DAVID C. KIBBE and BRIAN KLEPPER
One of the important decisions before Dr. Blumenthal and his colleagues at ONC and HHS is whether the national health information network will be one of closed appliances that bundle together proprietary hardware, software, and networking technology, or one of open data exchange and management platforms in which the component parts required to do medical computing can be assembled from different sources. If the former direction is chosen, power and control will be concentrated in the hands of a very few companies. If the latter, we could see an unprecedented burst of disruptive innovation as new products and services are developed to create the next generation of e-health services in this country.
Separating the data from the devices and applications, and maintaining a certain degree of independence of both from the networks used for transmission, is far more than a technical quibble. It can determine the economics of technology in stunning ways.
Then, in the summer of 2008, Apple released an upgrade of the software that runs the iPhone and iTouch which included an SDK (software development kit ) that allows third-party developers to create software that can be downloaded and run on the iPhone and iTouch; some of these apps are free and others cost up to a few dollars.
It is not coincidental that the Internet itself was made possible by several legal cases that lost AT&T its monopoly position mandating the use of a specific device – an AT&T telephone – to handle voice data, on AT&T’s private network. As Jonathan L. Zittrain writes in his remarkable book “The Future of the Internet and How to Stop It” —
These [legal] decisions [against AT&T] paved the way for advances invented and distributed by third parties, advances that were the exceptions to the comparative innovation desert of the telephone system. Outsiders introduced devices such as the answering machine, the fax machine, and the cordless phone that were rapidly adopted. The most important advance, however, was the dial-up modem, a crucial piece of hardware bridging consumer information processors and the world of computer networks, whether proprietary or the Internet.
With the advent of the modem, people could acquire plain terminals or PCs and connect them to central servers over a telephone line. Users could dial up whichever service they wanted: a call to the bank’s network for banking, followed by a call to a more generic “information service” for interactive weather and news.
The separation of devices that might utilize data, from the data itself, eventually made it possible for millions of computers of all kinds, makes and configurations to connect with one another over the Internet. Today, audio data in many different formats travel freely over the Internet, and these are consumed, interpreted and put to use by many thousands of devices and the applications running on those devices. If we send you an MP3 file of a song or a conference recording, we don’t need to specify the device or application you will use to decode and “play” that file. If we call you on the phone today, you might answer on a cellular phone, a land line-connected phone, or a computer running Skype. Each of these needs to “understand” the data in great detail: no one dictates which device the consumer has to use.
What do these examples have to do with health IT and HITECH? Just about everything. They illustrate the layers of standards that operate on a digital network: physical wires/cables (or wireless) components of the network; devices connected to the network; applications running on the devices; data of various kinds created and used by the applicaitons; and finally and most importantly, the social interaction based upon the content – that is, the purpose and meaning – of the whole by users.
The future of today’s health care IT systems seems to be converging on a handful of enterprise-run networks that do not interconnect, and a dozen or so health IT applications from companies that have the power to decide who can connect, what tasks they can accomplish, under what terms, and at what cost. The irony here is that while Microsoft and Google appear eminently capable of finding ways to exchange health data securely over the Internet in partnerships with New York Presbyterian Hospital and CVS Pharmacy, among others (see “Patient Records Inch Into 21st Century,” New York Times, April 5, 2009 , and “CVS joins Google Health Rx network: millions can access medication records online,” April 6, 2009), the traditional health IT industry seems to be declaring itself unable to free the data from its proprietary applications and devices quite so easily. Instead, they want a certification process to attach to the software applications from a select group of vendors BEFORE we get interoperability, based on the products’ features and functions, only one of which is the ability to handle the data layer. In the process, we believe they are acting more like the old AT&T than the new Apple iPhone.
Congress in its wisdom attached incentive payments to the social interaction layer of the network: “meaningful uses” are by definition the outcomes of the deployment of networked technologies and data exchanges, not necessarily referable to specific applications and devices themselves. This is real progress. Having done that, the next layer down, the data layer, is where attention ought to be paid in order for successful and widespread meaningful uses to proceed and economic stimulus to be unleashed in the health IT economy.
The government can and should facilitate the private sector’s arrival at agreement about the content of clinical data that it wants to become accessible to providers and patients over the network; it can also decide in what structures those data elements are formatted or packaged; and it can certainly set expectations and specifications for the protection of privacy of the packages and their contents, whether coded data, text, images, audio, or video. With a limited set of available and well tested standards, this could be done quickly and easily, and would immediately be seen as advancing implementation of the “meaningful uses” that Congress and HSS wish to see.
Here is where the Gordian knot of HITECH needs to be untied. There is no need for ONC to regulate the next layers, the applications, device layers or networks, or to link its specifications for the data to specifications for those products and services that will use the data. They need to know when to stop tying the knot, in other words.
Here’s the risk: limiting the kinds of devices and software applications that can handle standardized health care data to a few government (or government agent)-certified products would dramatically stifle innovation and utility while raising (or at least maintaining) costs. Imagine if the government told consumers they had to buy a Dell computer to connect to the Internet, or an iPod to listen to digital music recordings, or computers running Microsoft Vista to use email. Worse still, imagine we all had to use a modern-day version of the private network Prodigy in order to exchange health data of any kind. By locking in the use of data to a specific product/device/application or even a specific class of products/devices/applications, consumers suffer by not having the choice that an open market permits. Nor do they benefit from innovation and competition that arise from comparisons of service, features, usability, and price. And society doesn’t get the economic stimulus, or, rather, it goes to a few large corporations only.
Physicians are, generally speaking, fearful that they will see their EHR choices limited to the expensive and difficult-to-implement products that CCHIT now certifies. David Blumenthal is quite right to worry that doctors will “rebel” if given only the choice of a small number of products that are notoriously not “user-friendly” and not designed to meet the goals of “meaningful use” set by Congress.
But we don’t believe that the deeper problem is CCHIT or the dubious politics of hiring a vendor-run organization to certify the vendors’ products. That real or apparent conflict of interest is an issue, but it’s not the major one.Rather, it is the potential linkage of incentive payments to a certification process that would require specific applications to manage health data. That is the root of the the dilemma that Dr. Blumenthal et al. now face; this is the Gordian knot that needs to be undone.