Microsoft recently announced the impending shutdown of its HealthVault service, effective November 20, 2019 (see https://www.healthvault.com/en-us/healthvault-for-consumers/) . I am sorry to see it go, but I understand the competitive pressures and shifting landscape that led to its demise. We can learn from studying its triumphs and errors.
You’ve probably never heard of HealthVault (HV), which is part of the problem. HV is (soon to be was) a personal health record that lives in the cloud. Think of it as your Live email account, except that it holds an electronic copy of your data from all health providers. Imagine that Dr. A tests your blood and gives you the result on paper. If you are extremely organized, you take that paper home and store it in a binder, which you bring on your visit to Dr. B, so she can see what Dr. A did. But if you’re like most people, the paper floats around your desktop for a month, until you spill coffee on it and toss it. If Dr. B needs something from Dr. A, she’ll call his office, and they’ll send her a fax. Not exactly today’s standard of care for any other industry, but somehow medicine seems to muddle through with it.
With HV, Dr. A’s computer would upload the data (with your permission) to your HV Record (HVR). Home medical devices, such as blood pressure readers, could also upload their measurements to your HVR. When Dr. B wanted that data, her computer would fetch it (again with your permission) from your HVR. The doctors’ systems don’t need to know anything about each other. All each system needs is an HV gateway. Sounds like a good idea, doesn’t it? Yes, but:
Microsoft prioritized the user’s privacy, which doesn’t sound like a bad idea. All the data in a user’s HVR is owned at all times solely and absolutely by that user. The user could grant Dr. A’s computer the right to access it, and then rescind that right at any time, without consent or even notification of Dr. A. Fine, customer in control and all that, but here’s the snag: this means that Dr. A’s computer system couldn’t use HV for its primary data storage, because his access could be cut off at any instant. HV offers nothing to assist developers of the doctor’s main clinical program. HV only offers interchange with whatever other HV apps might exist. It’s purely a network app, the utility of which grows with the square of the number of users, but which has no utility at all in a standalone situation. Few vendors considered that benefit cost-effective.
Or sometimes even desirable. Given the consolidation in the health care industry, executives care greatly about communication within their own network, but not so much outside it. Boston’s Beth Israel Lahey medical center wants easy communication among its 4000+ doctors. It cares much less about easy communication with the 6700 doctors at Partners, the area’s other dominant practice.
HV could also be difficult to use. Allowing Dr. A’s computer to access your HVR was cumbersome. His system had to negotiate with HV, passing it a security question and answer, providing you with a GUID and a URL. You had to navigate to that URL, correctly answer the security question and then type in the GUID. It was very secure – Dr. A’s system never saw your HV credentials, or even your HV user name. But it was too hard for most civilian users, especially if they were old (hence less computer-literate) or sick or both.
I liked HV from its inception. I taught some classes on it, consulted on several HV projects, proposed a book on it (that MS wouldn’t pay for), used it for my Xamarin CHF weight tracker app (see https://msdn.microsoft.com/en-us/magazine/mt814422.aspx). But Microsoft never solved the chicken-and-egg problem: which vendors would put an HV gateway onto their apps, until enough others did? Almost every vendor took a wait-and-see attitude, and while everyone waited, nobody saw.
Because, finally and most telling, the HV ecosystem lacked a killer app – one that users would join an ecosystem to use. The killer app for the IBM PC was Lotus 1-2-3; for the Macintosh, it was PageMaker; and for Windows, it was multitasking and Solitaire, not necessarily in that order. If Microsoft had somehow provided, say, a HV app that would guarantee an extra fifteen minutes of sleep each night, they might have bought themselves enough adoption to start the virtuous cycle.
But Microsoft didn’t, and now they’re shutting down HV. So it goes. These are the lessons that Microsoft has, I hope, learned, and that you have now also learned on Microsoft’s dime. Bayete, HealthVault.