Direct Model or HIE Model

There’s a pretty fierce battle going on right now between all the various stakeholders interested in exchanging patient data. The stakeholders range from very large companies to government initiatives to startup companies. One of the major problems that I see is that it’s not completely clear which model of patient data exchange will win out. In fact, let’s not be surprised if a number of different options take hold.

With this said, I found it interesting that my favorite open source healthcare IT advocate, Fred Trotter, has chosen to get behind the Direct Project. In Fred’s post describing the challenges with the IHE-protocol HIE model approach is flawed and that the direct exchange of healthcare information is the way to go. In fact, he provides the following two illustrations in his post to show the difference:

HIE Model (click on the image to see it full size)

Direct Model (click on the image to see it full size)

Fred then offers this incredibly interesting conclusion:

At every level, organizations are deciding whether to invest in Direct or IHE-based exchange. At this point, I believe the only viable option is for a local exchange to either support Direct only, or both Direct and IHE. IHE is simply going to be too heavy weight for early adoption. Eventually, IHE may become dominate but for now Direct is much simpler, and puts the patient right in the center of everything. If you are a policy maker, you should be asking anyone involved with an HIE process to detail what their Direct-strategy is. If any effort is ignoring Direct and going with IHE-only I would lay odds that they will be broke and defunct before the decade is out.

Moreover, an IHE-only strategy is going to exclude direct participation from patients at this stage. If you care about patient empowerment, I recommend that you advocate for the Direct project at every level, including in your local HIE and REC.

Lots to consider with this complex challenge.

I guess you could say that the direct model is the patient centric model. Although, one could easily argue that the direct model doesn’t have the patient as the center of the model, but instead is a PHR centric model. So, the direct model will be a patient centered model only as much as the PHR software allows the patient to be involved.

Thus, it makes since why Microsoft HealthVault and Google Health are heavily involved in the Direct Project. Of course, they want to be involved in a project that puts them at the center of the communication.

The real question even with the direct model is what incentive do the various PHR vendors have to make this interaction happen? What will be the “cost” that PHR vendors pass on to consumers and/or doctors that use the PHR centric model? Basically, what’s the business model of the PHR vendors?

Unless we can find a PHR centric business model that works for the PHR vendor while still empowering the patient, even the direct model will fail or have adverse outcomes.

About the author

John Lynn

John Lynn

John Lynn is the Founder of, a network of leading Healthcare IT resources. The flagship blog, Healthcare IT Today, contains over 13,000 articles with over half of the articles written by John. These EMR and Healthcare IT related articles have been viewed over 20 million times.

John manages Healthcare IT Central, the leading career Health IT job board. He also organizes the first of its kind conference and community focused on healthcare marketing, Healthcare and IT Marketing Conference, and a healthcare IT conference,, focused on practical healthcare IT innovation. John is an advisor to multiple healthcare IT companies. John is highly involved in social media, and in addition to his blogs can be found on Twitter: @techguy.


  • There are two core aspects to HIE. The first, which is the mostly loudly and widely touted, is the “authorized-point-of-care-access-anytime-anywhere-24/7” sales pitch. All that “Use Case” business (why the simple word “example” doesn’t suffice escapes me).

    The second goes to the mining of clinical data (ostensibly “de-identified”) for “CER” leading to “EBM”/Practice Guidelines/Clinical Decision Support. To the Tinfoil Hat crowd, the latter means that Michelle Obama is comin’ after me with her celery sticks should my ONC-certified EHR auto-calculated BMI be “out of compliance.” (Which is why it gets comparatively less HIT policy cheerleading, and is only discussed at length within polite company.)

    CMS and health plans have been digitally mining administrative health care data for decades now for the most crude approximations of CER (I used to routinely run SAS and Stata code probes against Medicare Part-A claims data as part of my work during my prior two tenures with HealthInsight. We even bought larger set “All-Payer Data” from UNLV to do similar analyses stratified by payer types and other breakouts) .

    How, we might ask, does the “Direct Model” above facilitate CER? The “clinical-data-anywhere” patient-provider benefit is clear, but what of the national research goals, frightening as they are to those of Glenn Beck-istan?

  • “How, we might ask, does the “Direct Model” above facilitate CER?”
    They won’t. They facilitate patient care.

    However, the reasons you mention is why the HIE model will continue to go forward. The aggregated health data is far too valuable.

  • The HIE model will be chosen by the ACOs and big hospital groups for the value in the health data.

    However, the direct model would work well in the direct care or cash practices whose patients have chosen to stay out of the big medical systems.

  • The direct model seems like it would be difficult for providers to use. What happens when we have more vendors in the PHR field- would the provider have to be setup to access all of them (or receive data in all of the formats out there)? Do the labs have to be setup to send information to each of these PHRs, depending on which one their patient likes to use? It just doesn’t seem likely that every provider and every lab and ancillary provider would get setup to send\receive information from all of the PHRs out there. This looks to me like it really just rearranges the silos of information that we have now so that rather than having a silo with the original provider, we now have a silo with an outside application. How is this really better than old doctor sending the paper chart to new doctor?

  • Tiera,
    Those are good and interesting observations. The challenge I see with the Direct Model is that it puts the onus on the patient to get and provide the information. That will work great with some and not with others.

  • John-
    I agree that this will be useful for some and not for others. PHR centric models might work for people who are generally healthy, and technically savvy. However, that isn’t going to work for my grandma, or someone who is too sick to deal with being the primary organizer of their health information. It also doesn’t help at all in the scenario where a patient shows up at an ED unconscious or unresponsive and unable to give their medical history or grant access to their PHR. The ED will have little information in that scenario, and delays in treatment or negative treatment outcomes are more likely. I don’t think the PHR\direct model really achieves the goals of coordination of care, reducing waste, or increasing collaboration amongst caregivers. To me, it is a flashier incarnation of the old way of doing things- it doesn’t really change the healthcare delivery model. My impression is that the whole purpose of driving providers to use EMRs and achieve MU is to change the healthcare delivery model for increased efficiency and quality. While emergent care is only a small portion of the overall healthcare delivery system, it is one of the most impactful heathcare settings and one area that is highly impacted by timely and accurate patient information. I do believe that PHRs are useful, and I’ve seen some of your other articles out there about their potential. I’m not really convinced that they are the best way to to share patient information on a system-wide basis though.

Click here to post a comment