Detecting clinically related content in online patient posts

Reporting of Ethical Considerations in Qualitative Research Utilizing Social Media Data on Public Health Care: Scoping Review

BACKGROUND

The internet community has become a significant source for researchers to conduct qualitative studies analyzing users' views, attitudes, and experiences about public health. However, few studies have assessed the ethical issues in qualitative research using social media data.

OBJECTIVE

This study aims to review the reportage of ethical considerations in qualitative research utilizing social media data on public health care.

METHODS

We performed a scoping review of studies mining text from internet communities and published in peer-reviewed journals from 2010 to May 31, 2023. These studies, limited to the English language, were retrieved to evaluate the rates of reporting ethical approval, informed consent, and privacy issues. We searched 5 databases, that is, PubMed, Web of Science, CINAHL, Cochrane, and Embase. Gray literature was supplemented from Google Scholar and OpenGrey websites. Studies using qualitative methods mining text from the internet community focusing on health care topics were deemed eligible. Data extraction was performed using a standardized data extraction spreadsheet. Findings were reported using PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews) guidelines.

RESULTS

After 4674 titles, abstracts, and full texts were screened, 108 studies on mining text from the internet community were included. Nearly half of the studies were published in the United States, with more studies from 2019 to 2022. Only 59.3% (64/108) of the studies sought ethical approval, 45.3% (49/108) mentioned informed consent, and only 12.9% (14/108) of the studies explicitly obtained informed consent. Approximately 86% (12/14) of the studies that reported informed consent obtained digital informed consent from participants/administrators, while 14% (2/14) did not describe the method used to obtain informed consent. Notably, 70.3% (76/108) of the studies contained users' written content or posts: 68% (52/76) contained verbatim quotes, while 32% (24/76) paraphrased the quotes to prevent traceability. However, 16% (4/24) of the studies that paraphrased the quotes did not report the paraphrasing methods. Moreover, 18.5% (20/108) of the studies used aggregated data analysis to protect users' privacy. Furthermore, the rates of reporting ethical approval were different between different countries (P=.02) and between papers that contained users' written content (both direct and paraphrased quotes) and papers that did not contain users' written content (P<.001).

CONCLUSIONS

Our scoping review demonstrates that the reporting of ethical considerations is widely neglected in qualitative research studies using social media data; such studies should be more cautious in citing user quotes to maintain user privacy. Further, our review reveals the need for detailed information on the precautions of obtaining informed consent and paraphrasing to reduce the potential bias. A national consensus of ethical considerations such as ethical approval, informed consent, and privacy issues is needed for qualitative research of health care using social media data of internet communities.

Toward an Ethical Framework for the Text Mining of Social Media for Health Research: A Systematic Review

Background: Text-mining techniques are advancing all the time and vast corpora of social media text can be analyzed for users' views and experiences related to their health. There is great promise for new insights into health issues such as drug side effects and spread of disease, as well as patient experiences of health conditions and health care. However, this emerging field lacks ethical consensus and guidance. We aimed to bring together a comprehensive body of opinion, views, and recommendations in this area so that academic researchers new to the field can understand relevant ethical issues. Methods: After registration of a protocol in PROSPERO, three parallel systematic searches were conducted, to identify academic articles comprising commentaries, opinion, and recommendations on ethical practice in social media text mining for health research and gray literature guidelines and recommendations. These were integrated with social media users' views from qualitative studies. Papers and reports that met the inclusion criteria were analyzed thematically to identify key themes, and an overarching set of themes was deduced. Results: A total of 47 reports and articles were reviewed, and eight themes were identified. Commentators suggested that publicly posted social media data could be used without consent and formal research ethics approval, provided that the anonymity of users is ensured, although we note that privacy settings are difficult for users to navigate on some sites. Even without the need for formal approvals, we note ethical issues: to actively identify and minimize possible harms, to conduct research for public benefit rather than private gain, to ensure transparency and quality of data access and analysis methods, and to abide by the law and terms and conditions of social media sites. Conclusion: Although social media text mining can often legally and reasonably proceed without formal ethics approvals, we recommend improving ethical standards in health-related research by increasing transparency of the purpose of research, data access, and analysis methods; consultation with social media users and target groups to identify and mitigate against potential harms that could arise; and ensuring the anonymity of social media users.

BertMCN: Mapping colloquial phrases to standard medical concepts using BERT and highway network

In the last few years, people started to share lots of information related to health in the form of tweets, reviews and blog posts. All these user generated clinical texts can be mined to generate useful insights. However, automatic analysis of clinical text requires identification of standard medical concepts. Most of the existing deep learning based medical concept normalization systems are based on CNN or RNN. Performance of these models is limited as they have to be trained from scratch (except embeddings). In this work, we propose a medical concept normalization system based on BERT and highway layer. BERT, a pre-trained context sensitive deep language representation model advanced state-of-the-art performance in many NLP tasks and gating mechanism in highway layer helps the model to choose only important information. Experimental results show that our model outperformed all existing methods on two standard datasets. Further, we conduct a series of experiments to study the impact of different learning rates and batch sizes, noise and freezing encoder layers on our model.

Deep Sentiment Classification and Topic Discovery on Novel Coronavirus or COVID-19 Online Discussions: NLP Using LSTM Recurrent Neural Network Approach

Internet forums and public social media, such as online healthcare forums, provide a convenient channel for users (people/patients) concerned about health issues to discuss and share information with each other. In late December 2019, an outbreak of a novel coronavirus (infection from which results in the disease named COVID-19) was reported, and, due to the rapid spread of the virus in other parts of the world, the World Health Organization declared a state of emergency. In this paper, we used automated extraction of COVID-19-related discussions from social media and a natural language process (NLP) method based on topic modeling to uncover various issues related to COVID-19 from public opinions. Moreover, we also investigate how to use LSTM recurrent neural network for sentiment classification of COVID-19 comments. Our findings shed light on the importance of using public opinions and suitable computational techniques to understand issues surrounding COVID-19 and to guide related decision-making. In addition, experiments demonstrated that the research model achieved an accuracy of 81.15% - a higher accuracy than that of several other well-known machine-learning algorithms for COVID-19-Sentiment Classification.

The therapy is making me sick: how online portal communications between breast cancer patients and physicians indicate medication discontinuation

Objective

Online platforms have created a variety of opportunities for breast patients to discuss their hormonal therapy, a long-term adjuvant treatment to reduce the chance of breast cancer occurrence and mortality. The goal of this investigation is to ascertain the extent to which the messages breast cancer patients communicated through an online portal can indicate their potential for discontinuing hormonal therapy.

Materials and Methods

We studied the de-identified electronic medical records of 1106 breast cancer patients who were prescribed hormonal therapy at Vanderbilt University Medical Center over a 12-year period. We designed a data-driven approach to investigate patients' patterns of messaging with healthcare providers, the topics they communicated, and the extent to which these messaging behaviors associate with the likelihood that a patient will discontinue a prescribed 5-year regimen of therapy.

Results

The results indicates that messaging rate over time [hazard ratio (HR) = 1.373, P = 0.002], mentions of side effects (HR = 1.214, P = 0.006), and surgery-related topics (HR = 1.170, P = 0.034) were associated with increased risk of early medication discontinuation. In contrast, seeking professional suggestions (HR = 0.766, P = 0.002), expressing gratitude to healthcare providers (HR = 0.872, P = 0.044), and mentions of drugs used to treat side effects (HR = 0.807, P = 0.013) were associated with decreased risk of medication discontinuation.

Discussion and Conclusion

This investigation suggests that patient-generated content can inform the study of health-related behaviors. Given that approximately 50% of breast cancer patients do not complete a course of hormonal therapy as described, the identification of factors associated with medication discontinuation can facilitate real-time interventions to prevent early discontinuation.

Knowledge Discovery from Posts in Online Health Communities Using Unified Medical Language System

Patient-reported posts in Online Health Communities (OHCs) contain various valuable information that can help establish knowledge-based online support for online patients. However, utilizing these reports to improve online patient services in the absence of appropriate medical and healthcare expert knowledge is difficult. Thus, we propose a comprehensive knowledge discovery method that is based on the Unified Medical Language System for the analysis of narrative posts in OHCs. First, we propose a domain-knowledge support framework for OHCs to provide a basis for post analysis. Second, we develop a Knowledge-Involved Topic Modeling (KI-TM) method to extract and expand explicit knowledge within the text. We propose four metrics, namely, explicit knowledge rate, latent knowledge rate, knowledge correlation rate, and perplexity, for the evaluation of the KI-TM method. Our experimental results indicate that our proposed method outperforms existing methods in terms of providing knowledge support. Our method enhances knowledge support for online patients and can help develop intelligent OHCs in the future.