Title : Prevalence of antipsychotic-induced extrapyramidal symptoms in mental disorders using AI-powered data extraction: A meta-analysis and systematic review
Abstract:
This meta-analysis systematically examines the global prevalence of extrapyramidal symptoms (EPS) among patients with mental disorders taking antipsychotic medication. A comprehensive search of databases such as PubMed and Google Scholar was conducted for literature from 2018 to 2024 using predefined keywords. Observational studies reporting antipsychotic-induced EPS were included. Data extraction was performed using an artificial intelligence (AI) based retrieval-augmented generation (RAG) tool. Our proposed tool does not require any data-specific fine-tuning to extract data for meta-analysis studies. Meta-analysis was conducted using SPSS and R software. Subgroup analysis was performed based on geographical location. Nine studies, with 15,332 patients, met the criteria for inclusion. The pooled prevalence of EPS was 26.70%, with drug-induced parkinsonism being the most common (16.51%). Subgroup analysis revealed higher effect sizes in Asia (0.28) and Africa (0.27) compared to Europe and America (0.13). Our study shows that antipsychotic-induced EPS are prevalent globally, with regional population-specific variations. Drug-induced parkinsonism was the most frequent EPS. Further research on demographic and clinical factors influencing EPS is needed to optimize patient care.
