You can't get blood (or data) out of a stone

More details are now out on the Wal-Mart/Intel health records project. Turns out that it's a personal health records project called Dossia sponsored by the Omnimedix Institute. The project has gotten a lot of attention in the last couple of days. From what I can tell, the effort is, at best, a harmless sideshow. I just hope that it doesn't suck the wind from the sails of the many health IT efforts across the country today.

The reason I think this is a sideshow is that it is based on a faulty presumption, namely, that we can spur EHR penetration by giving PHRs to droves of individual patients. Those patients will want data to fill those PHRs, so the theory goes, and they will, in turn, pressure their physicians to purchase EHRs. Think of millions of patients, with their health-version of Quicken, pushing the "Download" button to get data from pharmacies, hospitals, labs, and physician offices.

The main problem with this approach is that it assumes that there's enough health data out there to make it worth my time as a patient to push the "Download" button. I don't think there is. Furthermore, I think it's a stretch to think that this approach will give enough push to the demand-side to affect HIT penetration.

First, on the availability of data. You have to get data into computers before you can get it out of them. Only 9% of physicians have a good EHR according to the CDC. Hospital use is higher but still not great. According to the American Hospital Association, 37% of hospitals have moderate to high IT implementation, and only 30% have implemented such functions as "access to medical records" and "access to medical history". Perhaps even more telling, a whopping 65% of hospitals say that fewer than 50% of their physicians use the IT that they've implemented.

So, the vast majority of data is still on paper. And even for the data that is electronic, no current systems that I know of have the ability to respond to an outside "query" (like a "Download" request from a PHR). With no physician data and no hospital data readily accessible, Dossia could prove to be very thin record indeed for quite awhile. There are no healthcare analogues to financial behemoths like Fidelity and Vanguard who can deliver a lot of data to a lot of people with a push of a button.

The structure of health care delivery will also make it difficult for even a Wal-Mart to have the leverage to drive higher IT adoption among physicians in this way. Walmart may have 1.3 million employees nationally, but most of them are distributed store-by-store in different health care markets across the country. With about 4,000 retail units in the US, each store employees about 325 employees. So they're spread pretty thin geographically.

On the supply-side, health care delivery is probably the most fragmented sector of our company -- it's very local, among both physicians and hospitals. With fragmentation on the demand- and supply-side of the equation, it's hard to see how a national player like Wal-Mart can exert influence on any given locality. For example, if I'm a doctor or hospital in a community with a Wal-Mart, I'm not going to have that many Wal-Mart employees as patients. So how much influence can their individual "Download" requests really have on my decision to purchase an EHR? In Bentonville, Arkansas, they can probably exert a lot of influence. I'm not sure they can in any other local markets. Sure, there are other employers involved, but it's hard to imagine that their employees' collective demand for PHR data will trickle through to the supply-side.

It's clear that Wal-Mart and many many other purchasers are absolutely fed up with spiralling costs and the seeming inability of the health care sector to modernize itself, and I think their frustration is absolutely justified. I just wish they'd channel their energy in more productive, collaborative ways.......