Clinical Decision Support Using Antimicrobial Susceptibility Test Results

摘要

Prescribing the proper antibiotic against an infectious agent is crucial not only for the patient's health but also for the community, due to fast development of new resistances in bacteria. To ensure the efficacy of the treatment, laboratory tests are performed on cultures of samples obtained from the patient, analysing the resistance patterns of the infectious microorganism against some antibiotics. In order to assist clinical microbiologists, the European Committee on Antimicrobial Susceptibility Testing proposes a catalogue of clinical rules to identify resistance patterns and clinical recommendations from the results of previous antibiotic susceptibility tests. The aim of our proposal is to automatise and evaluate this source of biomedical knowledge. To this end, we have implemented a knowledge module combining ontologies and production rules. This module was included in a clinical decision support system and evaluated using test results of 365 days. After its execution, 20.9 % of the final antibiograms were new resistance patterns not covered by the laboratory tests, having 44 % of them a high clinical evidence grade. These results could help clinicians to prescribe more efficient treatments against infections in the future.