Understanding Breast Implant Illness via Social Media Data Analysis.

Background: Breast implants have been increasingly popular over the last 20 years. There have been growing concerns with the risks of implants. Meanwhile, media phenomenon called breast implant (BII) has emerged. Objective: To identify and summarize key attributes of BII using social media data. Materials and Methods: We conducted social media data analysis to better understand the symptoms, signs, etc., that are associated with BII using Natural Language Processing (NLP) and topic modeling. We extracted mentions related to signs/symptoms, diseases/disorders and medical procedures using the Clinical Text Analysis and Knowledge Extraction System (cTAKES). Extracted mentions are mapped to standard medical concepts. We summarized mapped concepts to topics using Latent Dirichlet Allocation (LDA). Results: Our analysis identified topics related to toxicity, cancer and mental health issues that are highly associated with implant illness. We also identified pains and other disorders commonly associated with implant illness. Discussion: Our analysis suggests that implant illness can possibly lead to serious health issues such as autoimmune disorders, cancer, pain, fatigue. We also find that toxicity from silicone implants and mental health concerns are some underlying factors in BII. Our study could inspire future work to further study the suggested symptoms and factors of BII. Conclusion: Our analysis over social media data identifies mentions such as rupture, infection, pain and fatigue that are considered common self-reported issues among the public. Our analysis also shows that cancers, autoimmune disorders and mental health problems are emerging concerns, albeit less studied for implants.