• More on standards

    I’m really enjoying the comments from my last post on Health IT. I could respond to many of them individually, but it’s probably worth it to batch them here.

    I know that there have been organically derived standards in many other areas. One commenter even brought up HTML, which has gone through a number of revolutions, and has reached a reasonable amount of consensus without any guiding hand from the government. But it’s important to remember that, with respect to the web, everyone wants you to see their stuff. Any company with a website wants their information to be as accessible as possible. When it comes to medical data, the opposite is often true. Each EMR company does not want to share data with others. Moreover, each hospital or practice doesn’t want to share its data. So there’s no impetus for most of the stakeholders to try and find a common solution.

    On the other hand, HTML shows exactly why having the standard works. Once the stuff is in a common format, many browsers have cropped up to compete for transmitting and presenting it. They are cheap, they have different pros and cons, and they are mostly interchangeable. I wish EMRs functioned like that.

    I’m also not some starry-eyed believer that we will have one true standard, that encompasses everything. Waiting for that will leave us immobile forever. But there should be some consensus on how certain data (demographics, past medical history, allergies, medications, labs, visits, hospitalizations, etc.) could be housed in a common format, or easily translated into that format. I think it’s reasonable that all EMRs be required to export and import data in that format in order to be certified for “meaningful use”. That alone would make sharing data across providers much easier. It would also promote competition, since the barrier to changing EMRs would be significantly reduced.

    • If you are correct that the users/buyers of the EMR systems don’t care if different systems can share, meaningful standards seem a long way off. I don’t get why hospitals and practices would not want to share information though

      • It’s in the people’s interest but against the provider’s (“provider” widely defined). If data can flow between provider networks it means patients can flow too. By making switching networks as onerous as possible (e.g. by needing a new intake history) they can increase their retention.

    • Your posts discuss IT standards for data formats and connectivity. Beyond these technical IT standards, however, health care needs standards of care for managing clinical information, somewhat analogous to accounting standards for managing financial information. The needed standards of care for managing clinical information (both medical knowledge and patient data) must inform design standards for electronic medical records and decision support tools.

      These concepts are the subject of a new book, entitled Medicine in Denial, published earlier this year by an Amazon subsidiary. See the book’s web page, http://www.createspace.com/3508751. The book’s lead author is my father, Dr. Lawrence L. Weed (LLW). For background on his work, see a December 2005 article from The Economist at http://www.economist.com/node/5269189?story_id=5269189. As that article indicates, LLW is one of the pioneers of computer use in medicine. His work goes far beyond that, however. He has reconceived the practice of medicine as a scientific discipline. In doing so, he has taken into account centuries-old standards of scientific behavior, and the new possibilities created by information technology.

      Readers of The Incidental Economist may be especially interested in part V of the book. That part examines the foundations of modern science laid by Francis Bacon 400 years ago, and related concepts of Karl Popper. Part V then ties these concepts to F. A. Hayek’s analysis of economy of knowledge in market systems, where people constantly avail themselves of knowledge they do not individually possess. In health care, people need tools to avail themselves of both medical knowledge they do not possess, and information processing power they do not possess. The tools require simple rules for managing clinical informa­tion, just as market systems require accounting standards for managing financial information. This argument views health care as a complex adaptive system, but one that lags centuries behind the evolved systems in the domains of science and commerce.

      Compare the December 2010 PCAST Report to the President on Realizing the Potential of Health IT, at . The PCAST Report (p. 38) asserts that “any attempt to create a national health IT ecosystem based on standardized record formats is doomed to failure.” This assertion completely ignores the fundamental problems created by the unsound designs of current records, both paper and electronic. It also ignores the need to improve practitioner inputs to medical records, using integrated decision support tools, based on standards of care for choice, collection and analysis of patient data in light of medical knowledge. For further discussion see our response to the PCAST report at http://www.regulations.gov/#!documentDetail;D=HHS-OS-2010-0030-0099.