Tuesday, December 11, 2007

Standards for Personal Health Records

In my post about Personal Health records , I identified the 4 major types of Personal Health records - provider-hosted, payer-based, employer-sponsored and commercial. As more products are offered, it's key that all the stakeholders involved embrace national healthcare data standards to ensure interoperability of the data placed in personal health records.

To illustrate the point, I am posting my entire lifelong medical record on my blog (this is with my consent, so there are no HIPAA issues) in two ways.

The first is a PDF which was exported from a leading electronic health record system. It's 77 pages long and contains a mixture of clinical data, administrative data, normal and abnormal results, numeric observations, and notes. It's a great deal of data, but is very challenging to understand, since it does not provide an organized view of the key elements a clinician needs to provide me ongoing care. It is not semantically interoperable, which means that it cannot be read by computers to offer me or my doctors the decision support that will improve my care.

The second is a Continuity of Care Document , using the national Health Information Technology Standards Panel (HITSP) interoperability specifications. It uses "Web 2.0" approaches, is XML based, machine and human readable, and uses controlled vocabularies enabling computer-based decision support.

It's critical that Vendors, Payers, Providers and Employers embrace these standards. A standards-based personal health record can be used to prevent medication errors, ensure best practice disease prevention, and serve as the basis for decision support systems which recommend optimal care. Using CCD, data can be turned into wisdom , can be incorporated into EHRs, transmitted between PHRs, and can be easily expanded by the patient throughout life.

Today (December 13), HITSP will deliver the harmonized standards for Personal Health Records, Labs, Emergency Records, and Quality measurement to HHS Secretary Leavitt. These "interoperability specifications" will become part of Federal contacting language and be incorporated into vendor system certification criteria (CCHIT) over the next two years.

12 comments:

  1. Hi, John.

    I am attempting to get my EMR so I can upload it to Microsoft HealthVault and "kick the tires," on that offering.

    Although my healthcare system (which is also my employer) has a proprietary EMR system and a report for patients who want access to its contents, they're still working in a paper paradigm; I can only get a print-out or maybe a PDF.

    Based on your post, I'm curious if you think that the standards being issued today will remain specifications for interoperability, or are the first steps towards a national EMR standard that will create a standard upon which all EMRs/PHRs will be required by law to be based upon?

    I think it is an important distinction, but am not as knowledgeable about the topic as you are to be able to "read the tea leaves" and see which direction the current administration is moving in.

    Also, what's HITSP relationship to HL7? Will we see an internationally recognized standard for either communicating or maintaining these records emerge in the near future?

    ReplyDelete
  2. HL7, NCPDP, X12, OASIS, are all members of HITSP, so their standards are part of HITSP's deliverable to Secretary Leavitt.

    These standards refer to data exchanged between systems i.e. EHRs and PHRs and now how data is stored within an EHR

    ReplyDelete
  3. Great post again John. This is a topic that will only grow in importance and controversy.

    I linked to your earlier piece on Health Records today in the Science section of The Issue.

    ReplyDelete
  4. Are you aware that your HPHC member number is in that PDF, along with your MRN? Is that bogus data, or do you invite us to break into your accounts?

    ReplyDelete
  5. In this age of identity theft, you have posted quite a bit of personal data.

    Do you currently have a subscription to an identity protection service?

    If so, would you mind telling us which service.

    ReplyDelete
  6. John, is "Vegan" really something that belongs on a problem list?

    ReplyDelete
  7. Here's a comment emailed to me by Professor Peter Szolovits at MIT. Since the CCD supports narrative, we could have included the extra details he suggests below. Regarding my identifiers, my implanted RFID emits my unencrypted identifiers when I'm scanned, so I have to accept that my medical identifiers are public information. I'll post an article about that soon.

    "Pete Szolovits wrote:
    On a more technical level, I just spent a few minutes comparing
    > John's 2-page CCD record with the 77-page dump from his primary care
    > practice. The CCD is clearly much more well organized (despite the
    > lack of a slot for dietary constraints, hence the odd "active
    > problem"). However, I wonder if there isn't also a great deal of
    > potentially interesting data lost in that summary. For example, here
    > are just a few items that are obvious from the full record, even on
    > casual perusal, that are not in the CCD:
    >
    > 1. He was at one time much heavier, with worse lipids, BMI, etc.
    > Presumably he became a vegan to overcome these problems. Similarly,
    > he seems to have cut out caffeine because of arrhythmias.
    >
    > 2. His allergy to Amoxicillin is documented in the 2002 note about
    > treatment of his Lyme disease, but is simply stated in the CCD. I
    > know that poorly documented drug allergies often cause people
    > problems.
    >
    > 3. His immunization records only go back to the 2000's, so we don't
    > have a record (in either form) of childhood immunizations.
    >
    > 4. There is a note about his foreign travel indicating that he was
    > told to take various meds as needed if he ran into diarrhea, rashes,
    > etc. There is no hint of this in the CCD, and we have no patient-
    > entered data about whether he actually took any of these.
    >
    > I suspect that a more detailed examination would uncover lots of
    > other potentially interesting data that are in the 77 pages but not
    > in the CCD. For example, the CCD's vitals and labs only cover the
    > last 3 years, whereas older data are in the longer record. Are those
    > omissions clinically important?
    >
    > --Pete Sz."

    ReplyDelete
  8. John

    I'd love your perspective on HIPAA at some stage.

    Within the last year I had an eye exam at a Boston clinic. I asked whether I could take a copy of the dilated eye pictures home with me. Mostly this was so I could show my children. I figured they'd find them interesting (we've homeschooled in the past).

    Imagine my surprise when I was told that I'd only be allowed paper copies of the pictures.

    I thought HIPAA meant that I have control over what happens with my data. And I decide I want to take it with me and post it on the internet (as you've done), that's my freedom and right.

    If you've insight into this, or can point to a useful explanation I'd appreciate it.

    ReplyDelete
  9. HIPAA guarantees that patients have access to their records, but does not specify the method of access. Many hospitals and caregivers only share paper. I believe that will change as standards are more widely adopted and patients demand electronic access to their data.

    ReplyDelete
  10. John, in your example you use both a standard PDF format and the second, the CCD standard. But why not use the PDF Healthcare standard? As I am sure you're aware, this standard, if one can call it that, is currently being balloted and will leverage both the CCR standard within the PDF construct whereby the PDF becomes a container, with XML clinicals that can be extracted to populate a physician's EMR. This seems to be the best of both worlds - the ubiquitous PDF and the DRM tools that surround it with healthcare standards embedded.

    So, the big question is: Why did you not try this as well?

    And since you have been so public with your own experience, trust you won't mind my putting a link to your post in one I'll be putting up shortly over at www.chilmarkresearch.com

    ReplyDelete
  11. I am trying to find XML schema formats for HIPAA transactions such as X12 270 and 271, X12 276/277/278/835/837 etc. What would be a good place?

    Thanks,
    Pingala

    ReplyDelete
  12. Within the last year I had an eye exam at a Boston clinic. I asked whether I could take a copy of the dilated eye pictures home with me. Mostly this was so I could show my children. I figured they'd find them interesting (we've homeschooled in the past).

    Imagine my surprise when I was told that I'd only be allowed paper copies of the pictures.


    Recep Deniz MD

    DoktorTR.Net

    ReplyDelete