Credit: Technology Review
Web
How a Broken Medical System Killed Google Health
Google would have had to fix a balkanized U.S. health-care system to make the service catch on.
- Wednesday, June 29, 2011
- By David Talbot
At the end of this year, Google Health will flatline. The service couldn't encourage many people to import or analyze their health data, and experts say its untimely death is, in many ways, an extension of U.S. health-care providers' failure to share data across institutions, or make it easy for patients to obtain it.
Google's free online service lets people upload, store, analyze, and share their health information. But there are hundreds of different health-care institutions in the U.S. that use different systems to record and store data, and many doctors don't use electronic records at all, making the task of retrieving and updating data extremely difficult for the average person, says Isaac Kohane, who directs the informatics program at Children's Hospital in Boston, and codirects Harvard Medical School's Center for Biomedical Informatics.
For Google to make its service attractive, it would have had to solve this health IT mess, which is in the early stages of being addressed through recent national policy moves. These include 2009 federal stimulus incentives for doctors and hospitals to adopt electronic medical records, and for hospitals to share data with one another.
Kohane says it will be at least five years before data flows smoothly enough to make something like Google Health worthwhile. "Google is unwilling, for perfectly good business reasons, to engage in block-by-block market solutions to health-care institutions one by one," Kohane says, "and expecting patients to actually do data entry is not a scalable and workable solution."
Google did forge some partnerships—including one with the insurer Blue Cross Blue Shield—that let patients upload insurance billing and medical information into its service more easily. Even so, the user experience was uneven, as Technology Review described in 2009. Some patients, for example, would need to obtain copies of their records and then manually enter all the information.
Google announced on its blog late last week that Google Health would be canceled as of January 1, 2012. Users can retrieve their data for an additional year, but after that, the data will be deleted.
Other, similar services survive. Prominent among them is Microsoft's HealthVault. Microsoft was quick to announce ways that Google Health users could move their data over to HealthVault. In a blog spelling out this process, Sean Nolan, chief architect and general manager of Microsoft's health solutions group, also said "the only way to fix health care is to consumerize it," and added that "we'll get there, and Google Health moved the ball forward."
Some limited health data exchanges among hospitals and other health-care providers are cropping up around the United States, but right now, there's no requirement for full sharing. "There is still no flow," Kohane says. "There still has to be additional cultural shifts to actually make the information flow."
Through 2019, up to $36 billion could be spent under the federal stimulus for adoption and meaningful use of electronic medical records. (The final federal outlay—taking into account expected savings and income from penalties for failure to adopt the records—is projected to be $19.6 billion.) "The investment has been committed, but a lot of the money has not been spent yet, and we will see in the next couple of years whether it will happen or not," Kohane adds.
But Google isn't waiting to find out.
Related Articles
Health IT Saves a Life in Memphis
A medical data-sharing program saved at least $2 million and gave doctors crucial insight about a pregnant woman's complications.
A Better Personal Health Record?
In the wake of Google Health's collapse, one company hopes to promote personal health tracking through employers.
Searching for the Benefits of Google Health
Amid a growing emphasis on electronic health records, one patient tests online health services.
Marrach
34 Comments
- 693 Days Ago
- 06/29/2011
Too early. . .for the MD Culture, that is. . .
The Idea was great on its face. But the Technocrats failed to take into account the culture of the Medical Oligarchy.
The title of the article hi-lights the first fallacy-- There is NO Medical System in the US. There is a conglomeration of separate insurance plans, Hospitals, clinics and individual practices that all happen to work within the borders of the US of A. The CDC issues guidelines and occasional rules. The AMA attempts to say they speak for all Medical professionals (Or at least the ones who pay dues) and the Legal profession is trying to sue all the rest.
EMR's are Just beginning to make themselves known industry wide-- and the Numbers of EMR systems are Legion. And that's not counting the Special custom EMR projects that every Major Hospital is crafting for themselves.
Each and EVERY EMR variant is proprietary in their Data Format. Each and EVERY EMR hopes to be THE STANDARD for the rest of the Nation. Rest Assured that Colombia Presbyterian could Care Less what Data Format San Francisco General uses, much less Ohio State University Medical center.
And of all the EMR's out there, we don't know which of the 95% of the whole will die off and disappear.
So whose Data format is Google gonna talk to?
Then there are the Doctors and the Hospitals and their Lawyers. Internet Savvy Patients are Activist Patients and the LAST thing Hospital Administrators and nervous Doctors want to do is give these people direct, uncontested, LEGALLY problematic access to their Medical records that may or may not have 'Interpretative Outcome Differences'.
And INSURANCE Companies are not on board with Internet Accessible Data that can allow Insured Activist Patients to 'Compare Notes' on coverage issues like 'Denial of Services & Procedures'
First-- the EMR field has to thin out to clear winners. Data Standards must be then be battled out. And Doctor Use of any EMR must become the NORM instead of the exception. Then the rest will follow.
flyingmonster
29 Comments
- 692 Days Ago
- 06/30/2011
Re: Too early. . .for the MD Culture, that is. . .
Easy solution, mandate a common format that all hospitals etc must be able to export data to and import data from. All software suppliers get on board to agree to the common data format by a fixed date or get penalized. Is it not a matter of national interest and security to reduce costs across the board and protect the citizens? Medical costs are causing national harm already.
VCRAGAIN
37 Comments
- 692 Days Ago
- 06/30/2011
Re: Too early. . .for the MD Culture, that is. . .
any special interest is going to fight tooth and nail to keep control of it's area - even if that fight is just a do-nothing reaction - watch them
just NOT co-operate to kill this idea - the only way ANYTHING happens to benefit mankind in general is when it's DECREED and made law by a higher power - otherwise we just watch the shenanigans of each side trying to outsmart each other - humans
are so predictable !!
Basil Seal
3 Comments
- 692 Days Ago
- 06/30/2011
Re: Too early. . .for the MD Culture, that is. . .
If humans are so predictable, then...
Why did Google try this when failure was a foregone conclusion?
Why haven't you become a gazillionaire exploiting this predictability in the stock market?
Basil Seal
3 Comments
- 692 Days Ago
- 06/30/2011
Re: Too early. . .for the MD Culture, that is. . .
An sensible solution but too little opportunity for graft and corruption to have much appeal in our capitol. Instead we got a bill that mandates we become customers of these very same insurance companies. This makes the insurance providers very much like a public utility with similarly dismal prospects for a grid upgrade.
jjs
81 Comments
- 692 Days Ago
- 06/30/2011
There is one public EMR that I know of - VISTA, the award-winning system that runs the VA. Developed for and by the health care providers, it has won awards not just for design, but for it's ease of use.
Developed by US Government personnel, it is available to all (public domain). Two sites to get it from are hardhats.org and worldvista.org
In addition, HL7 is the standard for data interchange.
RebeccaCoelius
1 Comment
- 692 Days Ago
- 06/30/2011
Re: Too early. . .for the MD Culture, that is. . .
The title of this article and this comment are utterly ridiculous.
Google Buzz also failed miserably... is it because we are all antisocial and too backwards thinking to have appreciated an amazing new product?
Google created the wrong product for the current ecosystem, period. They didn't understand the complexity of data sharing between the current EMRs (which last time I checked were created to be that way by private technology companies), the serious roadblocks to freely sharing data due to HIPAA laws, and made a huge error when assuming consumers would want to spend their time uploading enormous amounts of their own data, or have anything in particular to do with the majority of it because for most people for most of their lives its both clinically meaningless, and meaningless in the context of their everyday lives. Did you notice that all of these enormous problems are actually EXTERNAL to the health system you are demonizing?
Create something that some part of the medical industry actually has an incentive to use, and when you do stop overlooking the fact that we are paid by the government and insurance companies per a limited number of services or procedures rendered. More importantly in this case though, create something that patients as consumers want to use, and it doesn't matter how inefficient the system is. Inefficient systems are opportunities.
Basil Seal
3 Comments
- 692 Days Ago
- 06/30/2011
This overlooks the more obvious cause of failure: that the average person finds the idea of turning their medical records over to a company that makes its money by invading your privacy a bit off-putting. Sure the average user might happily permit some tracking in exchange for a free email account but that same user has a pretty good visceral sense that just maybe handing over the entirety of one's medical records to such a company is not such a good idea.
davepruss
1 Comment
- 692 Days Ago
- 06/30/2011
Some of the earlier comments indicated that there is no common format to exchange information. As part of the Meaningful Use legislation, a requirement was made that all certified EMRs had to be able to produce information in either a Continuity of Care Record (CCR)- format specified by ASTM or a Continuity of Care Document (CCD) - format specified by HL7 with coded information included in the documents limited to a few standard code sets. GoogleHealth supported the import of CCR documents into their PHR.
One possible reason for the demise of GoogleHealth and the continued use of Microsoft HealthVault is the different approaches that the two companies took in targeting their product, the ease of use of the APIs and in the requirement of third parties to supply add-ins and users.
Microsoft's approach was to target communities that have the most to gain from a PHR, those with chronic conditions. This enabled them to partner with non-profits who would encourage their use and home care products that assist in the treatment of these diseases. Google's approach was considerably more broad brush.
An active community of developers and frequent releases of new API capabilities was also supported by Microsoft, enabling these partners to add value to the platform. I also know that Microsoft approached may EMR vendors and large hospitals to partner on integrations with their EMR and patient portals. My impression is that Google did not make this level of investment in their platform.
By the way HealthVault supports the import of CCDs and CCRs as well.
There is certainly much that can be done to improve the US healthcare system (such as it is). To blame the system and the number of EMR vendors for the demise of GoogleHealth seems a bit of simplistic. These facts may have contributed but they are certainly not sufficient conditions.
pameline
1 Comment
- 691 Days Ago
- 07/01/2011
The very concept of EMR is severely flawed, as is current medical organization according to Atul Gawande (http://www.newyorker.com/online/blogs/newsdesk/2011/05/atul-gawande-harvard-medical-school-commencement-address.html). We no longer need Cow-Boys, but teams.
What is needed are concurrent engineering tools when EMRs are typically document (sometimes data) management systems only.
A paradigm shift must occur, and Google Health was obviously not the proper system for this. This blog post resonates to me as saying "the paradigm shift couldn't occur because the old paradigm must get fixed before something can happen". Of course not!
tcaruso
7 Comments
- 687 Days Ago
- 07/05/2011
Re: Paradigm shift... or what?
I agree. What we need is a Paradigm shift. Or what Apple asks us to do: Think Different.
Health data in the future will be put into a cloud. Public Health Record applications will access information from that cloud after being given the authority to access it by the individual that the information is about, or by their proxies. One store of health information will exist, not multiple copies (except as backups of the cloud, of course). The only information someone could download from the health data cloud would be unidentified anonymous information, because that's all anyone needs to provide public health, disease prevention, healthcare and insurance services. For more information see my blog about Health Data Clouds at http://tpcaruso.com/2011/06/healthdatacloud/.
Yes, Think Differently...that means Google too.
ncalmd
5 Comments
- 689 Days Ago
- 07/03/2011
Having practiced primary internal medicine in a mountainous rural area for 30 years, if anyone needs the benefits that computerization can bring to the practice of medicine, its me. Yet my worst nightmare is that a new patients medical records are coming from an EMR. Given that EMR's can't communicate, I'm going to get the printed output from a system, that will produce medical records that take 5 times as many pages as a traditional medical records, and given that each approaches their output in a proprietary manner, its a bit like a scavenger hunt to find information that I want. In short, EMR's used in physician practices, make information sharing far more difficult.
The reason is that providing information to the physician is only part of the function of EMR. The most important function is to be able to meet the documentation requirements of Medicare and other payors and meet legal and regulatory needs.
If the feds goal is to improve patient care and avoid redundancy, they need to start at the beginning and consider what would improve the ability of physicians to care for patients, rather than how they can be sure they don't get "cheated".
Create regional databases, that capture prescriptions filled, lab and test results, all invasive procedures and key hospital reports.
When I see an acutely ill new patient the most useful information possible would be a list of their current medications with the starting date and the starting dose for each medication. It tells me what significant medical problems, something about the progression, and how long they've had the problem.
Throw in a list of allergies, and what the "allergy" consists of.(most allergies reported by patients aren't true allergies), and a chronology of their surgeries, I've got 80+% of what I need.
Then provide access to their laboratory reports organized in a uniform format and tools to manipulate the data.
Finally give me access to important documents: X-RAY and other medical prodedure reports. Hospitial history and physicals and discharge summaries.
In five to ten minutes, I could have a very good sense of even very complex patients, if I was using software that was customized to access these databases in a manner optimal for my specialty.
It would obviously be difficult to solve the access and security issues for these "public" databases. But it would end the problem of duplication and provide a starting point to a range of applications that could dramatically improve physicians ability to care for patients.
In short forget the EMR's for now, and focus on providing information and applications that improve patient care.
Marrach
34 Comments
- 686 Days Ago
- 07/06/2011
You have very good points.
We are starting the process of 'EMR'-ing our practice and I ran into a fundamental observation. Almost ALL of the EMR's were developed by the various Companies who focused on the types of practices/Clinics that would quickly be able to afford and move into them: Specialists.
The Immunization sections are skeletally scant. Surgeons don't care if you've had your MMR series.
The Systems are focused on SENDING reports. But the System for INPUTTING/SCANNING/CATALOGUING of OTHER Specialist reports is rudimentary. Again-- the Surgeon could care less what the Dermatologist thinks and Vice Versa-- but the Primary Care Doctor has to receive, Integrate and Catalogue ALL Specialist reports into a usable and intuitively useful scheme.
I have the same criticism for the Hospital Based systems. And WORSE, the Hospital Systems are heavily weighted towards Legal water-tightness. Those systems were less designed by Practitioners than they were by the Legal Dept. The Last time we got a requested medical record for a patient from a Hospital, I was stunned when my assistant dumped a REAM and a HALF package of Printed paper on my desk. It made a ghastly 'THUNK' when it hit my desk and scattered my work to the floor.
No-- EMR's have a long way to go. And NO-- No-one can even MANDATE a standard system yet, because people are JUST BEGINNING to see the DIFFERENT ways Various Medical professionals actually DO their Charting. The Internist Charts are NOT Pediatric Charts. The ALLERGIST Charts are NOT CARDIOLOGY Charts.
In short-- like I said before-- we don't have ONE MEDICAL SYSTEM. What we have is a large variegated patchwork that until now was multiply-mediated on the fly by a 'Medical Information Interpretation Algorithm'
We call that Algorithm by its simpler name: The Doctor.
wdavis283
1 Comment
- 686 Days Ago
- 07/06/2011
It is instructive to review recent past attempts to tame this beast.
http://en.wikipedia.org/wiki/David_Brailer
On Twitter
Become a Fan on Facebook
Subscribe to the Feed
mfolbe
49 Comments
medical records
The article insinuates that it is better to have electronic medical records that don't talk with other companies than paper records that don't talk with other companies. Actually, there is no difference. Except of course that doctors who have the electronic medical records have to pay the companies that make the electronic record. Until there is a standard that allows all companies to to share the information, meaningful use is just "pie-in-the-sky". Couple that with a system that protects information that you don't want other people to know about like AIDS, congenital heart conditions, and the like. It will not happen for years.
Even better, the government outlaying all this money to doctors to adopt medical records is just a pass-through and money grab for the technology companies. It's not like the doctors are making anything from the deal. And since it is a government program, the costs are outrageous. The big panacea is that tests won't get repeated and unnecessary tests won't get done. Really though, who is deciding what is necessary and what is unnecessary? It is very easy to say that your back MRI is necessary and everyone else's is a waste. There are already many insurance companies that require prior authorization for scans and surgical procedures. Electronic records may be coming, but I don't think it is going to save money.
If the government really wanted to save lives and money, they would entirely eliminate any tobacco subsidies to states and tobacco farmers. They would mandate exercise programs to our young people so they get off their butts. They would make internet only available for a few hours a day. The same could also be said for xbox, wii and all television for that matter. They would shut down the buffets in Vegas and every chinese buffet in town for those who have no control of what they pile on their hips.
Nice try Google.
Reply
sdmitch16
14 Comments
Re: medical records
Mandate exercise programs for everyone? even people who already workout? and how much would this cost? The only thing that should be forced is not smoking.
"They would make internet only available for a few hours a day."
If everyone only had internet for a few hours at the same time it would create huge power draws requiring massive servers, which would be completely wasted when they were off 20 or so hours a day.
I know plenty of people of healthy people who eat at buffets. The thing you should really go after, aside from smoking is alcohol.
Reply