Defunct Videotex reveals pitfalls in developing a healthcare ecosystem

By Shawn Vincent

Does anybody remember the Interactive Videotex (http://en.wikipedia.org/wiki/Videotex) systems from the 1970s and 80s? One of the more successful ones was Minitel in Europe. Videotex systems were an early internet variant – the telcos saw an enormous market for computers in homes, providing access to information and services: online purchases, searching for telephone numbers, sending electronic mail, and so on.

Fundamentally, the user experience of having a computer in your home to access these services was the same vision as the internet. Videotex systems were not technically inferior to the internet – they could provide all of the services that the internet, at the time, was capable of providing.

So why, today, is the internet a worldwide phenomenon, and Videotex terminals are interesting relics that occasionally show up in collections and computer museums?

The answer is one of economics. A single central agency that is attempting to build a monolithic system is less likely to succeed than are a number of independent agencies.

Why? Because most ideas fail to lead to successful implementations (https://docs.google.com/file/d/0B0QztbuDlKs_bHdnQ2h5dnNvcE0/edit). Studies have shown that up to 80% of ideas, many of which look good on paper, fail to be successfully implemented.

The real problem is that building on something like Videotex is not just one idea – it’s many. Each service provided to the consumer is somebody’s idea, and needs to be executed well. Unfortunately, for the service as a whole to be useful, you need multiple ideas to succeed. If the central organization that controls the environment must own the risk of all of these ideas succeeding, you have a recipe for cost overruns and overall project failure.

In a centrally controlled system like Videotex, it is very unlikely that multiple competing versions of an equivalent service will be provided. The barrier to entry for new services is high: You need approval from (and often need to pay money to) the controller of the service.

The internet, on the other hand, is an ecosystem. There are some basic connectivity services provided by the infrastructure, but end-user services are completely deregulated. Any organization or individual can, with an afternoon’s worth of effort, put a service onto the internet that is immediately accessible to every internet user in the world. The costs of getting a service in front of users are very low.

The result is astonishing. Rather than a single photo-sharing site, we have hundreds. Rather than one way to look up phone numbers, you have many competing choices. The good ones survive; the poor ones don’t get used, and die.

The effect of this is that the internet itself has had a much higher likelihood of success and providing value. Because there is no central authority that attempts to control the services provided, it’s much more likely that the constellation of services available to a user will be high-quality, and provide real value.

How does all of this relate to healthcare IT (HIT)? I claim that most jurisdictions are creating the next generation of Videotex. By adopting a strategy of central control and specific provided services, the jurisdictions own the risk that the strategy as a whole will fail.

What is the alternative? Here, things get dicier. This is entering territory that has not been well explored.

The U.S. government’s Meaningful Use initiative pays for outcomes rather than specific solutions. This forces the market to attempt to solve problems and rewards the victors. This pushes the risk to the market and investors, who are used to and comfortable with such risks.

Even in the United States, however, most HIT initiatives are centrally controlled, and suffer from the Videotex issues I’ve outlined above.

There are a few initiatives, like the Direct Project (http://directproject.org/) and Blue Button (http://www.va.gov/bluebutton/), that suggest a way forward: a number of low-level shared services that can be leveraged by various organizations to provide value without strong central control.

I worry that the current focus on central HIT programs that emulate the Videotex model will lead to many more failed projects, and consumer disillusion with HIT in general. My dream is that the industry will create an ecosystem that takes off like the internet, and that we will enter a new era of HIT that provides real value for care providers and patients.

Shawn Vincent is Vice President, R&D, at MD Practice Software LP.

Posted December 13, 2012