Introduction

Improving antibiotic therapy and reducing antimicrobial resistance

The clinical challenge

Despite half a century of antibiotic use, re-emerging and new infectious diseases - partially caused by the rise of antimicrobial resistance - have become again life-threatening and costly medical challenges. The EU-funded DebugIT research project has developed an IT-framework to allow health care systems to better address these newly emergent problems and improve their management.

The DebugIT response: integrating multi-site data-mining results in clinical monitoring and decision-support systems

To meet this challenge, DebugIT has developed a revolutionary approach which allows for detecting patient safety related patterns and trends through advanced data mining of data collected and stored in electronic Clinical Information Systems (CIS). It relies on (1) a semantic interoperability platform which allows the federation of patient or other data from different CIS and (2) a number of ontologies to describe the clinical domain under consideration. Results of data-mining are stored in a Knowledge Repository, which also includes existing guidelines and expert opinions. For the decision support workflow relevant medical knowledge is used which has already been collected and stored in a Knowledge Repository. The new knowledge gained in DebugIT is also applied to the monitoring of ongoing care activities and outcomes, and may help to predict future outcomes to give additional support to treatment decision on individual patients and for populations.

The DebugIT process begins with the extraction of microbial case data from Clinical Information Systems which are then mapped to a common domain ontology, concerning microbes and infection control. The resulting clinically-derived microbe data, aggregated across the different sites and harmonized with a common ontology, can be seen as coming from a global but virtual Clinical Data Repository.This is accomplished by building a semantic layer (SPARQL endpoint) on top of the existing systems.

Once available, the data is analysed by applying several statistical and data mining techniques. The resulting knowledge is then strored in a knowledge repository, ready for display in a monitoring dashboard, or to be used for decision support.. This monitoring dashboard is implemented as a stand alone web application or can be integrated in the Clinical Information Systems of hospitals.