Using Large-scale Social Media Analytics to Understand Patient Perspectives About Urinary Tract Infections: Thematic Analysis

Abstract

Background Current qualitative literature about the experiences of women dealing with urinary tract infections (UTIs) is limited to patients recruited from tertiary centers and medical clinics. However, traditional focus groups and interviews may limit what patients share. Using digital ethnography, we analyzed free-range conversations of an online community. Objective This study aimed to investigate and characterize the patient perspectives of women dealing with UTIs using digital ethnography. Methods A data-mining service was used to identify online posts. A thematic analysis was conducted on a subset of the identified posts. Additionally, a latent Dirichlet allocation (LDA) probabilistic topic modeling method was applied to review the entire data set using a semiautomatic approach. Each identified topic was generated as a discrete distribution over the words in the collection, which can be thought of as a word cloud. We also performed a thematic analysis of the word cloud topic model results. Results A total of 83,589 posts by 53,460 users from 859 websites were identified. Our hand-coding inductive analysis yielded the following 7 themes: quality-of-life impact, knowledge acquisition, support of the online community, health care utilization, risk factors and prevention, antibiotic treatment, and alternative therapies. Using the LDA topic model method, 105 themes were identified and consolidated into 9 categories. Of the LDA-derived themes, 25.7% (27/105) were related to online community support, and 22% (23/105) focused on UTI risk factors and prevention strategies. Conclusions Our large-scale social media analysis supports the importance and reproducibility of using online data to comprehend women’s UTI experience. This inductive thematic analysis highlights patient behavior, self-empowerment, and online media utilization by women to address their health concerns in a safe, anonymous way.