In a previous post, I set out to review Microsoft’s new health-related Web services, including its new online medical-record system, HealthVault. Plans went awry, however, when it took close to two hours simply to register for the service, so I decided to review the medical-records system in this post.
Microsoft clearly has grand plans for HealthVault, the main page of which you can see by clicking the thumbnail at left. Not only has it beaten rival Google to market with its offering (see our previous coverage here and here), it has lined up partners ranging from the American Heart Association to the Mayo Clinic and several other hospitals to a number of online health sites that have all agreed to support the Microsoft service and to make it easy to transfer at least some medical data into patients’ HealthVault account.
This is a fine idea, so far as it goes — which, so far, unfortunately isn’t very far. While all this will undoubtedly evolve over time, for now what’s most surprising about the HealthVault service is how little there is there. What is there is a real mess — a disjointed, confusing service that for now is mostly likely of limited use to the vast majority of patients. (For a screen shot of a classically incomprehensible page, click on the thumbnail at left.) And that’s setting aside perennial questions such as privacy and security, which are obviously going to loom large for many people.
Taking HealthVault at face value, there are two basic problems — one implementational, the other conceptual. The implementation issues are most obvious, starting with the fact that once you finally sign up for a HealthVault account, there simply isn’t much you can do with it at this point. You can add personal details to your profile and upload documents — what sort of documents isn’t at all clear, although there’s an intimidating “code of conduct” to read before you do anything — and set who you’d like to share your information with (presumably family and your doctor or doctors). Beyond that, the only clear suggestion your HealthVault account itself makes is to sign up for a service offered by MaxEmail that will let your doctor fax information directly into your HealthVault account, at a cost of $9/year for up to 100 pages.
There’s no sign, however, of Microsoft’s much-touted partners until you leave your account and return to the public HealthVault page, where you can visit just over a half-dozen sites that can transfer medical information such as blood pressure, blood-glucose values and the like to your personal HealthVault record. (Of course, you have to log in to HealthVault again each time you visit one of these sites.) The one I checked out in detail, the AHA’s Blood-Pressure Management Center, is every bit as confusing as HealthVault itself, since after you enter your profile — again — there’s very little indication as to how you’re supposed to get the data or where it’s going to go.
These are early days, of course, and over time I assume Microsoft and its partners will iron out a lot of these implementation issues — although my quick once-over of the service suggests that this territory is probably trickier to navigate than many people might have previously assumed. (This might also be one reason Google’s health guru, Adam Bosworth, bailed from the company a few weeks ago under less-than-clear circumstances. For an interesting, albeit conspiracy-tinged, take on that, see here.) Putting together an entire infrastructure that supports personal health records is a staggering undertaking — one that might well be beyond the resources of even a company as large as Microsoft.
All of which also leads us straight into the major conceptual question here, which is exactly what good a personal medical record is actually going to be. Ultimately, of course, the vision is for individuals to control every detail of their medical history at their fingertips so they can share it with any doctor or health organization they choose. Setting aside the enormous question of what it takes to get there from here, the big issue this vision raises is whether giving individuals “control” over their records — which is really shorthand for the ability to add, delete or change information — might undermine their usefulness in a significant way. In other words, there’s a downside to the electronic medical-records issue that, so far, virtually no one outside a small coterie of academics has even really begun to address.
I can’t claim to have delved deeply into the subject myself, but I was first tipped to the issue by Graham Walker, a third-year Stanford medical student who blogs at Over My Med Body!. Nearly two months ago, Walker wrote an impassioned post arguing that patients shouldn’t have the ability to cherry-pick and edit their own records if the resulting information is to be at all useful to doctors.
Walker offers several examples that gave me pause. For instance, what about patients who delete or restrict access to their sexually-transmitted disease status? Or those who simply don’t update their records very often? Are patient-controlled records intended to supplement or supplant those maintained by medical institutions? Of course, his questions originate from a doctor-centered perspective, but it’s still one worth pondering as the likes of Microsoft, Google and their ilk start us hurtling down the path to patient-centered medical records.
VB's research team is studying web-personalization... Chime in here, and we’ll share the results.