You have accessJournal of UrologyInfertility: Epidemiology & Evaluation I (MP21)1 Sep 2021MP21-09 TOPIC MODELING AND SENTIMENT ANALYSIS OF SOCIAL MEDIA MALE INFERTILITY DISCUSSIONS Krishna Ravivarapu, Alexander Small, Evan Garden, Micah Levy, Osama Al-Alao, and Michael Palese Krishna RavivarapuKrishna Ravivarapu More articles by this author , Alexander SmallAlexander Small More articles by this author , Evan GardenEvan Garden More articles by this author , Micah LevyMicah Levy More articles by this author , Osama Al-AlaoOsama Al-Alao More articles by this author , and Michael PaleseMichael Palese More articles by this author View All Author Informationhttps://doi.org/10.1097/JU.0000000000002006.09AboutPDF ToolsAdd to favoritesDownload CitationsTrack CitationsPermissionsReprints ShareFacebookLinked InTwitterEmail Abstract INTRODUCTION AND OBJECTIVE: Anonymous social media forums allow uninhibited and unbiased patient expression of prevailing concerns, attitudes, and experiences. For a sensitive condition like male-factor infertility, insights into online discussions can augment understanding of patient concerns. We used quantitative natural language processing (NLP) methods to identify key themes on one of the internet’s largest male infertility forums. METHODS: We extracted three years of posts (08/2016-08/2019) and comments (01/2017-12/2019) from the 3,063 member Reddit community r/maleinfertility. We employed a meaning extraction method followed by principal component analysis (MEM/PCA) to computationally distill major themes of discussion and compare the relative prevalence of discussion topics. Kaiser-Meyer-Olkin Measure of Sampling Adequacy (KMO>0.60) and Bartlett’s test of sphericity (p<0.001) indicated that the data is suitable for MEM/PCA. Furthermore, we calculated a sentiment score using the Valence Aware Dictionary for Sentiment Reasoning (VADER) model. RESULTS: 4,609 posts and comments were analyzed with MEM/PCA to generate five major themes of discussion ranked by prevalence (Figure 1). Posts are made to clarify specific semen analysis parameters (count 11.3%, motility 8.2%, morphology 6.4%). IVF 13.1% is mentioned more frequently than ICSI 5.4% or IUI 5.3% (p<0.001). Treatment effectiveness (chance 5.2%, luck 7.9%) and cost (expensive 1.7%) are discussed concurrently. Low testosterone (7.1%) is a major concern, leading to discussion of Clomid (4.6%), HCG (2.2%) and TRT (1.2%). Infertility also takes a toll on mental health with frequent mentions of concern 6.7%, worry 3.0%, and stress 2.1%. Suggestions for increasing fertility include lifestyle changes (vitamins 3.8%, diet 2.3%, exercise 2.0%, weight 2.0%), using ice (2.1%) to cool testicles and wearing loose clothes (1.2%). VADER sentiment analysis revealed that there were more positive-sentiment entries than negative-sentiment (3,675 vs. 934, p<0.001). CONCLUSIONS: This quantitative analysis of anonymous male infertility internet discussion is a step towards increased understanding of the complete patient experience. Providers can use this knowledge to better engage with patients in clinic and online. Source of Funding: None © 2021 by American Urological Association Education and Research, Inc.FiguresReferencesRelatedDetails Volume 206Issue Supplement 3September 2021Page: e349-e349 Advertisement Copyright & Permissions© 2021 by American Urological Association Education and Research, Inc.MetricsAuthor Information Krishna Ravivarapu More articles by this author Alexander Small More articles by this author Evan Garden More articles by this author Micah Levy More articles by this author Osama Al-Alao More articles by this author Michael Palese More articles by this author Expand All Advertisement Loading ...