A process mining and text analysis approach to analyse the extent of polypharmacy in medical prescribing

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Colleges, School and Institutes

Abstract

Prescription of conflicting concurrent medications, polypharmacy, is recognised as a significant problem in the UK, but its extent is not fully known. We combined a process mining approach with text analytics, to discover prescription processes for patients from five primary care sites in the UK West
Midlands. Free-text prescription instructions were combined with online knowledge about drug interactions to reveal that almost 62% of patients were prescribed with medication with some level of interaction, during a two year period. We describe a novel domain-specific approach to reduce the complexity of the mined processes, which will nevertheless be applicable to other flexible environments such as knowledge work. We also highlight difficulties encountered in accessing, interpreting and processing the data, which may be significantly mitigated through wider adoption of a mindset of curating data for automated analysis.

Details

Original languageEnglish
Title of host publicationThe Sixth IEEE International Conference on Healthcare Informatics (ICHI 2018)
Publication statusPublished - 26 Jul 2018
EventThe Sixth IEEE International Conference on Healthcare Informatics - Doubletree, New York City, United States
Duration: 4 Jun 20187 Jun 2018
http://hpr.weill.cornell.edu/divisions/health_informatics/ieee_ichi.html

Conference

ConferenceThe Sixth IEEE International Conference on Healthcare Informatics
Abbreviated titleICHI 2018
CountryUnited States
CityNew York City
Period4/06/187/06/18
Internet address