Well, I guess I am on a Microsoft bashing streak (the folks at the Leapfrog Group will appreciate the repreive). I read Dr. Bill Crouse’s blog with regularity and generally find it to be quite good. But this post makes me crazy.
Dr. Crouse is employed by Microsoft. In this post he blogs about a a Micorsoft partner, HealthUnity, that has developed something called a RHIO In A Box. The name alone makes me crazy. Anyone that has tried to exchange data between healthcare providers knows this is a huge challenge that cannot be solved with software or technology alone. All of the problems in creating a RHIO are outside the box. Posts like this only serve to make CEOs think that creating a RHIO is as easy as buying software.
I picked up this marketing gem from the vendor’s web site:
“A hospital that wants to tap into the HealthUnity™ network can do so using our low-cost, plug and play solution. Once deployed, this solution will enable your physicians to collaborate with internal as well as external parties over secure and HIPAA compliant mechanisms. We provide all the services required to maintain smooth operation of the system. This keeps your IT investments minimal while increasing the returns on your investment.”
When I got to “plug and play” I began to involuntarily heave.
Neither the HealthUnity web site, nor Dr. Crouse’s post, address the complexities of patient matching. For example, how does the RHIO know that John Smith with Medical Record Number 123 is also Jack Smith in another provider’s database with another medical record number?
Then we have the fact that most physician offices don’t collect allergies in a codified manner. In fact, many don’t even have the free form text online. The solution to that problem is not going to be in a box.
Dr. Crouse points out that the product has the ability to exclude parts of a patient’s chart from automated access, yet make it available for case by case requests (e.g. an AIDS test result). Driven by patient privacy and state laws. How does the RHIO In A Box know which of the lab tests are for AIDS tests? While there are some emerging standards for univerally coding lab tests (i.e., LOINC) less than 5% of health care providers use them. How does the box know what my hosipital lab system calls an AIDS test?
My mission in writing this blog is to add some realism into the healthcare IT debate. Healthcare IT has over-promsied and under-delivered. We have been selling unrealistic dreams. These are complex topics. To pretend they are not is the worse thing we can do to forward the IT cause in healthcare.