Automated recognition of clinical guidelines from electronic medical records Academic Article uri icon


  • Abstract Clinical guidelines are common in medical domains for prescribing a set of rules and policies that a physician should follow. To automate the runtime application of a guideline, the retrospective performance of quality assessment, or clinical research, we often need to determine by which guideline or guidelines the patient was treated (eg, in order to determine the degree of adherence to the guideline specification and the degree of possible match between the medical record and each guideline). In this paper we present a methodology for automatically generating a degree of possible match between a given set of medical records and each guideline within a given set of guidelines. We are currently evaluating our methodology, using a set of guidelines for management of hypertension and a set of medical records.

publication date

  • January 1, 2004