Selecting the "Best" EMR; Problems with the "Greatest Good" Doctrine
Posted Mar 05 2012 12:00am
Here's an interesting excerpt from a note posted on Hospital EMR and EHR regarding the selection of an EMR (see: Stalking the “Perfect” EMR ). It quotes the noted hospital CIO, Dr. John Halamka.
...[I]f you let every department and clinical constituency pick what they want to include in their EMR, you end up with “an unintegrated melange of different products that make care standardization impossible,” Dr. [John] Halamka suggests. As nice as it would be to satisfy everyone, there’s really only one approach that works, Dr. Halamka says. IT leaders need to pick an EMR for their enterprise that meets the enterprises overall strategic goals, one “providing the greatest good for the greatest number.” Then, follow up with substantial training, education, collaboration, user engagement support and healthcare information exchange, he says. No matter what your EMR turns out to be, it’s going to fix some workflow and process issues while creating others, he suggests. The best thing healthcare CIOs can do is simply go with smart enterprise-wide technology and help providers user it effectively.
Here's the first comment reacting to this post from a reader named Brian:
The HIS is unlikely to satisfy the unique workflow needs of every department. Other industries such as banking have firmly settled on a best-of-breed model. Improved interoperability standards will probably lead to more BoB over time, not less.
I, like Brian, am a best-of-breed (BOB) advocate for most hospital departmental systems and particularly for LISs. Here are some links to some of my prior notes about this topic, a couple of them dating back to 2006 (see: Defining a "Best-of-Breed" LIS ; Best-of-Breed Lab Software Vs. Integrated Clinical Systems ; Are You an Enterprise or Best-of-Breed CIO? Access to Cash May Make the Difference ). In my opinion, it's impossible for a vendor selling an "enterprise wide solution" (EWS) to provide all of the requisite functionality for key hospital units like surgical pathology and clinical pathology. However, if I were a CIO like Dr. Halamka, I would probably also promote a "greatest good" doctrine for EMR selection. Such an approach makes his job easier in that most of the integration comes shrink-wrapped from the vendor. However, it can also be seriously detrimental to the career of a lab director to hold this same opinion and can impair lab functionality.
The directors of surgical pathology and the clinical labs in a hospital, in the final analysis, will be judged by hospital executives on the cost-per-test generated in the labs. Lab tests are relatively inexpensive because of the high degree of lab automation that has been achieved over the years. A highly functional LIS is a critical component of this scenario. A lab director may be required these days to choose between a BOB LIS plus often some additional software to fill in the gaps or an EWS lab solution like Beaker (see: Here Comes Epic's Beaker LIS -- Ready or Not ; More Information about Epic's Beaker LIS and Its Sibling ; A Pathologist Describes His Firsthand Experience with a Demo of Epic's Beaker LIS ). This latter LIS seems to have good core functionality but also large amounts of "white space," which is to say functionality gaps (see: Details about Epic's Beaker LIS, Supplied by the Company ). Beaker is thus not free nor does it provide an escape route from Dr. Halamka dreaded "unintegrated melange of different products that make care standardization impossible." In fact, picking an LIS like Beaker exacerbates the lab software integration problem because there may be more software products to be interfaced to the EMR in association with it than if a BOB LIS had been chosen or retained in the first place.