One Step Forward for Dossia

Things are finally looking up at Dossia. The PHR project was launched awhile ago with all sorts of ballyhoo and misplaced optimism by a consortium of companies, led by Wal-Mart and Intel. After very low uptake by employees, and a disasterous falling out with their vendor, the project seemed to be on the brink of collapse due to an ill-conceived strategy, lack of expertise, and poor execution. It's most recent announcement suggests that they've solved at least the second problem.

The Wall Street Journal reports that Dossia has signed an agreement with folks at Childrens' Hospital of Boston to base the Dossia infrastructure on the Indivo architecture that has been under development at Childrens' for many years (a free preview of the article is available at WSJ, and there's a press release on the Dossia site). Dossia should be congratulated for bringing the Childrens' folks on board. If you ask me, Indivo is the most thoughtful and firmly grounded PHR project in the country, bar none. The folks behind Indivo -- Drs. Isaac Kohane and Ken Mandl -- are bonafide informatics heavyweights, and they've been thinking about this and experimenting with it for a very long time.

I don't think Dossia is out of the woods yet though. While I'm in complete agreement with the Dossia vision (and indeed, very few people in this field disagree with the vision), and they've now got the best informaticians on their team, I still think that their strategy is naive. The basic argument remains: Is higher health IT penetration going to happen from the demand-side (ie, consumers using their PHRs to push their doctors to adopt EHRs) or the supply-side (ie, fostering greater EHR use among providers to make PHRs relevant in the first place)?

My own view is that the demand-side approach is premature right now because there isn't enough electronic information available yet to make external PHRs (ie, PHRs that aren't connected to any particular provider or insurer) an attractive value proposition for most patients. Most clinical information right now is non-structured, non-electronic information that is held by fragmented, disorganized, paper-based provider networks. Some information can be assembled, like medications from health insurers. It falls off pretty quickly from there though. A lot of folks look to the national labs as a source of electronic info, for example, but their penetration is highly variable by market and generally quite low (on the order of 10% in Massachusetts is my guess). And don't even think about external PHRs getting information out of hospitals and doctors' offices -- it's so far from being scalable that it's not even worth talking about.

Overall, I think that we still need to focus on the supply-side -- figure out how to get more EHRs and HIEs into the hands of physicians so that more meaningful information is available to doctors and patients alike. Demand-side pressure can work too, I think, but not through PHRs, but rather, by patients' choosing providers who have EHRs.

That's the way it's working for me. I've signed up with a national PHR company, but for me it was too much work for too little gain. Rather than expending any effort on a PHR where I had to do the work and that was still unlikely to bring my info together easily or effectively anyway, I voted with my feet and moved my care to a provider (Harvard Vanguard) who already has an EHR and can make my information available to me through their own patient portal.

I may not have "control" in the sense that access to my records is governed not by me but by Harvard Vanguard's corporate policies, but if they gather and enter the information for me, AND store it, AND give me electronic access to it for free, that's a worthwhile trade-off to me. In return, the work that I'm willing to do as a patient is to channel my care to providers who are already integrated -- physicians at Harvard Vanguard and the hospitals they're connected with -- rather than spending my own time and money (or having my employer spending MY time and money) trying to integrate unwieldy information from disparate, disconnected providers.

I may not have direct "control" of my info, but I've invested my trust in an organization that I'm confident won't abuse it, and I'm not at all worried that my employer will get access to my info. It's in the hands of my doctors, which is exactly where I want it to be.