ALT Key Points Qualitative analysis of 7328 social media and e-forum posts captured candid, real-world perspectives from patients with IBS-C, revealing variability in perceived medication efficacy, treatment-related burden, intentional non-adherence, and barriers to adherence. NLP and AI methods demonstrated that the large language model reliably classified patient posts, achieving performance comparable to human annotators and supporting its use for larger-scale analyses of patient experiences. Given the complexity of patients' treatment experiences, these findings underscore the importance of shared decision-making and proactive management of treatment barriers to support personalized care and improved patient outcomes.