Google Health (GH) could be the event of the decade in advancing health care reform — not just healthcare information technology (HIT) reform, but health care system reform. GH promises simultaneously to create AND dominate the market for next generation personal health records (PHRs). There is nothing else in our solar system or in the entire universe like it.
While Google has not “officially” announced the details of GH, they’ve left a lot of clues. In this essay I’m going to play a combination of detective and tea leaf reader — deciphering existing clues and making some educated guesses about where GH is headed….it’s very exciting.
I get truly inspired when I see a business model that 1) is exceptionally powerful, and 2) offers dramatic benefits to humanity. Google Health is just such an animal.
Let me be clear. I am not speaking for Google; I don’t work for Google. This posting is my collection of inferences drawn from the clues provided about GH and from my general knowledge of health care and business. So when I write “GH will ______”, recognize that this is me putting together clues and applying my best judgment.
Here’s an overview:
- The Current Market Structure for Personal Health Information (PHI)
- GH’s Anticipated Technology Model
- Three Leverage Points
- GH Will Be Disruptive
- Concluding Thoughts
One last note before we dive in — this is a long blog posting and some of it will require careful attention to grasp the full picture.
I. The Current Market Structure for Personal Health Information (PHI)
The situation is sad, but not difficult to describe. Your health care and wellness information is 1) scattered everywhere, and 2) not in standardized formats suitable for a global information economy.
First, PHI is scattered. From a patient’s POV: “Information about my health and health care is scattered everywhere. Most of it is on paper, while some might be electronically stored. This information exists in the records of doctors, hospitals, pharmacies, labs, imaging centers, therapy centers, counselors, etc. Some of it is still in my head.”
The physician’s perspective is a mirror image of the view of an individual patient: “Information about all my patients’ health and health care is scattered everywhere….
Here’s how Adam Bosworth, VP, Google and head of GH, described the problem:
There is no place individuals can go to get a comprehensive set of health and medical information about themselves. Access to this comprehensive information can be vital to proper and timely diagnosis of the patient, to the patient getting the best possible treatment, and, perhaps sometimes overlooked, to the patient getting the best possible ongoing care and support after the initial treatment, especially for chronic illnesses. [Bosworth; December 7, 2006]
Second, not only is information scattered, but standards for defining and sharing the data are still evolving; where standards exist, many of them predate the Internet. Standards about how to define consistently the information (clinical standards) and to transmit and exchange the information (technical standards) are not yet formalized and agreed upon:
Despite all these initiatives [CCHIT, HL7, IHE, ASTM, EHRVA, HITSP ] there exists no overall plan for how standards should be defined and the overall principles that should guide their design to ensure an optimal societal use of ICT [information and communication technologies]. Equally important, no one has identified all the standards needed to support interoperability of health care ICT. [National Bureau of Economic Research, May 2007]
II. GH’s Anticipated Technology Model
We’ve been provided a number of clues about the technology model that GH is likely to develop:
- Patient centric
- A personal health URL
- Automated data mechanisms to gather and store PHI
- Interoperable technical standards: XML and the Continuity of Care Record (CCR) standard
- A user interface
- Appropriate security and confidentiality measures
- Value added functionality (over time)
Let’s examine the clues one at a time.
a) Patient centric
Let’s put the patients in charge of their health and medical information. Let’s build a system which puts the people who are sick in control. For every single medical and health-related event, let’s make sure that patients can effortlessly retrieve and share their information in its totality and then use it to ensure that they get the best quality of care possible. It is their health. [Bosworth; December 7, 2006]
This vision for the future of health care starts with the premise that consumers should own their own total personal health and wellness data (I’ll call it PHW for short) and that only consumers, not insurers, not government, not employers, and not even doctors, but only consumers, should have complete control over how it is used. [Bosworth; May 22, 2007]
b) A personal health URL
The personal health URL and the supporting infrastructure is the cornerstone of the GH offering.
…every ill person needs a “health URL,” an online meeting place where their caregivers — with express permission from the ill person — can come together, pass on notes to each other, review each other’s notes, look at the medical data, and suggest courses of action. This isn’t rocket science. It is online web applications 101. [Bosworth; December 7, 2006]
c) Automated data mechanisms to gather and store PHI
The people who treat, diagnose, test or dispense medications to patients should be required to deliver, instantly, over the net, at the speed of light, that information to those patients to use as they see fit. If these patients choose to share it with caregivers or health coaches or nursing services, that should be their right. [Bosworth; December 7, 2006]
…the labs you take and the prescriptions you fill out and even the images of your body are usually not available to you in electronic form even though they flow through the net speeding their way to insurers and Pharmacy Benefit Managers (PBMs) and formularies so that, in the end, people can get paid. We believe that consumers have the right to demand that if this information is flowing, it should also flow to them. [Bosworth; May 22, 2007
After attending one of Bosworth’s speeches, consultant and blogger David Williams suggested that GH will negotiate with a range of parties (presumably health plans, labs, imaging centers, pharmacies, PBMs) to allow PHI to flow freely and appropriately:
Google wants to include in their PHR transaction data between physicians and health plans, physicians and PBMs, labs and physicians and so on. They are not planning to rely on feeds from physician EHRs to do this –Bosworth made a point of pegging EHR adoption outside the hospital at <10% though I think he’s understating the truth– but to try to plug directly into the payment streams.” [David Williams, World Health Care Blog]
GH has been fairly clear about its intent — the patient owns the data and has the right to access their own PHI. While there are few specifics, Google likely will use a combination of carrots and sticks to encourage and cajole various parties to help PHI flow freely into and from Google’s repository that feeds the patient’s personal health URL.
d) Interoperable technical standards: XML and the Continuity of Care (CCR) standard.
GH will employ XML to make personal health information transferable across different information systems.
…we have spoken to leading health providers and institutions from coast to coast, we have heard people say that it is too hard to build consistent standards and to define interoperable ways to move the information. It is not! Ten years ago, I heard people saying the same things about how hard it would be to build consistent standards for allowing programs all across the world to share data. I set out with a small band of people to build a standard way to share any information, XML. [Bosworth; December 7, 2006]
GH will embed the Continuity of Care Record (CCR) standard into its offering.
With the advent of standards such as the Continuity of Care Record (CCR) and Continuity of Care Document (CCD) to follow shortly, we are almost there when it comes to computable health data. … as Google looks at these standards and we will, undoubtedly, over time want to support both so that we can help search intelligently for phenotypic information and connect people with the information, places, people, and online expertise that they need, we are likely to endorse these standards, but with the caveat that we will all need precise encodings and precise dosages or quantities to be spelled out. So this is practically a reality. [Bosworth; May 22, 2007]
e) A user interface. Google will need to present users (patients, physicians, others) with an interface to access their PHI through their GH personal health URL. Google is known for starting with basic functionality and adding capabilities over time. Google has a number of options to develop a user interface: 1) GH could build its own user interface, 2) GH could acquire an existing PHR/EHR company and adopt its interface, and/or 3) GH might develop a crude user interface of its own, and allow existing PHR/EHR companies to compete to provide more sophisticated user interfaces to patients.
f) Appropriate security and confidentiality measures. While the devil is in the details, GH seems to appreciate that it must keep PHI secure and confidential. I understand that this issue by itself stirs a tremendous amount of emotion and debate, so I’ll simply acknowledge the importance of the topic without delving into the rat’s nest of inevitable details.
g) Value added functionality. Bosworth’s speeches and writings have hinted at a wide range of possible options:
- Patient information indicating progress compared to evidence based guidelines
- Drug interaction alerts
- Medically reliable (certified) health information
- Recommendations for health care providers fitting patient specific criteria
- Physician/hospital quality information
- Health coaching
- Decision support
- etc., etc.
Many of these and other options could be added over time; Google could choose to develop these additional modules on their own and/or encourage 3rd parties to develop plug-ins to GH.
III. Three Leverage Points
Before you can fully grok the potential of GH, you first need to understand three conceptual building blocks for GH:
- GH can simultaneously create AND dominate the market for next generation PHRs.
- GH can overcome the Achilles heel of current PHRs — the “populating the PHR” challenge
- GH’s business model can create a network effect and could result in rapid, accelerating adoption
A. GH can simultaneously create AND dominate the market for next generation PHRs.
Since 99% of the world would probably struggle with even a definition of a first generation PHR, let’s start there. According to Stephen Downs, SM, senior program officer and deputy director of RWJF’s Health Group:
The current understanding of a PHR is of an online repository of all of the information in your medical record—and that’s way too limiting. [A New Vision for Personal Health Records, RWJF Project HealthDesign; May 2007]
There are two categories of first generation PHRs, each of which has significant limitations:
1) Stand-alone PHRs. These PHRs typically require patients to gather and input their own data. Only a small percentage of patients are motivated or interested in gathering together their PHI, which you’ll remember is 1) scattered everywhere, and 2) not in suitable, standardized formats.
2) Tethered PHRs. Tethered PHRs have a sponsor— typically a health plan, employer, or delivery system. One advantage is that the sponsor often brings “some” data to populate the PHR — perhaps EHR data in the case of a delivery system or claims data in the case of a health plan. However, this data typically is neither interoperable nor transportable — the data will not transfer across multiple providers, and the patient typically cannot “take” the PHR with them when they leave a sponsor.
Tethered PHRs also have other challenges. Adoption is stifled because patients do not always trust that a sponsor is acting in the patient’s best interests. Finally, data that a sponsor brings usually is not complete and/or comprehensive; for example, health plans primarily have access to claims data, which is a proxy for clinical data, but certainly not a substitute.
What’s a “next generation” PHR? Think of first generation PHRs as an application — a end in and of itself, a “place” to store and access PHI. Think of next generation PHRs as BOTH an application and a platform. As a platform, the PHR is not just a place to store and access PHI, but it also becomes a means to support a wide range of new applications.
Indeed, the power of a personal health record lies in its potential to be coupled with alerts, reminders and other decision-support tools that help people take action to improve their health or manage their conditions. From this perspective, the personal health record can be seen as part of a broader personal health record system. Health care and technology pioneers are beginning to develop solutions that harness the power of PHRs to create consumer-friendly tools people can use in their daily life to stay healthy and better manage illness. [Overview—Personal Health Records, RWJF Project Health Design]
The Markle Foundation has also written eloquently about the evolution of PHRs. Their concept of the “Networked PHR” is shown in the diagram below. Feel free to substitute “Google Health” for the words “Personal Health Record”.
While there are 200 PHRs currently in the marketplace, there will be nothing else like GH.
B. GH can overcome the Achilles heel of current PHRs — the “populating the PHR” challenge
Recall that the basic problem with PHI is that it is 1) scattered everywhere, 2) not in suitable, standardized formats. One of the biggest challenges of current PHRs is simply getting your PHI into the PHR in the first place and then keeping the data current.
GH will create systems and standards to address the “populating the PHR” challenge.
1) GH will develop automated data mechanisms to gather and store PHI
As described in the section above explaining GH’s technology model, GH will work with various parties to help PHI flow freely and appropriately into and from Google’s repository feeding a patient’s personal health URL.
2) GH will make the CCR standard the “MP3” of Personal Health Information
While explaining the elegance and the simplicity of the CCR standard is beyond this essay, you will need to grok the CCR standard to really understand the potential of the GH offering. Think of the CCR as 80% of the value of an electronic health record achieved with 10% of the effort.
David C. Kibbe, MD, MBA, — who is acknowledged as one of the developers of the CCR standard — analogizes the CCR standard to the MP3 audio format.
Let’s set one thing straight — the MP3 is an inferior audio format:
- It doesn’t contain all the information of an original recording. Audiophiles sneer at MP3s because much of the high highs and low lows are lost by compressing the audio file.
- Not everything is available in MP3 format.
- It is not a universal format — not every device can play MP3s.
Despite these inferiorities, the MP3 is “good enough” to have taken the world by storm. According to TechWeb:
By the end of the 1990s, music fans discovered that a CD song title converted to MP3 would still sound pretty good even though it was only 1/10th the size of the original CD track. Smaller files meant faster downloading. At an average of 4MB, it took less than 15 minutes to download a file over an analog modem. With a broadband connection on a college campus, it took seconds.
Similarly, the CCR standard is an inferior PHI format:
- It won’t contain all your medical history
- It won’t contain all the detail even about the medical history it does provide
- It’s not a universal solution to broader HIT standards issues
But it’s good enough. The CCR standard:
- Provides a constantly updated record of important PHI
- Is mobile (movable)
- Is accessible through a wide range of devices and media
C. GH’s business model will create a network effect and can result in rapid, accelerating adoption
The author of the Theory of Disruptive Innovation — Professor Clay Christensen — notes that technology by itself does not create disruption. It’s necessary to have the right business model to create disruptive effects.
The GH offering promises to create a powerful network effect. As shown in the diagram below, a network effect in a market results in the value of the network growing exponentially based on the number of users.
The classic example of a network effect is fax machines. When one person has a fax machine, the network has no value; when two people have fax machines, the network has “some” value because the two users can send faxes to one another. When the market reaches critical mass (the tipping point), network effects take over; adoption increases rapidly as people without fax machines feel compelled to join the network or be left out.
GH has potential to reach a tipping point rapidly: patients, government and employers will be openly enthusiastic about GH and will promote adoption (see the discussion below about disruptive potential).
The lack of suitable, formalized HIT standards creates an opportunity GH to set de facto standards. “A de facto standard is a technical or other standard that is so dominant that everybody seems to follow it like an authorized standard.” One example of a de facto standard is the Microsoft Windows operating system, which runs on about 90% of the world’s PCs,.
The timing for GH to set de facto standards around the use of the CCR standard is ideal. While the CCR standard is not yet widely used, the heavy lifting of creating and establishing it has been accomplished. The CCR standard has been approved by key standards bodies and has 11 sponsoring organizations, including HIMSS, the American Academy of Family Physicians, the American Academy of Pediatrics, the American Medical Association, and others.
Is Google powerful and influential enough to set de facto standards in health care? If not Google, who else could even try?
A business model with a strong network effect and defacto standards setting is necessary to solve the fundamental problems of PHI being 1) scattered everywhere, and 2) not in suitable, standardized formats. GH will place the data in one location (the patient’s personal health URL) and will present PHI in a suitable, standardized format (PHI that is updated through automated mechanisms using the CCR standard).
GH also likely will employ other tactics aimed at creating high barriers to entry
- free pricing to consumers
- economies of scale
- Using first mover advantage to create customer lock-in and switching costs
GH will build on build on Google’s existing capabilities — superior search technology, a strong brand, a proven revenue model, partnerships, and market leverage.
While there will be many opportunities to create revenue, Google’s existing primary revenue model (selling ads) undoubtedly will be a key aspect of GH. What’s not clear is how far Google can take PHI to created personalized and targeted ads. GH’s revenue model also cannot easily be duplicated by competitors whose core business is not Internet search.
IV. Disruptive Potential
GH will be very disruptive to existing economic interests. Here’s a quick first take how the GH might impact various players:
Patients/consumers – tremendous upside if Google can get past concerns about security and privacy.
Employers – significant upside. They will see GH as an important step toward HIT interoperability, transparency, greater consumer involvement in decision making.
Physicians – mostly positive, but GH could be confusing for doctors to understand.
- Can GH get past traditional physician caution to encourage rapid adoption?
- Can GH get physicians’ attention long enough to change their workflow? National physician organizations could take a leadership role here.
- GH is synergistic with the medical home model being promoted by primary care physicians. GH creates an opportunity for primary care physicians to reestablish a prominent role in delivering and coordinating care.
Health Plans – mixed.
- GH challenges proprietary IT. Many health plans are vested in their proprietary IT systems; they see proprietary IT as a source of competitive advantage. GH will promote interoperable exchange of information, and thus will challenge many health plans current thinking.
- GH challenges health plans “navigator” strategy. Many health plans are executing a strategy to position themselves as navigator of the health care system on behalf of patients. While this is a good strategy, it’s undermined by GH making the health system more navigable by consumers themselves with or without help from their doctor.
- GH turbocharges health plan cost containment and quality initiatives. The economic benefits of increased penetration of HIT accrue mostly to payers, not providers; the Markle Foundation estimated 89% of economic benefits go to payers. GH promises to be a catalyst for broader HIT adoption, which should make plans jump for joy.
- Bottom line: payers will have second thoughts (mostly privately), but will quickly recognize that resistance is futile and that GH provides many cost and quality benefits.
Hospitals – major disruption, but they will be unable to notice or react.
- Hospitals are a major target of cost containment efforts. Reducing admissions, ER visits and procedures is a major target of health care cost containment strategies.
- However, how does a hospital ever know about an admission that was avoided due to appropriate flow of patient health information?
Imaging centers, labs – mostly disruptive.
- Duplicative testing costs payers, even though it benefits imaging centers and labs
- However, the increased flow of electronic data promises some cost efficiencies over paper, fax, and phone.
V. Concluding Thoughts
Let’s go back to the beginning. The basic problem is that your PHI is 1) scattered everywhere, and 2) not in suitable, standardized formats. The GH technology model and business model offer solutions to both of these problems.
I find the prospect of GH very compelling. Patient care will improve. Costs will go down.
I hope that my projections about Google Health are only the start of a longer conversation. I look forward to your comments.