An exploratory machine learning workflow for the analysis of adverse events from clinical trials
Abstract: A new pharmaceutical drug needs to be shown to be safe and effective before it canbe used to treat patients. Adverse events (AEs) are potential side-effects that arerecorded during clinical trials, in which a new drug is tested in humans, and mayor may not be related to the drug under study. The large diversity of AEs andthe often low incidence of each AE reported during clinical trials makes traditionalstatistical testing challenging due to problems with multiple testing and insufficientpower. Therefore, analysis of AEs from clinical trials currently relies mainly onmanual review of descriptive statistics. The aim of this thesis was to develop anexploratory machine learning approach for the objective analysis of AEs in twosteps, where possibly drug-related AEs are identified in the first step and patientsubgroups potentially having an increased risk of experiencing a particular drug sideeffectare identified in the second step. Using clinical trial data from a drug witha well-characterized safety profile, the machine learning methodology demonstratedhigh sensitivity in identifying drug-related AEs and correctly classified several AEsas being linked to the underlying disease. Furthermore, in the second step of theanalysis, the model suggested factors that could be associated with an increased riskof experiencing a particular side-effect, however a number of these factors appearedto be general risk factors for developing the AE independent of treatment. As themethod only identifies associations, the results should be considered hypothesisgenerating.The exploratory machine learning workflow developed in this thesiscould serve as a complementary tool which could help guide subsequent manualanalysis of AEs, but requires further validation before being put into practice.
AT THIS PAGE YOU CAN DOWNLOAD THE WHOLE ESSAY. (follow the link to the next page)