Google Health — A view from the Inside

Google Health launched last Monday, which sent the world’s Google-watchers into a tizzy. I serve on Google Health’s Advisory Council – which met all day Tuesday – and so here’s a bit of inside dish, along with my impressions of the site and the company.Google Health pageFYI,
my work on the Council is covered by a Non-Disclosure Agreement, so I
won’t reveal anything that isn’t publicly known regarding Google’s
products or intentions. Also, in the interest of full disclosure, you
should know that I am compensated for my Google service. (No stock options, darn it.) With that as background, here’s the scoop. Google
began working on its version of the personal health record a couple of
years ago, after the company realized that a remarkably high percentage
of searches were for health information (I know, if that’s going to be
how priorities are set, you’re wondering if Google Sex is next). Google
put together an impressive team to develop the concept. One of the
leads is a former UCSF medicine resident, Dr. Roni Zeiger. Many of our
residents come to me for career advice, and I seem to recall Roni
asking me about pursuing his interests in informatics after residency
about a decade ago. Sage that I am, I probably told him that this
computer thing didn’t have legs. Luckily he didn’t listen, and now he’s
the top doc at Google. Go figure.

Another lead player in Google
Health is Missy Krasner, an effervescent woman who served as David
Brailer’s senior advisor when he was the first federal IT czar. Missy,
who sports an ever-present smile, serves as the glue that holds the
team together, and her knowledge of all the national IT players
(including those in DC) is crucial.

Google Health’s Advisory Council is chaired by Dean Ornish, the famous physician-author, and includes a bunch of luminaries, including Cleveland Clinic CEO Toby Cosgrove, Harvard’s CIO John Halamka, and Dan Crippen,
former head of the Congressional Budget Office and now a top McCain
advisor. Yesterday’s meeting was held in a large conference room in the
in Mountain View, CA; a huge Google poster was on the wall, with the
first “o” in Google morphed into a cartoon of Einstein’s face, and the
notation “=mc2” following the “e” in Google.

Google Health
began with a vision of aggregating personal health information in one
place. It rapidly became clear that asking a 75-year-old to type in 13
meds was a non-starter, so the team began to focus on establishing
links that would allow data-sharing between Google Health and large
healthcare entities that control key data. Ergo, Google Health would be
an aggregator of data, a trusted and portable repository of health
information. You can only imagine the challenges of establishing these
affiliation agreements, particularly working through each
organization’s privacy and proprietary concerns. Impressively, 14 partners
were signed up by the launch, including two large provider systems (the
Cleveland Clinic and Beth Israel-Deaconess), several pharmacy systems
(Walgreens, Longs, CVS), a pharma benefits company (Medco), and the
company that performs about half the lab tests in the U.S. (Quest
Diagnostics). Reps of all these companies were there for the launch and
Tuesday’s meeting.

Why did all of these companies want to play
(after all, they already host websites with significant functionality).
There were probably a number of business reasons, but I think that the
Walgreens representative captured the true motivation when she said,
“Hey, I work for Walgreens. This is Google.” One can’t put a monetary
value on the “cool” value of teaming up with Google.

What about the Google Health site
itself? I’d recommend you play around with it. The site looks “Googly”
– the lines are clean and attractive. I found it easy to enter diseases
and meds, and to check for drug interactions. There are already a bunch
of neat tools linked through “Personal Health Services.”

the other hand, the Find-a-Doctor function is a start, but needs work.
It should provide more useful credentialing and quality data – right
now the map of the doctor’s office is its most prominent feature, which
is fine if you’re searching for a Starbucks but not if you’re searching
for a cardiologist. When I look for an ENT doc in Philadelphia, at the
minimum I’d want to see board certification, med school, residency, and
fellowship. And ultimately, I’ll want to see quality data, volume by
diagnosis, publications, and patient and perhaps even peer ratings.
Ditto for hospitals.

The partner interface thing works well – I
found it easy to link to one of the partnering sites (in my case,
Walgreens), and, during a demo at the Council meeting, we were able to
import one of our member’s medical records from the BI Deaconess site
into his Google Health record. “Never forget this day,” said one of the
Council members, with forgivable hyperbole. “You’ve just seen a
demonstration of portability.” Is that how Thomas Edison felt?

Council meetings over the past two years, we really hammered away on
the privacy theme. Unless people feel 100% confident that their medical
records won’t be leaked, shared, or sold, they won’t feel comfortable
jumping in the Personal Health Record pool. Of course, the majority of
the reviews of the site (in traditional media and the blogosphere – a few of the best are here, here, and here)
scrutinized this issue with a fine tooth comb, and most came away
reasonably impressed. It seems to me that Google has done a nice job addressing the concerns about leaked personal data.

notice that there is no advertising on the site. “How are they going to
make money?” my mother asked me, cutting to the chase. I remember
sitting next to Google CEO Eric Schmidt at a prior Council meeting, and
he said simply, “Getting into health is the right thing for Google to
do.” And I believed him. Of course, you can afford to be a little
altruistic when your market cap is $100B. Marissa Mayer,
the tech genius and Silicon Valley rockstar with overall responsibility
for the Google Health project (she has the Googly title of “VP of
Search and User Experience”) made the point again yesterday. “Not
everything we do has to make money,” she said. “Google is an ecosystem”
– a term used repeatedly around Google – “and we’re trying to grow a
loyal, devoted user base.”

Where is Google Health going?
American healthcare is a series of disconnected and dysfunctional
silos. Unless you are lucky enough to be in a closed system (Kaiser
Permante or the VA, for example), your best hope for coordinated care
is if you or your family serve as your own orchestra conductor.
Although several companies are jumping into the mix (Microsoft, Revolution Health),
Google is probably best positioned to do it right. It has both the
trust and the eyeballs of the public, enjoys strong partner
relationships, and boasts a universe of developers itching to build
applications that add functions. As one example of the latter point,
when iGoogle
launched (which lets you customize your main Google Search page), there
were about 10 third party applications (those cute little widgets like
calculators and Letterman’s Top Ten list) available. A couple of years
later, the number is about 75,000.

From my two years working
with Google on this and another project (more on that one in a future
post), I have come to appreciate the company’s unique culture and
lingo. I’ll close with a few observations, along with reflections on
how they may impact Google Health over time:

  1. Google people are wickedly smart. And they’re nice. It is a bit scary.
  2. Googlers
    are also really happy (or so it seems), despite the fact that they work
    their tails off. People say things like “we think we’re changing the
    world” and truly believe it.
  3. Why wouldn’t they be happy?
    They work for a great company, with smart and interesting people, in a
    beautiful place with great weather. And they make a lot of money.
    Anybody who isn’t happy should have his or her head examined.
  4. Part
    of the way Google gets these young folks (you never see anybody over 50
    on campus, and most look to be about 25-35, even those in leadership
    roles) to work their tails off is by making the Googleplex into the
    nicest college campus you’ve ever seen. “How do I get to building 43?”
    I remember asking on my first visit. “Oh, walk between the hot tub and
    the sushi bar, take a right after the big dinosaur, and it’s right
    there.” No kidding. There is no reason to ever leave, and it seems that
    many people hardly ever do.
  5. The food is really terrific. And yes, it is
    free for staff and visitors. On Tuesday, the choices included Croque
    Monsieur, caramelized scallops in brown butter, and freshly made
    California Rolls. But before you get too jealous, realize that it isn’t
    all Utopia – the sushi was limited to six pieces per customer.
  6. How
    about information technology? Yes, they do that too. Here, Google’s
    obsession is on the user experience. As in Marissa’s answer re: making
    money off Google Health, their corporate philosophy hinges on making
    Google the destination for all your IT needs (including some you didn’t
    know you had until Google filled them) – they know they can “monetize”
    your eyeballs (read: display ads) eventually without having to nickel
    and dime you at every turn.
  7. An incredible culture of
    restraint and discipline is built into their corporate DNA. It is
    manifest in all kinds of ways. There are no fancy cars in the parking
    lot (I’m told that a few years ago, one of Google’s top brass
    threatened to take a chainsaw to any showy car he saw in the lot). I
    assume that many people have Porsches at home, but it is not a “flout
    your wealth” kind of place. I saw more bikes than Beamers.
  8. In
    addition, the iconic uncluttered look of Google pages (beginning with
    the famous home page but extending to everything they do) is a miracle.
    When you sit around the table brainstorming with Googlers or their
    advisors, everybody has a great idea. The trick is in the
    filtering – the recognition that the cost of clutter is simply too
    high. That takes extraordinary discipline.
  9. This philosophy also
    plays out in their launch strategy for new products. When they launch a
    new site, they want it to be good and stable, but they don’t obsess
    about making it perfect. (After all, it’s not like they have to reprint
    a book if they find a typo.) Instead, the mantra is “iterate, iterate,
    iterate.” Get it out there, “mine” user behavior religiously (they
    follow every click for signs that Search needs to be improved, for
    example), and continuously strive to make it better. It is a winning
    strategy. Obviously. And people now have come to intuitively appreciate
    this philosophy sufficiently to give Google the benefit of the doubt.
    So, even if you visit Google Health and say, darn, I wish it did X or
    Y, there is a pretty good chance you’ll come back to the site over
    time, because your past experience tells you that your wish list might
    just be filled by your next visit.
  10. The “Don’t be Evil”
    (Google’s corporate motto) thing is real (or, if not, they’ve fooled
    me). Here’s one small window into it, told to me by one of the key
    leaders in Google Search. You’d be shocked to know how often the Google
    search algorithm is tweaked to make it a teeny bit better (I can’t give
    you the precise number, but trust me, it is more often than you think).
    “I used to work at another company that did Search,” this Googler told
    me. “Before I ever touched the search algorithm, I needed to get the
    approval of the business types, since I might be changing their ad
    revenue formulas. But at Google, the message to me has always been,
    ‘just make Search into the best user experience possible. We’ll deal
    with the business consequences.’” This is remarkable, since a tiny
    recalibration of the search algorithm might cost the company millions
    of dollars in ad revenue (since ad charges were based on the old
    algorithm). “In years of working here, I’ve never heard a single person
    complain about this.” That is one amazing culture.
  11. Google is
    all about scalability, another Silicon Valley buzzword. Google Health,
    like many Google apps, is really a platform on which partners and third
    party developers can work their magic. So, although the functionality
    of GH is fairly limited today, come back in a few months and I’m
    guessing you’ll see all kinds of medical calculators, reminder systems,
    data sharing links, tailored searches, slick ways to find a doctor,
    connections to the literature, advice, maybe even user groups. Google
    will have developed only a small part of this – most of it will be
    third party products building on Google’s scaffolding.
  12. One key
    question for Google Health going forward will be how to parse all the
    great ideas that are likely to spring up. Here are a few: What if there
    is another Vioxx recall – should Google create a way to notify 100,000
    people on Google Health who are on a certain med? How about 1,000
    people who have a recalled pacemaker? Yes? OK, then, how about when a New England Journal
    blockbuster article comes out demonstrating a much better way to manage
    Type 2 diabetes – do they let all diabetics know? And how much should
    the data be “mined” looking for patterns? One way to identify an
    emerging epidemic or bioterrorist attack might be to notice a surge in
    searches for “shortness of breath” or “skin rash” from a given locale.
    Should Google’s mega-computers be sniffing these things out? And should
    researchers have access to Google Health data (stripped of personal
    identifiers) – what better way to find out the average creatinine of
    community-dwelling octogenarians with diabetes and hypertension? All of
    these issues – and more – will inevitably arise. In general, it
    appeared to me that Google’s philosophy on these “wouldn’t it be great
    if…” ideas is to go slow – and to never sacrifice the user’s experience
    and privacy.

Ultimately, patients need to be able to
control their own health records, and to have them reliably available
when they move from hospital to SNF, from primary care doc to
specialist, or from car crash to an ER far from home. At this moment, I
see Google Health as the best hope for accomplishing this in the near
future, and as a remarkable opportunity to blend the scrapbook aspects
of a personal health record with all kinds of functions that might well
lead to better health and healthcare. Congratulations to the Google
team for a great start.

Now go iterate…

Categories: Uncategorized

Tagged as: ,

14 replies »

  1. looks like proof that “nothing’ can be written about forever and ever and ever. go back to work or your discipline at least.

  2. I love the fire that’s been set to the healthcare IT industry. I think the days of just the big names in healthcare making all the rules is about over.
    The trend toward open source and free healthcare information services is going to force the big OEM’s to rethink their offerings. I’m hoping this will open the floor to better tools and customization of healthcare IT systems as I’m tired of seeing hospitals and imaging centers rework their workflow to meet that of the system rather than the system conforming to the workflow.
    True, it’s not quite there yet, but it will be.

  3. Am concerned that it doesn’t solve the problems. Yes it is cool to have digital importation of medical records and the site has two players involved… but it costs $10,000 for a digital link to a hospital EMR. It’s just too expensive now and I don’t see buy in for that reason. Teh patient will remain the center of the universe and guide the behavior as most docs still are using paper. I think we still have a way to go. BTW I like the interface of Worldmedcard.com better.

  4. Insightful article Bob and you nailed the Google Health culture perfectly. I believe that their ability to create the “community space” for the conversation and development to occur will be the real tipping point.
    Personally I frame the status of the current health information system as a public health emergency. If we already had the system in place and lost it and suddenly tens of thousands of people started dying unnecessarily or prematurely, administrative costs shot up 20%, redundant tests were being ordered, people had to wait days for calls from their doctors, and we had to go back to paper we would bring the resources to bear to implement a high quality, efficient, safe health care information system.
    We still need EMR”s in the 70% of provider offices that still rely on paper to feed into the PHR’s since none of us (including those of us who work in the industry) are going to hand enter our data into a PHR.
    It is still a challenge to integrate systems within a hospital and currently large vendors like Epic don’t even allow clients with the same vendor in the same city to share data. For example – Kaiser, Palo Alto Medical Group and Stanford can’t share patient information in Google’s own backyard nor can Group Health, Swedish and the University of Washington in Microsoft’s backyard. Rather then trying to build interface engines from each point to point it makes far more sense to deposit it in my Google health record banking account.
    Google Health will hopefully allow us to create a “hastily formed network” around this issue and has created the conditions that are necessary but not sufficient to solve the problem of information aggregation by building the organizational structure we need.
    For those of us out in the trenches implementing EMR’s at large health care systems and struggling to empower and engage consumers in complex systems this is one of the most exciting moves we have seen.
    In the banking system we have all become accustomed to a multi-tiered approach. Online bank records are similar to EMR’s, we can see real time data but we can only make deposits and withdrawals in a limited manner. When we download the data from our banks and brokerage accounts into applications like quicken that is the equivalent to a PHR. Most of our transactions occur via a debit card (so hospitals and emergency systems are now using smart cards).
    The key to the banking and credit system is a shared set of standards and information sharing agreements that are transparent to most of us. Google Health’s API and privacy agreements combined with the trust of their users will be the real transformative power, but until we get greater penetration of EMR’s in private practices it will be a challenge for consumers to aggregate their data. If we can pull in the state vaccination data and physician credentialing data along with the drug benefit firms data we would really provide some value to harried young paretns.
    BTW – Those of us volunteering on AHIC2 also had great lunches with Dr Robert Kolodner at Brooking’s discussing Chaordic Organizations, but I agree that dinner with Missy at the GPlex is more enjoyable. The one downside is that my face hurt from smiling back at her so much.

  5. Very helpful insight into the developmental process.
    Pioneering work by the Mayo brothers and others created a comprehensive and systematically documented record for each patient. The approach has proven to be a differentiating factor for integrated delivery networks in providing higher quality care with better outcomes.
    Now, some of the smartest people in healthcare are advising Google on how to move data out of the hands of provider organizations onto the www. I am not aware of any other industry doing this to the degree contemplated here.
    And who or what will provide the organization, discipline and interpretation of the data to achieve the improvement in patient care envisioned by the Mayo brothers?

  6. Well written, my only concern with GoogleHealth is no advertising, yet…
    Applicants: GOOGLE INC. [US/US]; 1600 Amphitheatre Parkway, Mountain View, CA 94043 (US) (All Except US).
    SACHS, Eric [US/US]; 8227 Riconada Court, Newark, CA 94560 (US) (US Only).
    Inventor: SACHS, Eric [US/US]; 8227 Riconada Court, Newark, CA 94560 (US).
    Agent: YAO, Shun; Park, Vaughan & Fleming LLP, 2820 Fifth Street, Davis, California 95618-7759 (US).
    Priority Data: 11/443,818 30.05.2006 US
    One embodiment of the present invention provides a system that serves advertisements within an electronic medical record (EMR) system. During operation, the system receives a request from a medical practitioner to access a patient-record from the EMR system. Next, the system looks up the patient-record in the EMR system and obtains one or more advertisements based information associated with the patient-record. The system then displays the one or more advertisements along with the patient-record to the medical practitioner.
    More at http://www.wipo.int/pctdb/en/wo.jsp?wo=2007145860&IA=WO2007145860&DISPLAY=STATUS

  7. My mind is not quite made up on what I think of Google Health, although this article has provided some tangible information (despite the adulation) that helps me to see its potential utility. I do have an answer for one question; however — “Why is Google doing this?” They are doing it because we are probably going to have a democratic president next, with a plan to install EMRs and high hopes that these will be a major health-care cost-saver. Google wants to be that EMR provider, and they are probably way ahead of the curve.

  8. “why do the Google Health terms of service go out of their way to insulate the company and its partners from all but trivial liability..”
    That’s the problem with a large central electronic data base, even a small breach could bankrupt the company. Maybe we could get Congress to shield Goggle from lawyers like they shield docs, heck let’s shield everyone from accountability.

  9. Fascinating post, Bob. I went in figuring you’d probably drunk the Kool-Aid as a result of your closeness to the process, and came out rethinking some of my own assumptions about the service.
    That said, one major issue still nags at me. If Google is as committed to patient privacy as folks like you and John Halamka insist, why do the Google Health terms of service go out of their way to insulate the company and its partners from all but trivial liability should legal issues (presumably including any resulting from privacy breaches) arise? I’ve gone into more detail over at BNET Healthcare, but legal clauses that limit Google’s liability — and that of its partners — to no more than $1,000 and which compel the user to “defend or settle” any third-party suit related to use of Google Health — don’t exactly inspire confidence in Google’s commitment on this front, at least to those of us who’ve never set foot on the ‘Plex.

  10. Great post Bob! It’s certainly time to digitize our medical records. With Google’s weight behind this initiative, success will be achieved. With HSA accounts globaliztion of medicine we will have options as never seen before. However, we must have a safe, secure, and seamless tool to have all our medical records at our fingertips. The vast majority of people don’t even know their blood types. We send patients around the world for various medical procedures, and having an online tool for accessing medical records will be tremendously helpful to all. Thanks for the info!

  11. That was the most informative, interestig bit of Google Info I’ve read so far. I’m tech savy, blog, invest heavily in IT in our clinic and I’m still not bitten by the Health 2.0 bug or PHR. The knowledge that Google is doing this to be the one-stop shop for IT needs (rather than as a platform to market health providers & products) gives me the warm and fuzzys rather than the Orwellian chill I had from the Google Health Launch press junket. Michael B summed it up well — insight into their motivation. That has been a big question on most bloggers minds – and you’ve answered it well.
    Given the cost of health care right now I’m not convinced that the IT infrastructure in clinics and hospitals is worth it (becoming Google compatible may be expensive) but if Google is willing to provide the service at least I have the option. Where do I send my EMR provider to tell them how to config the software?

  12. Thanks for the great post Bob. Particularly liked the inside view on how Google operates.
    Between Google and Microsoft, the visibility of PHRs has increased dramatically. Hopefully, consumer expectations have not risen too much yet as both of these solutions are new and honestly, there are simply not enough features in either o them to make them attractive long-term. But I do have confidence that if they can not do it, pretty much no one else can – the problem is just too complex.
    One last point, surprised by you referencing Revolution Health, which despite all the marketing, still has very little to show for it. Why not mention WebMD or for that matter Dossia, which is very similar to what Google and Microsoft are doing?
    And if I may be so indulgent, did my own review of Google Health over at http://www.ChilmarkResearch.com and did like Google Health, though in conversations since that initial post, have come across some issues with the platform that certainly demonstrate very limited utility.

  13. Great article…with some really compelling points regarding Google’s motivations and capabilities. I would really like to hear some discussion about how providers will be incented to participate.