This post is directed to the comment from my previous dialogue on Medical Home and Interoperability as a deeper explanation to encourage a continuing discussion.
In response to a real world example, diabetes mellitus serves as an excellent model for the need for semantic integration and population-based decision support.
Simple problems, such as comparing blood work from differing lab systems can be troubling. Comparing “blood sugar” to “serum glucose” to “glucometer” for instance; or “HBA1c” to “glycohemoglobin” to “glycosylated Hb” to “hemoglobin A1c” is in itself a daunting task. Let alone the myriad of other tests that need to be directly and indirectly compared.
What about the hurdles of simply identifying new patients in the diabetic population. Take for instance a patient who has a history of gestational diabetes diagnosed during a remote pregnancy. This information might only be present in the obstetrician’s office records or in the labor and delivery information system. If this patient were to have an eye exam years later, which demonstrated retinopathy, clearly we would like to be able to “connect the dots.” The problem, of course, is that the decision making is obfuscated by information silos and translational difficulties that result from differing medical vocabularies.
More difficulty arises from identifying negations of care. It is easier to use information technology to see if a patient has satisfied a particular quality measure, but providing a list of diabetics who “haven’t had a urine microalbumin level checked in 6 months” is technically challenging on many levels.
I believe that the times cry out desperately for improving diabetic care through the use of this kind of advanced technology.
The now famous UKPDS study showed that for every one point drop in a diabetic’s HbA1c level, there was upwards of a 14% decline in morbidity (i.e., kidney disease, blindness, loss of limb, stroke, etc.). It seems obvious to me that we must direct our most intensive efforts on treating those outliers with poorly-controlled disease, if we are ever to have any hope for a healthier, and thus less economically burdensome, population.
As a primary care physician, I am passionate about this subject, and I know my dedicated colleagues compassionately yearn to help this population. What are needed are the proper tools to accomplish this. I submit that a semantically organized interoperability platform, coupled with advanced population surveillance technology is a critical, yet difficult and poorly understood component.
