Tuesday, March 30, 2010

Health IT Strategic Framework – Field of Dreams

Health IT is on a binge. Every day we are served with a new acronym, a new committee, a new contract, a new grant and a new goal. Health IT is definitely creating jobs, which was indeed one of the goals. But there was one other stated goal.

It was supposed to help physicians and patients provide and obtain better health care by making pertinent information available at the point of care, by involving patients and their families in their own care and by using computers to improve accuracy of clinical information. It was supposed to benefit patients in a very immediate and tangible way. That was the original meaning of the over used term “patient centric”. It was supposed to be all about the individual patient.

There were secondary goals as well. A byproduct of computerization of Medical Records is, of course, the ability to conduct research to benefit populations and advance the science of medicine, but somehow, in the frenzy of spending federal funds, the byproduct became the main objective.

I am a big fan of Dr. Blumenthal and in perpetual awe of his work at ONC. Today I read the recently published Health IT Strategic Framework. The ONC’s articulated vision, which drives the entire framework, is the creation of a “learning health system”. An excellent vision to be sure; a vision which drives a framework of security, privacy, biosurveillance, data collection public health, medical research and reporting. I have no doubt that in due course, such learning health system will produce evidence based information to support cost effective care and benefit future generations of Americans and humanity in general.

How do you build such a learning health system? It seems that ONC is tackling the task head on, by actually building a system from the ground up, complete with a National network (NHIN), localized intermediaries (HIE), individual access points (EHR), education centers (HITREC), standardization (Meaningful Use), certification (NIST, NVLAP, ONC-ATCB, ONC-ACB, ONC-AA), all driven by rulemakings and generous financial awards.
This is the “Field of Dreams” theory (If you build it, they will come), and it does work sometimes, as Kevin Costner can attest.

We could of course try a bottom up approach.

Case 1. 60 year old John Doe presents at the ED with chest pain. What does the attending physician need as far as information is concerned? A good current medications and allergies list and a good current diagnoses list and a little time to listen to John and take a good history of present illness. What does the attending physician not need? John’s growth charts from when he was a baby, John’s immunizations records from high school, John’s record of taking antibiotics for an STD while serving in the Navy 40 years ago, John’s record of taking other antibiotics for a sinus infection 5 years ago, etc. It would be nice if the attending doc could “get online” and download a couple of paragraphs of information from John’s PCP.

Case 2. 55 year old Mary Doe has Diabetes and Hypertension and CAD. She is sitting across from a Cardiologist who is seeing her for the first time. What does the Cardiologist need in order to evaluate Mary? Pretty much the same information the ER attending needed, plus any pertinent test results that her PCP might have ordered. And just like the ER attending, the Cardiologist needs a little time to listen to Mary. Most likely Mary’s PCP already faxed the information over. It would have been nice if the PCP information would “magically” end up in the Cardiologist’s chart and the Cardiologist consult note would end up in the PCP chart. It would also be nice if Mary’s daughter could “get online” and look at everything the doctors are doing, or proposing to do, because Mary may want to discuss her options with her daughter who lives 1000 miles away.

Turns out that patients and doctors can pitch a good game in the backyard with just a bat and ball. So maybe all we need is a few more bats and balls – Internet access, simple file transfer, CCR and some very simple game rules, so no one gets accidentally hurt…..  Let’s get all of us in the game before pretending to be Shoeless Joe Jackson.


  1. Sounds like you're making a distinction between a comprehensive-in-every-respect "medical record" and a relevant-right-now "health record". You're right -- we should not even dream about the former before we can do the latter. This is pretty much what the St. Louis Integrated Health Network is trying to do among the safety-net providers using the system built at Vanderbilt. Nobody's thought about giving the patients access though. At least I don't think so.


  2. Hmmm... I'll have to look at that. I'm glad St. Louis folks are doing the right thing...

  3. Yes, if an EHR is simply an electronic version of paper chart Case 1 will be very probable, and even more so thanks to HIE. But if we succeed in storing structured data, at least, for new records, by applying standard vocabularies at the data source and having all labs, reports and other types of clinical documents linked to respective encounters, we will be able to sort, categorize, group patient information in a number of ways, and even make this customizable.

  4. I agree Alexander. My only problem is that we are starting with the discrete data when most docs out there don't even use a computer.
    It could be beneficial, I think, to insert an intermediary step, that has enormous benefits and is not so outlandish.
    I noticed you are involved with NHIN Direct. Is that one of the goals there? Lower the technological barrier to entry, so most docs can start participating?

  5. NHIN Direct has been conceived as a lightweight alternative to the IHE XD* (mostly, XDS.b) HIE frameworks, especially, in situations where healthcare providers don't have sufficient knowledge or resourses to meet the frameworks requirements and open up a TLS port to communicate with document registries and repositories. The NHIN Direct framework will closely resemble the e-mail model, with multipart mixed-content MIME messages being exchanged between participants. There are some discussions as to whether to use the XD* document set metadata, hData structure, or keep it really simple (and separate from XD*). At the end of the day, it's all about the ease of implementation, learning time and need to exchange information with XD* adopters.

  6. Margalit, you and I are often in disagreement, but not this time. Medscribbler has ALWAYS been focused on making information collection, storing and retrieval better than paper (what most use now)We have avoided all the other stuff, like CCHIT, and things like a discrete data templating becuase they are USELESS and even HARMFUL in the process of helping "physicians and patients provide and obtain better health care by making pertinent information available at the point of care."

    In this we added things like an HL7 dynamic and static interface that can interact with any other software or device that has an HL7 interface. Everyone should have this then the secondary goals could be achieved, right? Unfortunately few are capable of interacting, why? MONEY, The vendors like GE,Eclinical etc plus billing companies that think they can do an EMR, plus labs companies that think they can cash in have no interest in actually allowing their program to interact through an HL7 because that would mean they would have to compete with ideas and function not marketing. Marketing is about MONEY, design is the antithesis of the "quick buck" The government is just fuelling the "quick buck" Medscribbler is in for the long haul of innovative design because we want the whole market not pieces we can get handouts for or our egos stroked!! Are sales are increasing exponentially now!

  7. AS a PS - because we have focused on design and innovation we actually are about to release discrete data templating that does not require pick lists and and /or a provisional Dx. Took two years of design and the release is still only rudimentary - but rudimentary only in the secondary goals NOT in helping doctors with patient care.

  8. Howdy Mike, great to see you here....

    Funny, I always thought we had more in common than one could tell from the few "disagreements" we had in the past.

    I think we both agree that simplicity is a virtue in this business (and many others too) and I think anybody that can bring a simple solution to a complex problem, usually caries the day.

    I sure hope it's Medscribbler. I never really saw it... I don't have a tablet just yet. Waiting for that Courier thing from Microsoft, if it's not just vaporware.