Patterns extraction for monitoring medical practices Conference Paper uri icon


  • We present in this paper the extraction of relational patterns on a hospital discharge database to monitor for quality of care. We delineate the requirements set for the extraction of meaningful medical practices from the data and describe how relational patterns address these requirements. Patterns representing practices regarding hospitalization instances are comprehensive and easy to interpret — making them useful for quality management decision-making. We demonstrate how relational patterns can be applied to identify poor practices embedded in hospitalization processes and trigger subsequent inquiries.

publication date

  • January 1, 2001