宾夕法尼亚大学研究人员运用人工智能技术分析了约150,000条Reddit帖子，发现使用未经批准版本雷他鲁肽的用户报告的症状与临床试验结果存在差异1。研究团队从约40,000名用户的评论中识别出约14,000名声称个人使用未批准雷他鲁肽版本的用户1。在7,823名报告至少一种可医学分类症状的用户中，增加食欲成为最常见的可分类症状1，这与该药物抑制饥饿的设计目的相悖。
与临床试验数据相比，Reddit用户在食欲相关变化、睡眠障碍、心率升高和肌肉相关症状的报告比例均更高1。研究人员强调这些发现不应被解释为已证实的新副作用1，并指出来自灰色市场的产品成分、纯度和剂量可能与真实实验药物不同1。该研究表明人工智能可以帮助识别可能需要进一步调查的潜在药物安全信号1，相关发现已发表在《JAMA Network Open》期刊上1。
Researchers at the University of Pennsylvania employed artificial intelligence to examine approximately 150,000 Reddit posts and identified a striking discrepancy between symptoms reported by users of unapproved versions of semaglutide and findings from clinical trials 1. The analysis, which included comments from roughly 40,000 users, pinpointed approximately 14,000 individuals claiming personal use of the unauthorized drug formulation 1.
Among 7,823 users who reported at least one medically classifiable symptom, increased appetite emerged as the most common reported side effect 1. This finding contradicts the drug's intended mechanism, which is designed to suppress hunger. Compared to clinical trial data, Reddit users reported significantly higher rates of appetite-related changes, sleep disorders, elevated heart rate, and muscle-related symptoms 1. The researchers emphasized that these observations should not be interpreted as confirmed new adverse effects, noting that products from gray market sources may differ substantially in composition, purity, and dosage from legitimate experimental medications 1.
The study, published in JAMA Network Open, demonstrates how artificial intelligence can assist in identifying potential drug safety signals warranting further investigation 1.
评论
还没有评论，欢迎留下第一条。