Information Technology Reference
In-Depth Information
This is a typical scenario when a physician network has contracted to provide
care under a pay-for-performance approach. Common quality metrics such as
tobacco cessation, adult obesity, hypertension screening and childhood immuniza-
tion status are collected, aggregated and reported in a very simple, easy to under-
stand format.
The system can also provide more focused and detailed analysis at the provider
or the patient level.
Medicity
is based in part in Atlanta which has one of the largest clusters of health
informatics companies in the US. When I first came to Georgia Tech I made it a
personal goal to try to identify and become familiar with the most innovative
Atlanta-based health informatics companies, particularly with respect to chronic
disease management and patient-centered care. One of the first companies I “dis-
covered” was Novo Innovations. One of its co-founders, Robert Connelly, was even
a Georgia Tech graduate (the other was Alok Mathur). They became interested in
the problem of interoperability while working for Atlanta-based HBOC (now
McKesson Provider Technologies), the largest health informatics company in the
country. There they developed an early physician portal, a system to give an inte-
grated view of care to community based providers around a hospital or a hospital-
based health system. Such a system would typically consist of a hospital and
associated provider practices that might be owned by the hospital but, in those days,
would typically be independent and would have their own information systems that,
back then, would usually not include an EMR.
As is often the case in large companies, they got an idea for a novel interoperabil-
ity solution and, finding little interest on the part of their employer, left to start their
own company. The idea is a bit technical but it is based on “intelligent agents”,
software that can reside on the same server as an EHR or other clinical system
(a laboratory system, for example) and is sufficiently “aware” of that system that it
can observe its data flows (often in HL7 specified formats). The best way to explain
what happens next is to give a highly simplified example. Dr. Smith has ordered
some blood chemistries for her patient, Mrs. Jones. The agent running on the labo-
ratory system and the agent running on Dr. Smith's EHR can communicate and
cooperate. As a result, each knows who Mrs. Jones is in their associated system
(remember, she probably will have different identifying numbers in each of these
systems). When the agent observing the laboratory system's data flows sees a result
for a test ordered by Dr. Smith it “grabs” it and sends it, along with some informa-
tion identifying the patient, to the agent in Dr. Smith's office. That agent figures out
whose result it is and puts it into the right place in her EHR.
If you've been following along since earlier, you'll recognize that this is an
example both of interoperability - two diverse systems sharing data - and process
re-engineering - something once done manually now being automated, saving time,
expense and potentially reducing errors. The agents are “intelligent” with respect to
the system they are associated with. As a result, they are automating previously
manual processes. Unlike the much simpler DIRECT approach to information
