Archive for February, 2006

Louisville to be first eHealthTrust™

Wednesday, February 1st, 2006

Today, Louisville, Kentucky, became the first community in the nation to announce plans to build an eHealthTrust to provide lifetime health records for all its citizens. This is an exciting and important event in the development of our National Health Information Infrastructure (NHII) in the U.S. that should be closely tracked by the many other local and regional efforts working on these same issues. It also provides a tremendous opportunity for health information technology (HIT) vendors to participate in this truly groundbreaking project.

What exactly is Louisville planning to do?

After more than two years of research and community organizing, the leadership in Louisville concluded that an eHealthTrust would be the best approach to developing a community health information infrastructure (see last month’s posting for a description of the advantages). Having chosen the model, the question Louisville faced was “how to get started?” Working with stakeholders in the community, it was clear that the idea was appealing to purchasers of health care (e.g. employers) because it would make the complete patient information available that would allow higher quality and lower cost care. They expressed substantial enthusiasm about providing the eHealthTrust to their beneficiaries, and even subsidizing the modest cost (about $5/month). However, there was little appetite in the community for underwriting the development costs.

It became clear that if there were enough customers to guarantee sufficient revenue to cover operational costs (about $6 million/year), then the system could be procured as most enterprise software is — by offering vendors a contract to operate it profitably. Not only does this provide an incentive to offset development costs, but it gives the vendor the opportunity to market the software to other communities. This is better for both Louisville and subsequent communities that implement eHealthTrusts, because the software will be available on the open market rather than being owned by a community non-profit that is not focused on its dissemination.

So today, the Louisville Health Information Exchange (LOUHIE) issued an RFP inviting bids for operating their eHealthTrust. The successful vendor would receive a four-year cost plus fixed fee plus market incentives contract to provide management, marketing, operations, and security. The maximum allowable bid is $4.5 million/year plus a $500K fee and $10/net new member marketing incentives, making the total contract potentially worth over $20 million. A bidders conference is scheduled for 2/13/06 (in San Diego), required letters of intent are due 2/27/06, and the proposals must be submitted by 5/1/06. Anticipated start date is July 1st.

In addition, the community wanted to be sure that eHealthTrust members had access to other health information resources, customized by the information in their own lifetime health records. For example, being able to check your own list of medications for interactions without retyping all the information would be exceedingly convenient as well as potentially life-saving. The community’s goal is to provide a wide selection of such health information services from a variety of sources.

To address this issue, the capability for sending selected non-identified parts of a medical record to an outside information service vendor needed to be included in the system. In addition, all such services must have secure, established mechanisms for receiving the information. The result was the second RFP, which invites health information service providers to bid on licenses to link with the eHealthTrust to provide such customized services.

The successful bidders will have a link on the Louisville eHealthTrust web site that takes the user to a vendor-specific page with a list of available services. Selecting one of them, e.g. drug interactions, will result in a window being displayed with the non-identified information from the user’s medical record that needs to be sent in order to provide the service — in this case, the list of current medications. The user is then asked to approve the secure transmission of this information. Once approved, the data is encrypted and sent (via web services), and the user is shown a new window at the vendor’s web site with the resultant list of drug interactions.

A maximum of ten such licenses for these links will be awarded to information service vendors submitting the highest bids. Louisville will guarantee at least 100,000 subscribers to the eHealthTrust (although they cannot guarantee any minimum level of usage of the links). The minimum bid amount is $200K, with the same timeline for proposals as the first RFP.

Why is this important to other communities?

Many communities have started to plan and organize the development of a community health information exchange (HIE), recognizing its potential value in improving health care quality and increasing efficiency. While there is a great deal of excitement and optimism, these efforts are exceedingly challenging. In particular, communities typically are stymied by four key issues:

  • Business model: who will pay for the HIE? (startup and operations)
  • Privacy: how can privacy be protected effectively?
  • Complete information: how can complete electronic medical information be available (especially when most physicians still use paper records in their offices)?
  • Intercommunity connection: how can the HIE be built to ensure that it will be compatible with other communities?
  • Business model

    One of the most critical issues communities face in building HIE systems is funding — both for the startup and continuing operations. Louisville has addressed both these issues. The ongoing funding is being provided by securing commitments from subscribers to pay for the system when it is available. The eHealthTrust model lends itself to this approach, because it clearly can provide complete electronic medical information for each subscriber — which can then be used to provide better and more efficient health care. It is relatively easy for prospective sponsors to understand how the eHealthTrust model does this — thereby facilitating their decision to commit to participation.

    By obtaining these initial customer commitments, the need for startup funding to create the initial eHealthTrust product can be eliminated by asking the vendors to build it in exchange for a multi-year operations contract. This is also better for the vendors since it allows them to profit from marketing the system elsewhere. The risk to vendors is minimal since Louisville is guaranteeing the needed customer revenue for operations. Meanwhile, the second RFP is designed to help provide some startup funds for the LOUHIE organization itself. In addition, it can provide a buffer against any unanticipated shortfalls of operational revenue.


    Protecting the privacy of medical information is essential for an HIE system. This requires both establishing appropriate policies about what information is accessible to whom, and then enforcing those policies with good personnel and computer security. It is the area of appropriate policies where the most difficult problems arise. Devising general policies that cover every situation and satisfy all stakeholders is maddeningly complex, and may not even be possible. The Louisville eHealthTrust solves this problem by making each person responsible for controlling access to their own health information. In this way, each subscriber can determine their own policies about who can see what information and when — and the decisions can change as needed. By adopting the principle that each person must authorize all use of their own health information, every person’s privacy can be protected as they wish — and each person can make their own choices in their own way.

    Complete Information

    Missing information is what leads to inefficient, error-prone health care. Therefore, successful HIE systems must make complete (or nearly complete) patient information available when and where needed. The most difficult information to provide is from outpatient records, which remain largely paper since over 85% of physicians do not yet have electronic health records (EHRs). There has been a tremendous amount of discussion and study of the “slow adoption rate” of EHRs by physicians, but the basic issue is that most of the benefits do not accrue to the physicians being asked to foot the bill, but to those who pay for care. To overcome this barrier, the eHealthTrust provides direct payments of about $3/outpatient encounter to physicians for submitting standard reports of encounters, resulting in $10-20,000 of additional annual revenue to offset the costs of an EHR system (which can require over $40,000 of initial investment for each physician). This will be a potent stimulus to physicians to acquire and use EHRs. No other proposed HIE model even attempts to solve this problem. Unless physicians are using EHRs in their offices, it will not be possible to have complete patient information in an HIE system, since the outpatient records will be missing.

    Intercommunity connection

    There has been a great deal of discussion about the need for HIE systems in different communities to interconnect. For example, the current Federal demonstration projects REQUIRE such connections. The theory is that to be able to immediately retrieve information about a patient from any location where care has been received, queries must potentially be able to travel nationwide. The need for such intercommunity connections creates understandable anxiety for HIE planners. Since the standards and procedures for such connections have yet to be defined, how can communities build HIEs without the risk of being incompatible with some as-yet-unknown future requirements?

    But the eHealthTrust model does NOT require connections between communities — each person’s entire record resides in the community eHealthTrust. No queries beyond the eHealthTrust that contains each person’s complete record are needed. If a community member ever receives care from an out-of-town provider connected to another eHealthTrust, any information sent to that provider’s eHealthTrust will automatically be “redirected” to the “home” eHealthTrust. This “redirection” does not have to occur instantaneously, since no one is waiting for the record to be retrieved when the information is sent.

    Why is this an exciting opportunity for HIT vendors?

    HIT vendors should be excited by the opportunity to respond to this RFP for a community eHealthTrust HIE system. With a potential value in excess of $20 million, it is the largest such RFP ever issued. The selected contractor (or consortium) will be the first to implement the eHealthTrust model, resulting in significant competitive advantage in other markets as additional communities turn to this approach to solve the difficult HIE problems discussed earlier. The potential market for eHealthTrusts is very substantial — at $60/person/year, nearly $18 billion in annual revenue across the entire U.S. (1% of overall health care costs). With the expectation that such systems can reduce health care costs 8% or more, the potential benefits for health care stakeholders in communities are compelling.

    Furthermore, much valuable knowledge and experience will be gained in this first eHealthTrust. The actual availability of complete electronic medical information has never before been achieved in a community, and there are undoubtedly many key lessons that will be learned along the way to this exciting goal.

    Finally, the simplicity of the eHealthTrust model allows for a rapid timeline, which means early results — the Louisville system should be fully operational by January, 2007. So the experience gained will not only be important, but will be available soon for application elsewhere.


    The Louisville eHealthTrust project, along with a number of other demonstrations underway across the country, will make 2006 a very exciting year in the development of the NHII.

    Next time: A Framework for Measuring HIE Progress in Communities