Active neural networks to detect mentions of changes to medication treatment in social media.

Active neural networks to detect mentions of changes to medication treatment in social media.
复制标题

DOI:
10.1093/jamia/ocab158
复制
发表时间:
2021-11-25
期刊:
Journal of the American Medical Informatics Association : JAMIA
影响因子:
--
通讯作者:
Gonzalez-Hernandez G
Gonzalez-Hernandez G
中科院分区:
其他
文献类型:
--
作者:
Weissenbacher D;Ge S;Klein A;O'Connor K;Gross R;Hennessy S;Gonzalez-Hernandez G

文献摘要

参考文献

相似文献

我们提出了使用社交媒体数据来补充当前监测人群用药不依从性的努力的第一步:检测药物治疗的变化。未经医生监督的药物治疗改变,例如剂量或摄入频率的改变,就是不遵守药物治疗。尽管会产生包括健康状况恶化或死亡在内的后果,但估计有 50% 的患者没有按照指示服用药物。目前识别不依从性的方法有很大的局限性。直接观察可能是侵入性的或昂贵的,而通过患者调查进行的间接观察在很大程度上依赖于患者的记忆和坦诚。在这些研究中使用社交媒体数据可能会解决这些局限性。我们对 9830 条提及药物的推文进行了注释,并训练了一个卷积神经网络 (CNN) 来查找提及药物治疗变化的内容,无论这些变化是否是医生建议的。我们使用来自 WebMD 注释的 12972 条评论的主动学习和迁移学习来解决 Twitter 语料库的类别不平衡问题。为了验证我们的 CNN 并探索未来的方向,我们对 1956 条积极的推文进行了注释,说明它们是否反映了不遵守规定,并对给出的原因进行了分类。我们的 CNN 在这个新语料库上获得了 0.50 F1 分数。对正面推文的手动分析显示,在包含 9 类不遵守原因的子集中,不遵守情况很明显。我们发现社交媒体用户公开讨论药物治疗的变化,并可能解释他们的原因,包括何时构成不依从。这种方法可能有助于补充当前依从性监测方面的努力。
We address a first step toward using social media data to supplement current efforts in monitoring population-level medication nonadherence: detecting changes to medication treatment. Medication treatment changes, like changes to dosage or to frequency of intake, that are not overseen by physicians are, by that, nonadherence to medication. Despite the consequences, including worsening health conditions or death, 50% of patients are estimated to not take medications as indicated. Current methods to identify nonadherence have major limitations. Direct observation may be intrusive or expensive, and indirect observation through patient surveys relies heavily on patients’ memory and candor. Using social media data in these studies may address these limitations. We annotated 9830 tweets mentioning medications and trained a convolutional neural network (CNN) to find mentions of medication treatment changes, regardless of whether the change was recommended by a physician. We used active and transfer learning from 12 972 reviews we annotated from WebMD to address the class imbalance of our Twitter corpus. To validate our CNN and explore future directions, we annotated 1956 positive tweets as to whether they reflect nonadherence and categorized the reasons given. Our CNN achieved 0.50 F1-score on this new corpus. The manual analysis of positive tweets revealed that nonadherence is evident in a subset with 9 categories of reasons for nonadherence. We showed that social media users publicly discuss medication treatment changes and may explain their reasons including when it constitutes nonadherence. This approach may be useful to supplement current efforts in adherence monitoring.
DOI: 10.7326/m18-0543
发表时间: 2018-07-03
影响因子: 39.2
作者:
De Geest S;Zullig LL;Dunbar-Jacob J;Helmy R;Hughes DA;Wilson IB;Vrijens B
通讯作者: Vrijens B
药物疗法不合规。
DOI: 10.4103/0253-7176.101762
发表时间: 2012-04
影响因子: 2.8
作者:
Reddy MS
通讯作者: Reddy MS
DOI: 10.2196/jmir.9222
发表时间: 2018-03-14
影响因子: 7.4
作者:
Abdellaoui R;Foulquié P;Texier N;Faviez C;Burgun A;Schück S
通讯作者: Schück S
DOI: 10.1155/2015/217047
发表时间: 2015
影响因子: --
作者:
Lam WY;Fresco P
通讯作者: Fresco P
DOI: 10.1001/jamanetworkopen.2020.8953
发表时间: 2020-06-25
期刊: JAMA NETWORK OPEN
影响因子: 13.8
作者:
Golder, Su;O'Connor, Karen;Gonzalez-Hernandez, Graciela
通讯作者: Gonzalez-Hernandez, Graciela