Eric Schmidt at HIMSS

Uploaded by Google on 04.03.2008


ERIC SCHMIDT: Thank you very, very much for that kind
So why would Google be here?
A very interesting conference, not a conference we would
normally attend.
And one day we started thinking, how much do people
use Google for health?
And then I realized, what's the most important
search I could do?
So I typed, how long will I live?
It seems like a reasonable question.
And up comes an age calculator.
And I programmed it, and it said 67.
Wrong answer.
ERIC SCHMIDT: So I reprogrammed it again and I
got 86, right answer, and I'm done.
So you see Google is very important to my health because
I have to meet the new programming goal.

Nearly one out of every two Americans have one or more
chronic health conditions.
I was struck by this.
You all are doctors, medical professionals,
you all know this.
I didn't realize it.
Hypertension, arthritis, respiratory diseases,
cholesterol chronic mental conditions, heart disease, eye
disorders, asthma, and diabetes, and these are people
who live with these things day in and day out, and they use
the internet a lot now.
There is emerging evidence--
here's some of the numbers-- eight million Americans
research a health-related topic on the internet.
More than 2/3 start their research with a search engine.
It turns out that 2/3 of internet search engines, users
trust the internet, whereas a smaller percentage trust their
own doctors.

All right, this is like a problem.
It's something we can work on together.
Usually it is especially strong for younger users.
Doctors are learning how to work with patients that are
better educated about their health.
So, for example, 2/3 now of all US physicians are using
the internet.
And they use it for prescription drug interaction,
this sort of thing.
We have a lot of studies that show this.
So what has emerged is a controversy over this question
of this new form of internet use.
Is it good or is it bad?
So TIME Magazine, Dr. Scott Haig--
you know you love these controversies--
Dr. Scott Haig says, and these are his words, "A seasoned doc
gets good at sizing what kind of patient he's got and how to
adjust his communication style accordingly." And he's talking
about his particular patient. "I knew Susan was a Googler,
queen, perhaps, of all Googlers.
But I couldn't dance with this one." Because he had so much
trouble with her aggressive knowledge and the way she
approached it.

This elicited a doctor fight with Dr. [? Perak ?]
And this is from him a month later.
Quoting from the earlier article, "'I was a unnerved
about how she brandished her information, too personal and
too rude on our first meeting,' he wrote.
He proceeded to call her the 'queen, perhaps, of all
Googlers, a class of patients he referred to as brain
suckers.'" This is a compliment.

He goes on. "So the problem with Dr. Haig's article, other
than petulance, is that he's ignoring every single internet
trend in health care over the past decade.
The medical establishment, in fact, has taken way too much
time to understand that the internet is a disruptive
innovation that has overturned the status quo.
It has leveled the playing field between expert and
novice, in this case, doctor and patient.
And while some doctors like Dr. Haig may find that
challenge threatening to their status as an expert, the web
is now providing the kind of information doctors need to be
aware of if we want to continue to be good at our job
and the kind of trends that will help the patients be
healthier and smarter."
So I think this sort of fun fight between two senior
doctors lays out the problem that Google
wants to help solve.
It is a fundamental problem.
It is certainly one of the fundamental medical problems,
and it is certainly one of the fundamental information
I was alarmed to find out how much information was being
used at Google about personal information.
We had no medical training.
I have a Ph.D. so I'm called doctor.
And I always say, well, I'm not a real doctor.
So my first few weeks at Google I show up, and we get
this letter from a fellow this says, thank you
for saving my life.
And I go, that's pretty interesting, a startup
So it turns out he was having what you all would know is
heart attack symptoms. And so he types them into Google, and
the first result says, you are having a heart attack.
Dial 911.
By the way, that's the correct answer.
We're very proud of this.
So what happens is he does 911, and they show up.
What is the drug that they give you to make sure you
don't die in the middle of a heart attack?
Whatever that drug is they gave it to him.
And they said, had you not called us
immediately you'd be dead.
So we told that story to our engineers to explain why it
was so important to have answers within 1/10 of a
second because even seconds matter in health.
And we've since received many such you saved my
life kind of letters.
And it's one thing to run a company.
It's another thing to save somebody's life.
It's pretty phenomenal.
So we got interested in this question about medical health
and health in general not knowing much about it.
And we started looking at the interesting problems in the
world where our technology could help.
We've formed a group called, and we're working
on global public health.
So, for example, we've picked prevention and cure of
blindness where we're donating money to an incredibly
important cause.
Eradication of polio, another huge, huge
and important cause.
Eradicating the guinea worm which is called the world's
most painful disease.
We're working with organizations who are, in my
view, just heroic, just every one of them a hero trying to
solve something which will compound for
the next 5,000 years.
We've also started looking at programs related to global
public health.
And information technology, which is what we do well, can
really help here.
We like the notion of predict and prevent.
So, for example, if our computer systems can be
programmed in such a way that they can detect early
outbreaks we can get ahead of these waves of information.
So early detection, preparedness, and response
systems for emerging threats, especially in the third world
where a little bit of uncorrelated data could give
us just enough information that voom, all of a sudden
there's an outbreak of this particular strange
disease in like Laos.
And we can get there before it crosses over into the much
larger, and, in many cases, more dangerous areas because
of the crowding.
So in a country like Laos where it could take almost
nine weeks for a reported case to get into the information
system, maybe we can get that information into the reporting
systems earlier and then detect them using new
mathematical techniques, pretty interesting.
We can also, using the information that we have, get
a lot more information to people who don't have it.
It turns out that there are many, many examples where
textbooks are really not available.
Virtually all of the money is going into textbooks.
Well, if these systems are online, you don't need the
textbook, or you can have them be current.

And, in many cases, we can also use Google to make sure
people know that these services, especially the free
services in Ghana, for example, that these health
services are, in fact, available to them.
People in countries which are not as developed as ours spend
an awful lot of time not knowing about systems and
services that are available just down the street because
the connections are not so great.
So if you take the model that health is important, that
information is important, and that we have both the
resources and the will to work on it, what is the underlying
architectural trend that we're working in?
And let's just talk about it in the context of the growth
of the internet.
Everybody here knows the internet is a big deal.
It is by far the fastest growing medium in history now,
more than $1.3 billion users on the order of $200 million a
year new users, the ray of underlying technology
innovation is not slowing down.
A technology-based case would offer you that Moore's law,
which is the rule that semiconductors double in the
capacity or speed every 18 months, is going to continue
for another at least 10 years or so until we hit various
photolithographic limits.
There is another bizarre law called Kryder's law that says
that memory doubles every 12 months.
So if you have CPUs going at every 18, which is, by the
way, a factor of a 100 in 10 years, and you have memory
doubling every 12 months, which is about a factor 1,000
over 10 years because of the compounding, you can see the
enormous things that can happen.
An example would be that in the year 2019, if current
trends continue, a device the size of an iPod will have 85
years of video in it, which means that you carry a device
which you cannot watch until after you're dead.

It's like the ultimate dissatisfaction device.
There's always something I'm not going to be able to watch
on this device.

The one I really got me going, by the way--
we were reviewing this yesterday--
there are 10 hours of video being uploaded into YouTube
every minute.
God knows what the quality of that video
is, but it's coming.
So you take a look at this rate, and this is going to
become much more massive than anything that we have seen.

Blogging is another one.
They're on the order of 70 million blogs and 120,000
blogs being created every day.
More than half of those are created by people who are less
than 18 years old.
As we know, if you have teenagers, they have a lot to
say, and they're saying it online.
And if you're a parent you might want to read it.
So users are going to use this technology, and they're going
to use it to say a lot about health.
They're going to have various forms of not only communities
but ways in which they help each other.
And they provide advice to each other.
Did this work?
Did this not work, and so forth.
And the notion of daily support groups, the
traditional people around the room, is now going to become
very much online.
There's something called DailyStrength which has more
than 500 support groups online.
It works really, really well.
There's something called Psych Central which has more than
600,000 users who visit their websites, libraries, and
We do things like we have bird flu, we have both reported
cases on a Google Map as well as the communities to study it
so that the scientists and the people who think they might be
victims of the same disease can all see the same
information, et cetera.
We're beginning to see people tell their stories on YouTube.
Humanity is fascinating because of our need for
A young woman named Kat created a series of 34 videos
about her battle with anorexia and got more than a million
people viewing it as she suffered through this terrible
disease with obviously a lot of support and help to help to
try to address her problems.
It's interesting, by the way, that the professionals of the
room will say, oh my god, we've got all of these crazy
people out there who have diseases now commenting it and
sharing information.
And a lot of people have studied this, and it turns out
that the vast majority of user reported health information
is, in fact, accurate, including the diagnosis which
is a surprise to me.
And the most recent study said that only about 6% of it is
By the way, 6% still means you should go to the doctor.
You shouldn't just read online and just do that.
You should talk to a professional.
But the fact of the matter is that 94% is accurate, and is
pretty impressive.
And it shows you that people do want to share accurate
information for each other.
Now, architecturally, to me this sets up the premise for
what Google is doing.
The change in power here is transformative, and it has
occurred in other industries.
And everyone else is struggling with it.
And we want to work with you to make this one be
In the entertainment and media industry--
I have lots of statistics--
42% of users 18 to 29 use the internet as their
primary news source.
I find this very disturbing.
30% of users 18 to 29 use a video sharing site such as
YouTube daily.
So if you're not in that age group, you're not seeing this
cultural shift which, if you remember when you were that
age, seemed obvious to you at the time.
To them this is obvious.
It's obvious this is how the world
is going to be organized.
And one of the consequences of this is the traditional
industries are declining.
The one that I worry the most about a CD music.
Sales of CDs have declined 19% because people either purchase
or illegally steal the information online.
These are very, very real issues for those industries.

Now when you use Google, you can do a number
of interesting things.
So one of the things that you start, is you start learning
about the history of medicine.
So I said, well, what are some interesting things to learn?
We're scanning all these books, right?
Well there are a lot of books written about
health 150 years ago.
So here's an example, The full text of The Medical Times and
Gazette, which is a British medical journal written, this
one, in December, 1858.
So it's 150 years ago.
The surgical procedure for treating conjunctivitis, which
is pink eye, he gives the patient a mixture of laxatives
and tells him to apply a dozen leached to the
eye if the pain returns.
I presume that the leaches just change the pain to a
different paradigm.
The full text for a Treatise of Military Surgery and
Hygiene in 1865 with more than 40,000 surgical operations
performed during the US Civil War, presumably all with a
large amount of alcohol and vodka, including medical
treatment for gunshot wounds, amputations, gangrene,
tetanus, and general hygiene in military hospitals.
So I have all this information, right?
I've got this transformative phenomenon.
I have all of these searches and so forth.
I need some solution to this.
What I really want, by the way, is
something very personal.
I want access to my cholesterol test. I want the
X-ray on my sprained ankle.
Why can't we solve that problem?
Now we've decided to bring sort of a
different model to it.
We're going to partner with leaders in health care to
cross-connect, to make this problem and
literally get it fixed.
And we want to apply the principles of the internet,
but we want to apply them in conjunction with the leaders
in the medical community to get the right
outcome for the end user.
So the first principle that we established was, it's the
consumer's data.
It's not anybody else's data.
It's the consumer's data.
So from our perspective we take a consumer focused view.
So in this model users can access their data and control
who gets to see it.
And the data follows the consumer wherever they go.
So if they move from one provider or one doctor it's
still with them.
They take it with them to the next doctor, institution,
insurance company, what have you.
And this is an important distinction in many of our
systems. Think about closed versus open.
Cell phones are typically closed.
If you buy a cell phone all your contacts are there.
It's very, very difficult just to switch phones.
Whereas if you think about banking ATMs when you go from
one bank to another, it doesn't sort of matter except
for transaction charges.
You can pretty much get any banking ATM to give you the
money, the same thing.
So you want a system where it sort of doesn't matter.
The system takes care of all of these complicated things.
It's really end user focused.
There are more than 200 personal health record systems
in the US, and most of them are closed.
That is they're tethered to a particular health system.
And this is a system that we see this commonly in
industries that have not yet been fully
internetized, if you will.
And it makes sense because it's not possible to have a
single standard.
And so smart people tend to build a system that solves the
problem that they see in front of them, and then someone else
duplicates some of that work.
So here's an opportunity to get these systems tied
together and get the best of breed out of everyone.
So in our case, if you take the position that 30% switch
health insurance companies each year, which is data which
is a huge surprise to me, maybe not to you all.
The benefit of consumer interoperability is extremely
There's one study that said that literally open health
care standards, which have been, I think, discussed at
the [? hymns ?] for a long time, could deliver savings of
$78 billion-- and that's billion with a B--
annually just in terms of the ability for these systems to
interoperate, let alone the health care benefits which are
very important.
So it seems to us that consumer control over the user
information will only work if there's a strong privacy and
security policy.
So in our case, our model is that the owner of the data has
control over who can see it, and trust for Google is the
most important currency on the internet.
It's easy to understand.
If you have a user-centric model, and you violate that
trust, the users will go somewhere else.
So you have to start off with the premise that the
information in your health record, or whatever you want
to call it is yours, and it doesn't get shared or given to
anyone else without your permission.
If you do so then it happens, and otherwise is won't.
Now we're in the midst of this enormous shift to what we call
cloud computing.
And cloud computing, the model here, is that rather than
having all that information stuck on my personal computer,
it's stuck on a set of servers.
We call them cloud computing or clouds because we used to
draw crowds to describe it.
And then you can pick up any computer and just access that.
And the easiest way to see that model is imagine--
everyone here basically carries a
laptop, one kind or another--
imagine what happens when you drop your laptop.
It's like a really bad day.
Sometimes it's a bad week or a year.
So what you want to have the ability to do is pick up any
laptop, and with appropriate permission, login and
password, get everything there.
So this new emergent model of server-based information, and
technically what happens is the computer
connects to the internet.
The program that you need comes down really quickly, in
less than 1/10 of a second, and there it is as if it had
always been there.
But the data is managed, I like to say, by professionals.
Because we know how to back your information up.
We make sure you never lose, and so forth and so on.
This is a core part of Google's overall strategy, but
it's particularly applicable here.
Because why don't I have my X-rays in my cloud?
After all, they're pictures of me.
Why doesn't the doctor just pump it in there.
And then when I have my next situation it'll just be there.
And it doesn't matter what viewing device or so forth?
Maybe I'm at a different country.
Maybe I've upgraded.
Maybe I've switched from a PC to a Mac or
something like that.
ERIC SCHMIDT: Sorry, sorry, sorry, sorry, sorry sorry.
I'm on the board of Apple.

Everyone here has those little yellow immunization cards.
I'm terrified I'm going to lose this card.
And I don't know what it says, by the way.
It's just scribble.
But it's really important to go in and out of the country,
and that's my job.
So why can't I just have that in my cloud, and when I get
their pop up my thing and just sort of show it.
A more serious example--
I guess these are all serious--
is in Hurricane Katrina.
A tremendous amount of health information was lost during
that terrible disaster, which again, had it been in a cloud
server it would have been kept.
So you get the idea.
Now so you it there and you go, this will never happen.
We have skeptics in the audience.
People say, well, he's a nice guy, and he's from Google.
And they're ahead of things.
And 5% of the people will adopt this stuff.
That's always true in year one.
But in year 10, it's usually 70% or 80% have adopted it.
And let me remind you that 10 years ago, when we started but
looking at electronic commerce, the studies were 80%
to 90% of the people will never use electronic commerce
because they do not trust that their credit card will be safe
on the internet.
And I'm not suggesting your credit card is safe on the
internet now, by the way.
But 80% of people now trust the internet with
their credit cards.
So as people become comfortable with these models
tens and tens of millions of people switch over.
And as they do we develop the systems that make sense.
So when you think about this model, it's particularly
applicable for something like, let me pick X-rays.
There are two billion X-rays annually, and
each X-ray is 10 megabytes.
That's 200 petabytes.
Petabytes is a very large number.
In my world it isn't.
So you could just put them all online and then we wouldn't
have to argue about this.
And then wherever you went you'd be able to have all that
information, and you would have it historically.
62 million CAT scans annually.
These are even bigger files than X-rays.
Why are they are not available to me wherever I go?
Why are they in that one building that I can't remember
where I went when I had my CAT scan?
And they probably lost it anyway because that's not
their primary focus.
Again, this is a problem that can be easily solved.
So the important thing is that any scenario where information
is sort of isolated is a scenario where health is not
well delivered.
What we want to do is we want to make sure that all that
information, however wacky, and however relevant, and
however irrelevant, is available to the professional
when they're in a situation like in the emergency room.
So, if god forbid, I was in an emergency room here in
Florida, I'd want whoever is sitting there trying to figure
how to keep me going to have access to the last N years of
my radiological experiences, and I'd like them to have it
And we can do that now.
So in order to do all of this, we organized ourselves around
a health advisory council.
I wanted to take a minute, and what I'm going to do is I'm
going to show you a video of what they had to say.
And then I want to do a demo of the system
that we're now trialing.
I think it'll give you a good sense of where we
think this is going.
This is version one.
And before I say anything else, I want to mention that
Google is not a company that designs a product then ships
it, and then just sort of waits.
We iterate, and iterate, and iterate, and
iterate, and iterate.
And we iterate on a weekly basis.
Our products are in beta test, if you know what that means,
or sort of general testing for a couple of years as we try
them here and try them there.
And we were fortunate to have an initial beta test partner,
Cleveland Clinic, to do this.
But first, in looking at the health advisory council, we
were sort of overwhelmed by 39 new pathogens have been
How do we deal with this?
Modern travel is dealing with these sort of disease and
spreading them very quickly.
Prescription drugs, they're more than 13,000 prescription
drugs on the market today, but only a few hundred are
actively prescribed.
How do we get the other ones, the information of the other
ones available?
There are 110 medical specialties in the AMA Guide.
Half of the doctors in the US work in practices with fewer
than five physicians.
So we have this explosion of information, but we have the
structure and we have the limitations that exist in the
medical community today.
How do we bridge that gap?
So we formed this health advisory council.
And I think maybe what we should do is just run the
video, and you'll see for yourself what
they have to say.
DEAN ORNISH, MD: I'm training is in internal medicine.
I'm a clinical professor of medicine at UCSF, and I'm also
the founder and president of the nonprofit Preventive
Medicine Research Institute.
MOLLY COYE, MD: I'm a physician, and I was a public
health officer in two states.
PAUL TANG, MD: I'm the vice president and chief medical
information officer at the Palo Alto Medical Foundation.
ROBERT M. WACHTER, MD: It's a disparate group of doctors
offices, and hospitals, and pharmacies that really aren't
tied together.
SHARON TERRY: People often say the health care system, but
there isn't one.
MOLLY COYE, MD: It's about time that consumers had all of
the information they need in order to really manage their
own health.
PAUL TANG, MD: One of the biggest challenges is really
to promote and support developing partnerships
between patients, their family caregivers, and their
professional health care providers.
SHARON TERRY: For example, to get my own medical records,
it's almost impossible.
To coordinate the various pieces of my medical care is
ROBERT M. WACHTER, MD: What we know from good research is
that 50,000 to 100,000 Americans die every year from
medical mistakes.
DEAN ORNISH, MD: So much of health care so fragmented.
It's in silos.
Not only the inefficiency if you have to fill out the same
form every time you go to another doctor, but there's so
little communication between the doctors.
ROBERT M. WACHTER, MD: We have this extraordinary workforce
of doctors, and pharmacists, and nurses.
But we abuse them because we handle information so poorly,
and we make their work so difficult.
SHARON TERRY: So the ideal role would be that we all have
the information that we need at our fingertips, that it's
really accessible, that it's well protected in the sense
that it's private.
PAUL TANG, MD: There's a tremendous amount of
information available online.
What we need is a way to organize it.
ROBERT M. WACHTER, MD: And so some sort of entity in the use
of information technology is necessary to try to create a
level of coordination.
DEAN ORNISH, MD: So working together with the health
advisory team at Google, we're trying to change that.
SHARON TERRY: If we accept that currently there isn't a
system for health care, and that we really need to look at
how to interconnect the various pieces, what we see is
the ability of Google to aggregate information to give
us a great user interface to use it.
PAUL TANG, MD: I think Google is in a terrific position to
be able to organize health information so that it's
useful to an individual.
ROBERT M. WACHTER, MD: Hopefully at the end of the
day the docs and the patients will actually be looking at
the same information so that we're not acting across
purposes, but we're acting as members of the same team.
MOLLY COYE, MD: That's a powerful
weapon to improve health.

ERIC SCHMIDT: So we organized this group.

We organized a set of partners both as advisers, but also
some companies that we're trying.
And I want to get their logos up so you can see them.
But the basic idea here was to go to everyone we could find
who had a lot of patient data and then work with them to
develop standards that were secure, by the way, that would
take information that these folks have in their
proprietary databases and suck it into the Google Health
infrastructure that I have been describing.
And I suspect when you look at the list you'll see almost all
of us interact with many of the firms that
are here on the chart.
So the basic idea here is that we developed a set of
which is sort of what Google does--
which are easy for these guys to connect their proprietary
data systems to, and with user permission, take that
information and put it into a user place.
And that user place, call it a personal health record, call
it what you want, can then be worked on.
The problem that we have is that without this information
we would be making the end user duplicate a lot of work.
So we need these folks as partners.
And it's in their interest because
they want better health.
They want people to have more information and more choices.
It makes good sense.
And so, for example, some of them will help with lists of
doctors because their business is insurance.
So they know what doctors they have, which
doctors offer this insurance.
Others have drug information, drug interaction information.
Other ones have just health information in general.
And obviously we want to do this as broadly
as we possibly can.
I think I've talked enough here.
And I think it's more interesting
to hear a demo anyway.
I'd like to introduce Dr. Roni Zeiger.
Roni is a Google employee who is also an emergency room
doctor and the unusual aspect that he's both a doctor as
well as a masters in information technology.
Roni, where are you?
DR. RONI ZEIGER: I'm here.
ERIC SCHMIDT: Ah, here's Roni.
There you are, Roni.
And Roni is one of the chief architects of this vision.
He has been working on this for a long time.
His first task was to try to understand how accurate or
inaccurate Google was without any help.
And he started off looking at taxonomies.
And today when you use Google, and you type in one of these
long words that are medical words, the results have been
shaped by the judgment and the algorithms that Roni and his
team invented.
So once he put that in place he decided to work on this
broader initiative.
DR. RONI ZEIGER: Thank you, Eric.
OK, the friendly login page we see here
actually is not live yet.
But I assure you that everything else you see is
real live product.
So here we.
This is the home page, the Google Health
homepage, of Diana.
She's a fictional user who is also part of the Cleveland
Clinic pilot.
Now Diana just came back from visiting
relatives out of town.
Unfortunately she came down with a bad sinusitis.
She saw a local doctor, and he prescribed for her amoxicillin
to treat her sinusitis.

Now if we drill into the details of her conditions list
we see that some of the data was entered by Diana herself,
and some of the data she imported from
the Cleveland Clinic.
Now because she explicitly gave the Cleveland Clinic
permission to also pull data from her Google Health
account, if we hop over to her Cleveland Clinic MyChart
account from Epic Systems, we see that it also now contains
her new prescription and her new diagnosis.

Now some of you probably noticed that Diana's allergic
to penicillin.
The drug interactions feature of Google Health checks for
interactions between drugs, allergies, and conditions.
ERIC SCHMIDT: I'm still confused.
How did this happen?
Shouldn't the doctor have figured that out?
DR. RONI ZEIGER: So I'll admit, Eric, that when I see
patients I do sometimes forget to ask about allergies.
In this case, Diana herself may have forgotten about her
penicillin allergy.
Fortunately her Cleveland Clinic doctors do know about
her penicillin allergy.
And we just saw that they're also now aware that she was
prescribed amoxicillin.
Diana herself now has a safety check available to her that
reminds her to talk to her doctor about this.
Now another very cool feature about Google Health is
something that we call Google Health Reference Pages.
So the user studies that we've done so far have taught us
that consumers really want some basic context, especially
about conditions that they may not know much about or that
they may be wondering if they have. We include here also
some informative, if sometimes a bit spooky, illustrations as
well as relevant and dynamically generated news,
web search results, research articles from Google Scholar,
and pointers to discussion groups.

Now the last thing I want to share with you is what I find
most exciting about what we're doing in Google Health.
Diana can choose to connect her Google Health account to
any of a growing number of third party services that have
integrated with Google Health using our
soon-to-be-published APIs.

And what we're seeing here is a live application that, with
the user's explicit permission, has pulled their
data from their medication lists in Google Health and can
display it in a variety of interesting ways that Google
Health does not.
ERIC SCHMIDT: So who wrote this app?
Did we write this app?
DR. RONI ZEIGER: No, this is written a
company called Solventus.
ERIC SCHMIDT: Did we know this?
DR. RONI ZEIGER: We gave them access to our APIs.
ERIC SCHMIDT: OK, did we pay them?
DR. RONI ZEIGER: We did not pay them.
Should they pay us?
DR. RONI ZEIGER: I don't think they should pay us either.
ERIC SCHMIDT: Sorry, just checking.

So it's like they can do what they want and just connect
into our system.
DR. RONI ZEIGER: That's correct.
And Diana can choose to work with them if she wants to.
So I might want to print this out, this weekly view, and put
it by my medicine cabinet.

Another developer created a medication reminder gadget
that I can put on my personalized iGoogle homepage.

And I cannot wait to see the amazing and innovative tools
other developers can create for the Google Health users.
ERIC SCHMIDT: For benefit of the audience, do you have some
tools or ideas?
I think drug interactions is an obvious one.
But, as a doctor, there must be like 500 other categories,
if you had all that information, that
you could go over.
What are some others that you think would be most powerful?
DR. RONI ZEIGER: Well I think that from my own experience,
and more importantly from hearing from our users and the
experts that we're working with, I think some of the
things that would be really neat is if I could enter the
immunizations of my children and to get a useful dashboard
of everything that they've had done, what they need to do
next, and when.
If I could get customized feeds of news and research
articles that are targeted to my conditions and my
ERIC SCHMIDT: Wouldn't it be nice, for example, if there
was a corpus, and then they said, the disease you have has
been cured.
Call here.
DR. RONI ZEIGER: I would want to know immediately.
ERIC SCHMIDT: Right, speed matters.
Well, thank you very much, Roni.
DR. RONI ZEIGER: My pleasure.
ERIC SCHMIDT: I wanted you to see it because we did this
partnership with Cleveland Clinic.
Cleveland Clinic is a really neat group.
They're very, very large.

They're very large in general.
They have a lot of different sites.
They have more than 100,000 people inside their medical
health system.
And so they were willing to work with us to help define
this standard which is, of course,
non-exclusive to anyone.
But hopefully this will show you the benefit.
And we'll see how well this goes over the
next month or two.
And as we broaden this, we hope to broaden this to
essentially everyone that's possible in the United States.
The technology that's used is extremely simple from our
It's an internal interface which allows you to move data,
and we use it for a lot of our other applications.
It's a security model and so forth, so it's all
So we've managed to layer this on top of just Google.
And that's why this will move so quickly and be so exciting.
Cleveland Clinic today has more than 1,370
people in this trial.
And over the next few weeks we'll find out do we really
make a difference in their health?
What is missing?
What are the next key apps?
And one of our messages to you are, if you have an
opportunity to build an application
on top of this platform--
because remember, this is not just a personal health system.
It's really a platform for interacting on a user's data
with their permission.
If you've got an idea that can really change the world in
medicine, we want you to build it on top of this platform.
So with that, thank you very, very much for your time.
MALE SPEAKER: To moderate our questions and answers, let's
bring out our president and CEO, Mr. Steve Lieber.
STEVE LIEBER: Good morning.
Good morning.
Well, Eric, I think you showed us and told us what we've been
waiting to hear this week.
You created a lot of buzz with the announcement last week,
and we certainly wanted to see and hear what you had to show
and tell us today.
We're going to turn the lights up.
And we've got some a little bit of time
left for some questions.
And while we've got people coming up to the mic, let me
ask the first one.
You talked about consumer adoption of internet
technologies, the uptake, and all that, low percentage first
year, picks up after that.
Have you got any predictions in terms of what that cycle is
going to be?
ERIC SCHMIDT: With respect to these tools, I think it will
completely be determined by two things, the ease of use of
the interface and the services that we can provide.
Ease of use turns out to be one of the most important
things in one of these internet services.
It's true not just for health.
If people get confused, if you start asking them the wrong
questions, they quickly become sort of tired
and they move on.
So we worked hard with this user interface to be able to
capture health information, with people's permission,
very, very quickly.
The moment you do that you have to immediately show them
something that's useful like an oh wow moment, like, oh my
god, I I'm healthier, or I got this piece of information, or
I'm sick, or something.
And then once they have that experience they'll come back
and come back.
So my guess would be that this will grow very quickly to at
least early adopters during the first year.
It's hard to know after the earlier
adopter phase how quickly.
But our goal is everyone or at least everyone who wants this
kind of information.
OK, let's start over here, Joyce.

AUDIENCE: It was a great talk and very interesting to see.
I'm an analyst in this space.
And one of my colleagues was hired to look
at the payment space.

We were expecting to find that Google Payments was really
well received and incredibly successful.
And our sense was that in the payment space it hasn't been
that successful and that there hasn't been a sufficient
commitment to that space after the product launch.
How do we know how committed Google is to this space?
And how do we get a sense to judge where
we'll be in a year?
ERIC SCHMIDT: Well I disagree with your question about
Google Payments.
Google Payments is a product that was designed to make it
quicker for advertisers to get their money.
And on that metric it works extremely well.
And we certainly put a lot of money and focus on it.
We have a lot of partners.
This is an end user product.
Payments was not.
Payments was really an infrastructure part.
So the question here about Google, a consumer product
will be whether consumers like it.
And successful consumer products take
off extremely quickly.
So you'll know very quickly.
But if we make a mistake or if we don't get the UI right then
it'll take longer.
But I can assure you we've been working on this for a
couple of years.
So many of our queries are health-related that we must be
successful here either with this approach or modification
of this approach as we learn what works and what does not.
STEVE LIEBER: Great, thanks.
Yes, my name's Corey Ziegler.
I'm in a small rural hospital in Northern New York.
And as I listen to this, first of all, I applaud you for
pushing standards.
Because those are the issues that are really kind of at the
core of what we're trying to link all these separate
systems together.
But in speaking with our physicians, when we present
them with the data from another physician or outside
of something they're familiar with, they have some
reluctance to trust that data.
And there's some compliance and some risk management
issues for the facilities in trusting that data.
So if they use it for clinical decision making I'm concerned
about the liability.
Do you have any comments on that?
Have you guys discussed that?
ERIC SCHMIDT: We have, and we have found that more
information usually solves this problem.
Of course there are people who say, well, that's not my
information so I'm going to ignore it.
Such groups have existed for thousands of years in our
history, and they've often done really bad things.
It's really much better to be aware of the information that
other people have even if you don't use it
as your final diagnosis.
So even in a situation where you are unwilling or, by
regulation, you're unable to use the information that
Google Health has in it, it has to be helpful.
It must fundamentally at least inform your decision as a high
quality caregiver.
We're not trying to change the way doctors work.
We think that the doctor profession is obviously a
very, very important and very well thought of.
And doctors need to figure out how to use this information to
achieve their objective which is greater patient health.
We're providing information.
They'll sort it out.
STEVE LIEBER: OK, over here on this side.
AUDIENCE: One of the things you spoke about was the
building of trust for people to use these kind of systems.
And I think a lot of people will probably feel OK having
their information in Google's cloud knowing that you're
probably real good at keeping things safe.
But a lot of people I've spoken to are also wondering
what's in it for Google to have this information.
Talking to one of your engineers at the booth, the
phrase he used is monetization path.
And is there a monetization path for this information that
you're going to be holding for people?
ERIC SCHMIDT: Not in the short-term.
We're making a commitment that the data itself will never be
shared with anyone without the end user permission.
One of the things that we've learned is that if we have a
powerful, so-called vertical site that does something
really, really neat, that person is more likely to use
Google in its traditional ways and therefore
click on our ads.
We have a great success story there with Google News.
You can make a lot of money with ads and other services on
Google News, but we've decided not to because we know that
people who use Google News are more likely to
do more Google searches.
So we believe that if you, as a citizen, as a customer, if
you will, of Google Health, we believe--
and there's lot of anecdotal evidence--
that you will be using Google for many other things,
ultimately click on ads, and it's a net positive.
And we'll measure that.
And I suspect it will be true.
OK, back over here on the right.
AUDIENCE: My name is Sreedhar Potarazu.
I'm the CEO for a company called VitalSpring.
Interestingly enough, about eight years ago, I was a
physician at Johns Hopkins.
I practiced there for several years.
And my MBA thesis eight years ago was about how we can build
community-based networks building
communities around that.
Eight years ago people thought the model would fail.
Over the last eight years we have 40 of the Fortune 500
companies now that we've been doing business with where
we're building, essentially, the next Facebook model of
health care.
And essentially, as you said earlier, there's an open model
and a close model.
Employers are the ones still paying for health care today.
And I can tell you speaking on behalf of 40 of the nation's
largest employers, they have a big problem in terms of
empowering consumers to get information outside of a
closed model where advertising is not allowed.
And now we're building, essentially, the next social
network amongst consumers' employees to get all of this
information in that closed model
because this is not banking.
This isn't buying consumer goods.
It's financed by people, and yet we're giving them the
opportunity to find the information outside, and
they're not paying for it.
How do we solve that?
ERIC SCHMIDT: [? Dossier ?]
is actually one of the partners.
I don't know if you saw them on the board.
Our general answer is that open standards will allow
people to take these proprietary information data,
put it into the end user, and then that creates
a balance of power.
The end user then has a choice of moving.
The doctor has a choice of using that information or not.
In my experience, looking as a scientist, in most cases
people are not very empowered in the system.
And this is a step to give them more power.
STEVE LIEBER: Great questions, but we're running out of time.
I've only got time for one more, sorry.
AUDIENCE: All right, thank you very much.
I'm Paul [? Schadler. ?]
I am a practising physician in Denver, Colorado.
And this is exciting, exciting stuff.
I do have an occasional pang of paranoia about Big Brother
Google knowing everything about me and everybody else.
But as a physician I--
It's just me.
AUDIENCE: It's just a little ole boy from the country.
Nothing to worry about here.
ERIC SCHMIDT: Yeah, I grew up in rural Virginia.

AUDIENCE: But as a physician, I and my colleagues, waste so
much time collecting data that's been collected in the
past, and reviewing data that's been reviewed in the
past, and going over data that's been gone over.
And the oh wow moment that I perceive is when Sarah comes
in my office--
patients switch doctors all the time now--
Sarah comes in.
Instead of handing me the form we say, do you
have the Google password.
Well, yes.
Can we access it?
We access it, boom.
Does everything look here good?
Her medical information comes in.
Her payment information may come in.
Her insurance card number, it all downloads.
Sarah, the doctor will see you right now because he's not
wasting his or her time reviewing the data.
It's all in our system.
He'll be with you in just a moment.
That's an oh wow moment and that will improve health care.
Thank you.
STEVE LIEBER: Thank you.
Thank you very much.

Was there a question in that?
AUDIENCE: There was a question.
I forgot the question I was so excited.

So then the question was--
and your optimism led me ask this--
oh that will be simple.
50,000 gargantuabytes, no big deal.
So while Google is doing all this easy stuff that you're
about to do, which a lot of people have struggled with,
are you going to go ahead and just easily create a little
interface so that these doctors, the 80% of the small
practices can just use Google EMR to put data into the
system and have Google EMR create an SCD-9 code and send
a bill to the insurance company?
ERIC SCHMIDT: What I was going to suggest is that's a classic
example of a third party app.
And we've got a number of entrepreneurs here in the room
who could see.
Let's take your idea because it's a really good one.
You've got a situation where you've got a patient and
you've got their medical history and you have their
business relationships.
And with their permission you could imagine a whole bunch of
applications which did exactly the kind of thing you're
talking about.
We're unlikely to do it because we don't really
understand that part of the business.
But our system is designed as a platform.
And this is exactly what we're looking for.
So I hope you found that company and make yourself a
lot of money.
STEVE LIEBER: Thank thank you very much.
Eric, you have certainly opened our eyes this morning
to a whole new world of consumer-facing health care.
And we certainly appreciate you joining us.
Please join me again in thanking Eric Schmidt.