he kick himself? If I have colon polyps
from the colonoscopy, will I kick myself? The point is that bad
outcomes can occur either way. And we have to let them know
with the intervention too. If we push Coumadin and
they bleed in their brain, we have to make that very clear
also that that can happen. So I make it clear that
whichever choice you pick might be the wrong choice. And I wish he'd do it. AUDIENCE: What happens
if you find it? You lose that patient? ANDREW LAZRIS: I don't. I actually find that because we
had that discussion beforehand, it's actually something
that people appreciate. AUDIENCE: Hi. Thank you for your talk. It's great. I'm in one of the
hospitals here. So my question is,
how do you partner? Do you partner with the
consultants in the community in that some of what you're
potentially recommending or sharing with patients
may be contradictory to the cardiologist or
gastroenterologist, whoever. And so are you
making connections with those specialists
with this information to make sure that you're in
some ways on common ground and that you're not somewhat
adversarial with the patient in front of you,
so that you share trust as opposed to losing it? ANDREW LAZRIS: I think
that's a great question, and that's a question
that's been, in my practice, very difficult. We have a
group of very aggressive cardiologists where I work that
literally stress test, echo, and holter everyone once a
year, and do a lot of stents. And these people, over the seven
years I've had my own practice, know me pretty well. So the relationship
they have with me might be different than
the relationship they have with other doctors
because my patients walk into their office
with information that they give them, and I
say have the doctor explain it to you. My relationship with them
is through my patients. I say, your cardiologist
has an obligation to explain to you why he
or she is doing this test, and what the outcome is
going to be if you do it and if you don't do it. Here is the information I have. I show them– we have
peers on all this– and I say so if they have
different information, I'd be interested in
having that information. And I've seen the cardiologist
doing much fewer tests on my patients. But they don't want to do that. So I don't know– I think I have a
great relationship with the cardiologist
in town, but they know where I'm coming
from, and that's helped. I think just as a
primary care doctor, just being pretty bold– and I believe the
primary care doctor– I can take care of
pretty much everything, other than intensive care,
that patients run into, and I feel like I
have every right to talk about the cardiology
world as the cardiologist, and I make that
very clear to them. And that's worked. I still have a
great relationship with the cardiologists,
and my patients sometimes hear two different stories. I'll say well, there
are two stories and then here's the facts, and
you can make your decision. And a lot of them do
choose to continue getting the annual stress test. It makes them feel good, even
though deep down they know. I had someone recently wanted
a PSA even though we've had this discussion. And they said, look, I just– if it's normal
I'll feel so good. I said, well, what
if it's not normal? And probably you
shouldn't feel good if it's normal either,
or any more normal. They said, I just will
feel good if it's normal. Just do the test. But we're not rational
creatures at the start of it, so you have to
work with patients. Not all specialists
are like this. The person in charge of this
institute, the Lab Institute, which deals with overtreatment
is a cardiologist. So you find the ones who are
willing to work with you too, and that's really
important also. ROSS BUERLEIN: Time
for one more question. AUDIENCE: Thank you, sir. My name's Mike Williams. I'm actually a surgeon,
so forgive me [INAUDIBLE]. ANDREW LAZRIS: [LAUGHS]. If we If we were on
an island, deserted, I'd want you there, not me. [LAUGHTER] AUDIENCE: I've been that guy. So my question is– and
I love this information, love the perspective you
guys are providing for us, and I'm hopeful that
we can find a way to disseminate this
across the country, because we need to, obviously. My parents are having this
talk about unnecessary tests that they're undergoing
and they're in their 80s. If I live to be 70, I've
had my last colonoscopy, because whatever
happens after that, that's what's going
to happen to me. But my question is,
there's one risk that we haven't
talked about that I haven't heard mentioned today. Increasingly, as
insurers shift cost back to individual patients
through their employers who can't afford the
health care premiums, those costs shift back to
the individual pocketbooks. One of the risks, the
one with the black dot, is how much it's going to
cost me, the individual, for the test. Because increasingly, I got
to pay for the test, which I'm OK with, and that's part of
why we have the problem now is, get a test, because somebody
else is paying for it. So I'm wondering if
you have considered the possibility of adding in the
direct or indirect costs of all these tests people get to them– as one more piece
of data they can use to make a decision of
what the value is in the test. ANDREW LAZRIS: There's
certainly a lot of people in the
political world who have talked about that,
from the right and the left. Paul Ryan talks about that
all the time and believes– I don't agree with
everything the man says, but he does believe
patients should have a stake in everything they do. He's even talked about
getting rid of the 20% copay that you can get– I mean, 20% secondary insurance
you can get with Medicare because he says
it takes that away from patients– that patients
can make a better decision if there's money involved. And yeah, that may well be
part of the equation too– that they're much
more willing to look at this– pros and cons. Again, I found in
my own practice, even in the Medicare
world I live in, where most of my patients have
secondary insurance, that they still don't want a lot
of tests and they're looking for any way out of it. And this actually gives them
the freedom to move on and say, I don't need that. So even without
the money part, I think this has
been very helpful. MARCIA DAY CHILDRESS: I'd like
to thank Andy Lazris and Erik Rifkin again for
a wonderful hour. I'm sitting here thinking
about clinic rooms with pads of paper that
has a theater on it and Sharpies so you
could color in the– ANDREW LAZRIS: That'd be very– MARCIA DAY CHILDRESS: –squares. We hope that you can join us
next week for Medical Center Hour. The topic is Ethics and
Emerging Threats to the Health of Pregnant Women,
Zika and Beyond, with Annie Drapkin Lyerly
from UNC Chapel Hill. So please join us then. And again, please thank
Andrew Lazris and Erik Rifkin. [APPLAUSE]

Leave a Reply

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