Agglomerative Clustering-Based Drugs Graded for the Risk Signals of Drug-Induced Cognitive Disorder: A Study on the FAERS Database

The study introduced an agglomerative clustering-based model (ACM) to evaluate drug-induced cognitive disorder (DICD) risk signals using the FAERS database, surpassing traditional methods like Lasso regression. The ACM effectively grouped drugs into clusters based on risk levels, identifying Finasteride as a high-risk drug and Carbidopa/Levodopa, Topiramate, and Clonazepam as moderate-risk, which Lasso regression missed. This model provides a more comprehensive assessment of drug-related risks, aiding clinicians in safer prescribing practices. However, it relies on FAERS data, which may have biases, and further clinical studies are needed to refine risk assessments.