I have been interviewing a handful of state health information exchange leaders over the past few weeks for a feature article. One of the states I was interested in was Tennessee, because it has had experienced promising developments with HIE, with three RHIOs in operation: MidSouth eHealth Alliance (MSeHA), Middle Tennessee eHealth Connect (MTeHC) and East Tennessee Health Information Network (etHIN). Tennessee has also experienced some of the pitfalls of HIE with the closure of CareSpark last year.

But to my surprise, when I set up an interview with Keith Cox, CEO of the Health Information Partnership for Tennessee (HIP TN), he told me the organization, instead of taking off, had chosen to wind down operations. HIP TN was created to provide a statewide “network-of-networks” model for sharing of health data. There was also a plan to provide several enterprise services, starting with links to immunization registries. But the state and the HIP TN board of directors decided to change direction and focus on getting more providers to sign up for Direct, which allows for the secure e-mail transmission of health information between providers.

Cox told me it was disappointing to wind down after hundreds of stakeholders volunteered their time to work on an HIE framework.

But Tennessee may have realized at the last minute that they were investing in what Micky Tripathi, CEO of the Massachusetts eHealth Collaborative, has described as an “over-architected HIE"?

What is his definition? An over-architected HIE “tries to do too much for too many with not enough money and time. It tries to establish an all-encompassing infrastructure and service to meet multiple, heterogeneous current and future requirements of multiple, heterogeneous current and future customers. It tries to do all of this with a shoestring budget and staff. And worst of all, it focuses more on long-term potential "big-bang" value at the expense of short-term, realizable, incremental value.”

It appears other states besides Tennessee are downshifting from more ambitious — and perhaps “over-architected” — plans to more basic ones involving Direct. HCI just reported that Oregon is looking to a new subsidized Direct project to be the onramp that many Oregon providers take to health information exchange.

It may be that many states are recognizing that they will not have enough time or money to stand up a platform for broader exchange and find a sustainable business model to support it. Some states may not have a business case for many statewide interoperability services. Or perhaps they will several years down the road, once other business drivers in the Affordable Care Act make HIEs seem more essential to doing business. And the Office of the National Coordinator’s focus on Direct does offer a means for all providers to meet the meaningful use goals about exchange and can perhaps be used as a stepping stone to more robust exchange.

Yet some HIE leaders I interviewed were unhappy about or even disdainful of Direct. Some pointed out that by promoting point-to-point exchange, it may actually hamper interoperability solutions that allow more structured data exchange.

Others said that they had already made progress on interoperability and felt ONC’s pushing Direct was forcing them to take their eye off all the other balls they are juggling to add one more. And they see Direct as much less sophisticated than what they are already doing. “Direct is a step backward from where we are, but there are some providers for who it is appropriate,” one HIE official told me, “so we are rolling it out.”

Another said: “Direct is ONC’s push and it is evolving, but the question is how it fits in the clinicians’ work flow. If it is just a way to check off the meaningful use criteria, we are not going to arrive at what we want to achieve.”

I am wondering if others see Direct as a valuable addition to the HIE toolbox or as a stopgap distraction for the states that are making a push at interoperability? Could ONC be doing more to support states that have made more progress?