Answers are grounded in this paper and cite their sources.
This study evaluates ASReview, an open-source machine learning application for study selection in systematic literature reviews, using data from a review of whether screening for sexually transmitted infections reduces the incidence of pelvic inflammatory disease. A systematic literature review was conducted in accordance with the PRISMA guidelines, and manual screening produced a fully labeled dataset that served as the reference standard. ASReview was configured with four machine learning classifiers (Naïve Bayes, Random Forest, Support Vector Machines, and Logistic Regression) and two feature extraction methods (TF-IDF and Doc2Vec). Simulation experiments assessed screening efficiency under sampling-based and heuristic stopping rules. The systematic review suggested that annual screening for sexually transmitted infections may reduce the incidence of pelvic inflammatory disease by up to 40% compared with routine practice, although the evidence base was limited. In the simulation experiments, Naïve Bayes with TF-IDF achieved the highest recall and screening efficiency, particularly in datasets with a low prevalence of relevant records. Conservative stopping rules increased the likelihood of complete retrieval but required greater screening effort. Overall, these findings highlight the limited and heterogeneous evidence on sexually transmitted infection screening for pelvic inflammatory disease prevention and show that ASReview may improve the efficiency of study selection when evaluated within a real systematic review workflow.
No takes yet. Share an insight, caveat, or question.
Guadarrama-Atrizco et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: