34399 Analysis of online psoriasis support group reveals great interest in adjuvant therapy

Abstract

Background: Patients with psoriasis are using social media to gather and disseminate information about disease management. There is concern over a potential spread of misinformation in these online communities. Objective: We hypothesize that analyzing patient-generated posts in a large online forum would allow for an assessment of bias toward various types of treatment. Methods: We collected 100,262 posts by 14,565 users made from July 1, 2016 to June 30, 2021 in a popular online forum for psoriasis. We applied Latent Dirichlet Allocation (LDA), an unsupervised machine learning model, to organize the posts into topics. Keywords for each topic supplied by LDA were used to manually assign topic labels. Results: We recognized 40 significant topics of conversation, organized into 6 major categories, including “Adjuvants” (31.98%), “Clinical Presentation” (13.77%), “Mental Health” (19.61%), “Treatment” (13.61%), “Information” (12.83%) and “Logistics” (8.20%). Within “Treatments” there was more discussion about “Biologics” (5.81%) than “Systemic” (3.76%) and “UV” (1.53%) therapies. Popular adjuvant therapies discussed included “Moisturizers” (9.91%), “Diet” (9.3%), “Makeup” (2.80%), and “Wrapping” (1.92%). The bulk of hyperlinks shared pointed to other forum posts (27.78%) and evidenced-based journal articles (9.20%). Limitations: LDA classifies posts into topics based on frequencies of words within the posts without an understanding of language or context. Prospective studies are needed to validate these findings. Conclusions: This online forum shows evidence of a balanced discussion of treatment options. There is stronger interest towards adjuvant therapies in managing psoriasis. Patients also use social media to share and receive emotional support.