Google Health launched this week to the public. It is pretty much what everyone anticipated — an online electronic record that people can maintain if they wanted to. Highlights include quick information at your fingertips to your health concerns that you enter in. A user’s profile includes pretty much what everyone has seen in the prior screenshots: Age, sex, height…; Conditions; Medications; Allergies; Procedures; Test results; and Immunizations. Under each profile area, you can enter things in by hand, or choose from long, long alphabetized lists of medications, procedures, etc. Too bad it doesn’t have that “Google magic” and recognize that when you enter in a condition, you may want to step through everything related to that condition in a tutorial-like manner.
Lowlights include the fact that it only allows importing data from 8 different personal health record services (none of which are software-based records which you may have from a Windows program or what-not), and no exporting of your data whatsoever. So much for being “open.” Also, apparently Google is licensing health information from ADAM and incorporating it into the record when you want a quick reference, taking Google down the road of yet again competing against publishers and playing favorites in this space. The user-interface is… clunky, to put it gently. You have to add a medication, test or procedure, and then click on it separately to add its details. It looks like the interface was designed by programmers, not people who actually deal with health conditions on a day-in, day-out basis.
The current service, which carries the Beta tag and likely will for years to come (just as Google’s Gmail service has for years), is labor-intensive and the benefits seem… distant. There are no graphs, no real tracking abilities, nothing to make this health record stand out from any other attempt at this sort of personal health record in the past decade. In fact, I was completely underwhelmed by what it is currently offering.
Over at The Health Care Blog, you’ll see a different point of view from one of Google’s 8 launch partners (you know, the same 8 that you can import data from). Granted, it may be convenient to have a hospital do a lot of the heavy work for you in terms of importing all of your existing data into the Google Health record. But once it’s done, I’m not really sure what you get from it.
John Halamka notes that once you’ve imported your data, you can “receive drug/drug interaction advice, drug monographs, and disease reference materials.”
So what? You can get all of this today and have been able to for years.
He goes on to say, “[Patients] can subscribe to additional third party applications, share their records if desired, and receive additional health knowledge services.”
In other words, patients can then choose to: send their data to yet another service (which may or may not require payment and means your data goes out to yet another third-party who gets to market to you based upon it); or subscribe to additional fee-based services. (I couldn’t find any way to share my record with anyone else.)
If we take this out of corporate-speak, it largely sounds like the Google Health record is an ideal way for other companies to market to you based upon your specific health concerns!
One last note:
The data standards underlying Google interoperability include a proprietary form of the Continuity of Care Record, called CCR/G. Google has committed to supporting the standards which have been recognized by HHS Secretary Leavitt including the Continuity of Care Document. The vocabulary standards used by Google and its decision support partner, Safe-Med, include SNOMED CT, LOINC, NDC, RxNorm, and ICD9.
Naturally, like many, I’m a little leery of anything that’s proprietary and not open-standards in this day and age. I hope Google will publish its proprietary form of this record so all can make use of it in the future.
Updates: The Google implementation of the CCR standard is open and available here as I later learned through some digging.
You cannot “hide” your main profile. When I attempted to do that, it would hide it, but then create another “unhidden” profile with the same name. Perhaps a bug?
Google says they delete your data permanently when you delete a profile. What they don’t say is what happens to the data you’ve shared with third-party providers because of Google Health. Do they get the delete request as well, or do you have to delete your data at every third party service you sign up for?
I should also note for readers that Google is not a HIPAA covered-entity. That means that while your hospital records are protected by HIPAA regulations while in the care and convenience of your hospital, they leave the security of HIPAA the minute you import them into Google.
What this means in reality is that a hospital won’t let themselves be held liable if you’ve transferred your health data to Google, and then had your Google password cracked or what-not. And of course Google won’t be held liable either.
This an important component buried in their 1,600 word TOS and Authorization you agree to when you sign-up for this service. Most people will miss and not understand the importance of the lack of HIPAA oversight of their health data once it’s in Google.
Excellent analysis, John. Thanks. I’ve read a fair amount about Google Health for a while but hadn’t caught the HIPAA angle, nor the cascading-delete issue.
Even without that, I had a lot to say in January. Nobody should dare have an opinion about this without understanding the CNet.com issue and caving in to the Chinese government.
It’s made much worse, I think, by Google being completely disingenuous about the privacy issues. Do they not take the subject seriously, or are they arrogant enough to think they can do anything and get away with it? Either way, I want nothing to do with it.
To me it’s an indictment of the whole company’s integrity – but maybe it’s just really bad judgment (or naivete) on the part of some middle managers.
IMO, it all boils down to whether you think you can trust Google to be responsible with your data, or whether Google can be pressured into doing whatever a government tells it to, or simply might just screw up, at which point there’s nothing you can do about it.
If you’re comfortable with that, give Google your data. If not, don’t.
The community on Slashdot doesn’t seem too pleased with the HIPAA issue, either. That thread has a lot of discussion about why HIPAA doesn’t apply to Google, and some are arguing that HIPAA seriously needs to be updated for that reason.
Others are cheeky about the much-rumored likelihood of GOOG selling vendors the ability to somehow market SOMETHING to you (on an anonymized basis, I am SURE). (I mean it when I say “rumored” – I have no evidence about anything of the sort.)
Amusing wisecracks there: “I can see “Need Liver or Kidneys?” coming up in the recommended searches”, followed by “I’m concerned about what happens when they combine information about who has healthy kidneys with streetview. And then display google ads offering discounts on bathtubs and ice.”
This is a “big yawn” even to most of the world. Nobody asked for this, nobody’s clamoring for yet another PHR, and virtually nobody is going to use this without a “partner” organization making it available as an option. Only eight launch partners?! There’s like a few hundred major organizations they could’ve launched with in health, and they could only find 8. Since the vast majority of Americans have no relationship with any of these 8, you can bet that’s not going to be a factor for a person deciding to use this platform.
The real concern I have is the business model. Google does nothing just for the fun of it. Since they’ve already publicly said they won’t run advertisements on the service, you have to legitimately ask, “How do they make money from it?”
Partnering with hospitals won’t be enough (e.g., hospitals pay Google to get access to the system). The answer is simple — these third-party services that will pay to be featured by Google and allow a user to send all of their patient data to them with a simple press of the button. It’s the Facebook model of data exchange.
While that works great for Facebook apps that are largely for fun (“Be a vampire and bite a friend!”), it seems like a questionable model which to emulate serious health data exchange (especially given the broad permissions granted — write/read, or read-only — no discrimination about type of data sent to whom and under what conditions).
This is billed as a “beta” launch but it looks largely unchanged from what some of us saw from nearly a year ago. In other words, there’s not much going on in this space, and not much for consumers to get excited about.
If you are just tuning in to HIPAA, here are a few links:
Center for Democracy & Technology
http://www.cdt.org/healthprivacy/
Health Privacy Project (now partnered with CDT)
http://www.healthprivacy.org/
Exposed Online: Why the new federal health privacy regulation doesn’t offer much protection to Internet users
(a 2001 report from the Pew Internet Project and the Health Privacy Project of the Institute for Health Care Policy and Research at Georgetown University)
http://www.pewinternet.org/PPF/r/49/report_display.asp
Great post. I’m also in the skeptics’ camp where PHRs like Google Health are concerned, for reasons I’ve spelled out over at BNET.
I agree with John that nothing much has changed since Blogoscoped published screenshots last August. It sure looks like Bosworth’s departure took a lot of wind out of the project’s sails.
In addition, has anyone taken a close look at the indemnification language in the TOS? I’m no lawyer, but something about those provisions struck me as odd, particularly the phrase that (and I’m paraphrasing from memory) “You will defend or settle all third-party cases brought against Google in connection with your use of this service.”
I thought of another source for commentary on HIPAA. Peter Swire, chief counselor for privacy under President Clinton, summarizes how the sausage got made in this 2005 article: Justice Department Opinion Undermines Protection of Medical Privacy
If you are ever at an event where he is listed as a speaker, run to a table up front so you can hear everything and be first in line to ask questions. He knows what he’s talking about.
Look what leapt into my hand on the way to the loo today – the 4/21/08 issue of InformationWeek, p. 29ff, “Cat-Herding Nightmare” (re Web 2.0 collaboration tools).
It’s about the pros and cons of employees putting data out onto third-party Web 2.0 tools like SocialText, PBWiki etc. In a sidebar titled “Impact Assessment,” in the “Risk” column, it says:
“Whenever … information resides on third-party systems, the risk of loss increases – particularly if the information is being shared with business partners.”
If you consider your personal health data to be worth as much protection from vandals as a company’s business data, you might reflect on why InformationWeek would consider this newsworthy.
I have been looking around for more Google Health and bring back these links:
Musings of a VC in NYC: Making My Personal Health Record Public
http://avc.blogs.com/a_vc/2008/05/making-my-perso.html
Google Public Policy Blog: Google Health, privacy, and HIPAA
http://googlepublicpolicy.blogspot.com/2008/05/google-health-privacy-and-hipaa.html
Are other people seeing interesting threads out there? Please share!
Since privacy is a major topic, I thought I’d share this law review article by Daniel Solove:
‘I’ve Got Nothing to Hide’ and Other Misunderstandings of Privacy
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=998565
The article is focused on government surveillance and data mining, but I think it has implications for health care.
For example, I’m pretty taken with his taxonomy of privacy, which he developed as a way to “shift away from the rather vague label of privacy in order to prevent distinct harms and problems from being conflated or not recognized.”
Here it is:
Information Collection
– Surveillance
– Interrogation
Information Processing
– Aggregation
– Identification
– Insecurity
– Secondary Use
– Exclusion
Information Dissemination
– Breach of Confidentiality
– Disclosure
– Exposure
– Increased Accessibility
– Blackmail
– Appropriation
– Distortion
Invasion
– Intrusion
– Decisional Interference
Sorry, one more link about privacy attitudes being different among different groups, say patients dealing with cancer vs. HIV or 20-somethings vs. 30-somethings vs. 60-somethings:
NYT magazine: Exposed
Blogs are obviously completely different from medical records, but in reading this I realized that the author & I (being of different generations) are experiencing the world in very different ways.
A comment posted today on John Halamka’s blog, by David:
I personally don’t know how much I’d count on the FTC to enforce anything, but even then, the point about Google’s extremely weak authentication seems quite valid.
I don’t know what it signifies, but every single individual I spoke to today (none of whom read my blogs) said Google Health didn’t seem like a very good idea because of privacy concerns. And these aren’t cyber-wizards, just ordinary folk.
That second comment on Halamka’s blog wasn’t mine — guess there’s more than one David bouncing around this subject :-). In any event, I did some more digging into the Google Health terms of service and privacy policy and wrote it up over at BNET.
The lack of sharing functionality seems the most glaring early feature omission in GOOG’s personal health info application. Otherwise, they appear to be following their generally effective ‘do something, pay attention to what people do with what you’ve done, make appropriate changes’ formula.
As for what appears to be a mounting fetish for HIPAA as security/privacy protection benchmark, I confess I’ve been almost as underwhelmed by the purported ‘protection’ HIPAA provides, over & above preceding legal protections for the those ever-elusive qualities, privacy & security* , as I am by the desultory enforcement of that statute. The only figures I can find indicate around 26,000 complaints, & 4 convictions over the statute’s existence. Let me be first to caution these figures may be out of date, and/or taken out of context.
Either the risks are exaggerated, or the damages have generally been less than anticipated, or the law requires substantive overhaul – or I need much better tutelage in the statute’s potency than I’ve gotten to date (from a group of people that includes at least one present at the law’s creation).
*(kudos to John for providing some links to thought-provoking if not particularly conclusive materials on those topics. I particularly enjoyed the Solove piece, and this passage in the conclusion: “…understanding privacy as a pluralistic conception reveals that we are often talking past each other when discussing privacy issues. By focusing more specifically on the related problems under the rubric of “privacy,” we can better address each problem rather than ignore or conflate them.”)
Hi 1st GOOG no harm,
The HIPAA fetish may turn out to be a side show, but I am continuing my education by reading Daniel Solove’s book, Understanding Privacy. I’m only on p. 22 but I’m never going to field another broadly-worded survey question related to privacy again b/c of his insights. So if you liked his article, buy the book!