Bubble

From time to time, I can't resist spouting off on where the overall
health care IT market is going. Two items really tickled my fancy in the latest post at HIStalk. Enjoy.

Investor's Business Daily is rah-rah
on healthcare IT (and heavy on the pun-slinging like HDM.) The[y] point
out the wonderful work being done by Sun and IBM on infrastructure
projects. Here's a breezy summary:
"There's
no national framework yet for such a system, but a number of regional
projects are working toward that goal. Several large hospitals have
already set up systems that let patients see lab results and other
records online. But digitizing records is the easy part. Setting up a
universal network that links all parts of the health care system
together — and standardizes medical information between all of them —
is a much tougher task."

Digitizing records is the easy part? Check out the quantity and quality
of data available for those fancy networks to share. Hospitals can't
even generate enough real information for themselves to monitor
outcomes, costs, market share, and physician performance. Existing
systems and optimal use of them will have to improve a lot to make
shared data useful, although we're lucky that in healthcare,
e-prescribing and e-mailing of PACS images is still seen as cutting
edge.

Mr. HIStalk's Guide to RHIOs for CIOs, in [10] easy steps:

  1. Participate
    eagerly in meetings with your competitors as they talk about sharing
    data, making sure to speak in generalities and offering no
    participation beyond attending more meetings.
  2. Go
    back to your co-workers and sneer at how stupid your competitors are
    and how backward their systems are compared to yours.
  3. Report
    back to your fellow VPs that there's really nothing in it for your
    organization, but that you'll listen politely and avoid all commitments
    just so no one gets mad.
  4. Respond
    to latest of 100 surveys asking about RHIOs, making sure to wax poetic
    about the wonderful possibilities that will result from the electronic
    hand-holding that RHIOs will bring, knowing full well it won't happen
    until benefits are offered to those involved.
  5. Go
    back to your co-workers and make fun of all the folks who've forgotten
    CHINs and therefore are doomed to repeat history.
  6. Ignore sales pitches from vendor RHIO participants who got involved only to troll for new business.
  7. Go
    back to your co-workers and make fun of the IT organization of the
    local IPA or medical society, consisting of one doctor's brother-in-law
    armed with an AA degree and an A-Plus certificate.
  8. Prioritize
    your IT shop's involvement in RHIO work somewhere between "get rid of
    all the cubes and give everyone an office" and "seriously consider
    moving all desktops and servers to open source operating systems."
  9. Go
    back to your co-workers and explain to the bright-eyed among them who
    ask about RHIOs that it's "no margin, no mission" and that it will be a
    cold day in hell when you voluntarily share your exquisitely created
    and managed information with the clueless barbarians across town with
    their pathetic IT systems.
  10. Get
    on the speaking circuit and HIMSS advocacy groups to make sure your
    attendance at RHIO meetings is rewarded with industry visibility as a
    RHIO thought leader.

The point Mr HIStalk has hammered relentlessly is that hospitals must
reach a certain level of data management in order to have information
to share - and the great majority of hospitals have yet to achieve a
level of automation to make RHIOs and their ilk worthwhile. Focused on
IT and medical device integration at the point-of-care as I am, I
couldn't agree more. If we want to reduce the cost of health care and
improve communications, the industry should focus on replacing all
those phone calls and faxes that serve as the principle means of
communications and coordination between physicians, payors and
hospitals. The savings gained from automating eligibility look-ups,
referrals, and pre-certifications would be huge - on the scale touted
for EHRs. Most claims these days are submitted electronically. Sadly, a
minority of these claims are
"auto adjudicated," i.e., processed by computer. Most claims must be
reviewed by an individual who makes and receives numerous phone calls
and faxes between the payor and provider to straighten things out. Add
some software (and maybe some government mandated standardization) to
enforce data quality at the point of submission for claims and national
health care savings would far exceed EHR benefits. Of course all of
this is boring and mundane, no human interest like with those sexy
longitudinal EHRs and RHIOs.

The health care market has had its fads in the past; I remember those 6
years that were "the year of the LAN"in the eighties, the time when
hospitals wanted to own physician practices, the HIT portion of the
Internet bubble, and now the electronic health record and RHIOs. Health
care is a funny market; peculiar funny, not ha-ha funny. Those of us
who've been in health care for years understand this. One reason is
that health care is the world's largest cottage industry. There is no
meaningful supply chain linking physicians, payors and hospitals; each
entity has one set of people they deliver services to and another set
that pays for those services - and they differ from physician to payor
to hospital. Politicians, bureaucrats, large multi national
corporations without significant long term health care revenues, think
of health care like manufacturing or the retail trade - industries that
respond to market and technological stimuli in a rational and timely
manner. That's not health care. Maybe someone can offer some similar
kinds of fads that have swept through other industries, and what it
might mean for us in health care.

UPDATE: Shahid, The Healthcare IT Guy, also opined on the same IBD article as HIStalk. Shahid's take is somewhat different, noting additional differences between the health care industry and industries outside health care.

Last I checked, I don’t really have a unified financial record
anywhere. When I need to do business with Bank of America, they don’t
really ask me for a record of my dealings with my Navy Federal Credit
Union. When my CPA does my taxes, he doesn’t have all my financial
records in one place that he can just peruse through and give me great
advice. When I shopped at CompUSA last week they didn’t have the
ability to tell me that the new USB thumbdrive I bought was already
sitting in my closet because I bought one just like it last year. I
have to manage my own financial records.

So, please,
spare us the comparisons with retail and banking. When you’re dealing
with a complex ecosystem like healthcare you can’t just assume that
putting networks in place, tossing in computers, adding some smart
engineers, and shaking it up a bit will make a great health IT
milkshake. The problems in healthcare have nothing to do with
technology, they have to do with the complexity of a multi-payer,
decentralized healthcare system. Competing hospitals need to share
medical records, physicians (small businesses) have to share orders,
pharmacies (suppliers) have to share clinical records.

Great examples. Read the whole thing.

The original post on HIStalk is also getting some good comments that are worth a look. Here's my favorite from Anony-Mouse:

I nearly soiled myself with the RHIO top 10 list.
We have a set of "very important" discussion
meetings coming up. Who are we kidding? What has
changed since the days of the CHIN concept? Let's
see:

  1. Web services/XML are present (but XML without a
    data model is garbage) and every healthcare
    organization is building their own
    service-oriented architectures as we speak (or at
    least the CIOs will say they are to be da'bomb)
  2. HL7 v3.0 which is being adopted at blazing
    speeds (NOT QUITE)
  3. 10 more organizations are live on comprehensive
    EMRs with order entry, all areas entering data
    (nurses, physicians, techs, therapists, etc.) -
    hey another pre-requisite that is being adopted
    quickly
  4. Every "well-constructed" vendor system
    is compliant with the most "well-adopted"
    controlled vocabularies known to man (wait, I am
    just coming off the Nitrous and realized what I
    said... oops - scratch that)
  5.  Every individual entering data into the
    clinical systems has been trained to understand
    the meanings of the vocabularies so it is really
    standardized (oops... scratch that one too)
  6.  The business model for initial and continual
    funding has been worked-out (damn, missed on that
    one too)
  7.  The oversight models are well-established and
    take into account that there is still competition
    in healthcare (wait, how many of these things are
    live/working and really doing some good and where
    is the Bizarro world where all the competition
    wants to share data with their competitors)
  8. There is a published meta-model and data
    pointer recovery standard that is well adopted and
    implemented (oh yeah, who has a HL7 Query message
    live?)

Let's start with the basics first, get clinical
data repositories live, define a workable meta
model, define a standard security model and use
case, figure out how to finance this venture and
then let's talk!