Karsten Hilbert wrote: > > John, > > >...clip... > > Clinicians simply do *not* perform these sorts of data mining > > activities in their practice. > "Give me the last mention of tetanus immunisation > status anywhere in this patient's EMR." i.e. "Give me the most > recent occurrence of 'tet.' or 'Td'." Admittedly this can be > done with grep, too.
Umm... I'm experiencing a little cognitive dissonance here. I am a clinician (1). I work in a clinic (2) of a couple dozen docs. Despite (or because of) my best efforts, we do not have EMR (though we can access all dictation and transcribed notes electronically), so I think this applies to my practice. 1: I do data mining daily with patients, mostly regarding labs. I flow-chart -- on paper or in my head -- labs continually, on any patient with a chronic condition that requires monitoring. PSA, lipid profile, creatinine, potassium, sodium, hemoglobin, white cell types, platelets, glucose, HgbA1c, sedimentation rate, urine protein, immunoglobin level -- these are only the most common. 2: Our practice does data mining continually. We have 2 RN's who together spend almost "full-time" activity on pulling random charts, looking for data on quality measures, and reporting. We study diabetes-management practices, hypertension treatment and efficacy, prenatal care, immunization practices in adults and children, and well-child care routinely and review other data as well. Mayo Rochester, Jacksonville, and Scottsdale have EMR, but "Mayo System" sites around Wisconsin, Minnesota, and Iowa do not. The main benefits of the long-promised system-wide deployment of EMR to our smaller clinics will be more efficient quality review and easier data sharing among sites. I don't think physicians are expecting patient visits to suddenly become easier or shorter when this invades our exam rooms. I am not an IT technician, so I don't know what the feasibilities are ...it seems as though, if the goal is to have the front end of the EMR to be browser-enabled, that for some tasks (e.g. data presentation), the solution is to have a sub-process, perhaps a function of the database system, create an html table ,or .pdf file that can be viewed directly by the browser or a pdf viewer via a browser plugin, or a spreadsheet file that can be viewed by a helper application such as Applix or OpenOffice. I.e., I don't understand John's apparent need for browser singularity. As a clinician I don't care what the mechanisms are that get the information in wonderful format, as long as it's not inefficient and cumbersome (organizationally, technically, financially) to create or operate the mechanisms. It is not difficult, on a current open-source desktop, to quickly view a wide variety of file formats. Or am I missing some important technical truth(s)? Happy New Year to all! DJohnson
