Decision tree-based expert system for adverse drug reaction detection using fuzzy logic and genetic algorithm

Early detection of unknown adverse drug reactions (ADRs) could save patient lives and prevent unnecessary hospitalizations. Current surveillance systems are not ideal for rapidly identifying rare unknown ADRs. Current methods largely rely on passive spontaneous reports, which suffer from serious underreporting, latency, and inconsistent reporting. A more effective system is needed as the electronic patient records become more and more easily accessible in various health organizations such as hospitals, medical centers and insurance companies. These data provide a new source of information that has great potentials to detect ADR signals much earlier. In this paper, we have developed a methodology that uses both decision tree and fuzzy logic to generate a decision model. The developed model is equipped with a fuzzy inference engine, which enables it to find the causal relationship between a drug and a potential ADR. This could assist healthcare professionals to early detect previously unknown ADRs. Optimizing fuzzy rule weights and fuzzy sets parameters using genetic algorithm has been embedded in the proposed system to achieve excellent performance and improve the accuracy of the developed model. To evaluate the performance of the system, we have implemented the system using Weka and FuzzyJess software packages, and generated simulation results. To conduct the experiments, clinical information on 280 patients treated at the Detroit Veterans Affairs Medical Center was used. Two physicians on the team independently reviewed the experiment results. Kappa statistics show excellent agreement between the physicians and the developed model.