How do others cope? Extracting coping strategies for adverse drug events from social media.

Patients advise their peers on how to cope with their illness in daily life on online support groups. To date, no efforts have been made to automatically extract recommended coping strategies from online patient discussion groups. We introduce this new task, which poses a number of challenges including complex, long entities, a large long-tailed label space, and cross-document relations. We present an initial ontology for coping strategies as a starting point for future research on coping strategies, and the first end-to-end pipeline for extracting coping strategies for side effects. We also compared two possible computational solutions for this novel and highly challenging task; multi-label classification and named entity recognition (NER) with entity linking (EL). We evaluated our methods on the discussion forum from the Facebook group of the worldwide patient support organization 'GIST support international' (GSI); GIST support international donated the data to us. We found that coping strategy extraction is difficult and both methods attain limited performance (measured with F1 score) on held out test sets; multi-label classification outperforms NER+EL (F1=0.220 vs F1=0.155). An inspection of the multi-label classification output revealed that for some of the incorrect predictions, the reference label is close to the predicted label in the ontology (e.g. the predicted label 'juice' instead of the more specific reference label 'grapefruit juice'). Performance increased to F1=0.498 when we evaluated at a coarser level of the ontology. We conclude that our pipeline can be used in a semi-automatic setting, in interaction with domain experts to discover coping strategies for side effects from a patient forum. For example, we found that patients recommend ginger tea for nausea and magnesium and potassium supplements for cramps. This information can be used as input for patient surveys or clinical studies.

[Making a diagnosis with the aid of the internet in times of large language models]. / Diagnose stellen met online tools.

The internet is an excellent aid in making diagnoses. One can retrieve diagnostic information from a reliable source such as a continuously updated textbook or search specifically for a diagnosis in bibliographic databases such as PubMed. Entry of a patient summary in a general search engine or a large language model such as ChatGPT can suggest differential-diagnoses to the expert user,but one must be conscious of the limitations of current large language models. There seems little room left for the traditional differential-diagnosis generators. Ideally, large language models will be combined with transparent algorithms with which medical data can be retrieved, to create a new generation of diagnostic decision support systems.

Sample Bias in Web-Based Patient-Generated Health Data of Dutch Patients With Gastrointestinal Stromal Tumor: Survey Study.

BACKGROUND: Increasingly, social media is being recognized as a potential resource for patient-generated health data, for example, for pharmacovigilance. Although the representativeness of the web-based patient population is often noted as a concern, studies in this field are limited. OBJECTIVE: This study aimed to investigate the sample bias of patient-centered social media in Dutch patients with gastrointestinal stromal tumor (GIST). METHODS: A population-based survey was conducted in the Netherlands among 328 patients with GIST diagnosed 2-13 years ago to investigate their digital communication use with fellow patients. A logistic regression analysis was used to analyze clinical and demographic differences between forum users and nonusers. RESULTS: Overall, 17.9% (59/328) of survey respondents reported having contact with fellow patients via social media. Moreover, 78% (46/59) of forum users made use of GIST patient forums. We found no statistically significant differences for age, sex, socioeconomic status, and time since diagnosis between forum users (n=46) and nonusers (n=273). Patient forum users did differ significantly in (self-reported) treatment phase from nonusers (P=.001). Of the 46 forum users, only 2 (4%) were cured and not being monitored; 3 (7%) were on adjuvant, curative treatment; 19 (41%) were being monitored after adjuvant treatment; and 22 (48%) were on palliative treatment. In contrast, of the 273 patients who did not use disease-specific forums to communicate with fellow patients, 56 (20.5%) were cured and not being monitored, 31 (11.3%) were on curative treatment, 139 (50.9%) were being monitored after treatment, and 42 (15.3%) were on palliative treatment. The odds of being on a patient forum were 2.8 times as high for a patient who is being monitored compared with a patient that is considered cured. The odds of being on a patient forum were 1.9 times as high for patients who were on curative (adjuvant) treatment and 10 times as high for patients who were in the palliative phase compared with patients who were considered cured. Forum users also reported a lower level of social functioning (84.8 out of 100) than nonusers (93.8 out of 100; P=.008). CONCLUSIONS: Forum users showed no particular bias on the most important demographic variables of age, sex, socioeconomic status, and time since diagnosis. This may reflect the narrowing digital divide. Overrepresentation and underrepresentation of patients with GIST in different treatment phases on social media should be taken into account when sourcing patient forums for patient-generated health data. A further investigation of the sample bias in other web-based patient populations is warranted.

Automated gathering of real-world data from online patient forums can complement pharmacovigilance for rare cancers.

Current methods of pharmacovigilance result in severe under-reporting of adverse drug events (ADEs). Patient forums have the potential to complement current pharmacovigilance practices by providing real-time uncensored and unsolicited information. We are the first to explore the value of patient forums for rare cancers. To this end, we conduct a case study on a patient forum for Gastrointestinal Stromal Tumor patients. We have developed machine learning algorithms to automatically extract and aggregate side effects from messages on open online discussion forums. We show that patient forum data can provide suggestions for which ADEs impact quality of life the most: For many side effects the relative reporting rate differs decidedly from that of the registration trials, including for example cognitive impairment and alopecia as side effects of avapritinib. We also show that our methods can provide real-world data for long-term ADEs, such as osteoporosis and tremors for imatinib, and novel ADEs not found in registration trials, such as dry eyes and muscle cramping for imatinib. We thus posit that automated pharmacovigilance from patient forums can provide real-world data for ADEs and should be employed as input for medical hypotheses for rare cancers.

Symptoms reported by gastrointestinal stromal tumour (GIST) patients on imatinib treatment: combining questionnaire and forum data.

PURPOSE: Treatment with the tyrosine kinase inhibitor (TKI) imatinib in patients with gastrointestinal stromal tumours (GIST) causes symptoms that could negatively impact health-related quality of life (HRQoL). Treatment-related symptoms are usually clinician-reported and little is known about patient reports. We used survey and online patient forum data to investigate (1) prevalence of patient-reported symptoms; (2) coverage of symptoms mentioned on the forum by existing HRQoL questionnaires; and (3) priorities of prevalent symptoms in HRQoL assessment. METHODS: In the cross-sectional population-based survey study, Dutch GIST patients completed items from the EORTC QLQ-C30 and Symptom-Based Questionnaire (SBQ). In the forum study, machine learning algorithms were used to extract TKI side-effects from English messages on an international online forum for GIST patients. Prevalence of symptoms related to imatinib treatment in both sources was calculated and exploratively compared. RESULTS: Fatigue and muscle pain or cramps were reported most frequently. Seven out of 10 most reported symptoms (i.e. fatigue, muscle pain or cramps, facial swelling, joint pain, skin problems, diarrhoea, and oedema) overlapped between the two sources. Alopecia was frequently mentioned on the forum, but not in the survey. Four out of 10 most reported symptoms on the online forum are covered by the EORTC QLQ-C30. The EORTC-SBQ and EORTC Item Library cover 9 and 10 symptoms, respectively. CONCLUSION: This first overview of patient-reported imatinib-related symptoms from two data sources helps to determine coverage of items in existing questionnaires, and prioritize HRQoL issues. Combining cancer-generic instruments with treatment-specific item lists will improve future HRQoL assessment in care and research in GIST patients using TKI.

Perspectives on automated composition of workflows in the life sciences.

Scientific data analyses often combine several computational tools in automated pipelines, or workflows. Thousands of such workflows have been used in the life sciences, though their composition has remained a cumbersome manual process due to a lack of standards for annotation, assembly, and implementation. Recent technological advances have returned the long-standing vision of automated workflow composition into focus. This article summarizes a recent Lorentz Center workshop dedicated to automated composition of workflows in the life sciences. We survey previous initiatives to automate the composition process, and discuss the current state of the art and future perspectives. We start by drawing the "big picture" of the scientific workflow development life cycle, before surveying and discussing current methods, technologies and practices for semantic domain modelling, automation in workflow development, and workflow assessment. Finally, we derive a roadmap of individual and community-based actions to work toward the vision of automated workflow development in the forthcoming years. A central outcome of the workshop is a general description of the workflow life cycle in six stages: 1) scientific question or hypothesis, 2) conceptual workflow, 3) abstract workflow, 4) concrete workflow, 5) production workflow, and 6) scientific results. The transitions between stages are facilitated by diverse tools and methods, usually incorporating domain knowledge in some form. Formal semantic domain modelling is hard and often a bottleneck for the application of semantic technologies. However, life science communities have made considerable progress here in recent years and are continuously improving, renewing interest in the application of semantic technologies for workflow exploration, composition and instantiation. Combined with systematic benchmarking with reference data and large-scale deployment of production-stage workflows, such technologies enable a more systematic process of workflow development than we know today. We believe that this can lead to more robust, reusable, and sustainable workflows in the future.

DeepADEMiner: a deep learning pharmacovigilance pipeline for extraction and normalization of adverse drug event mentions on Twitter.

OBJECTIVE: Research on pharmacovigilance from social media data has focused on mining adverse drug events (ADEs) using annotated datasets, with publications generally focusing on 1 of 3 tasks: ADE classification, named entity recognition for identifying the span of ADE mentions, and ADE mention normalization to standardized terminologies. While the common goal of such systems is to detect ADE signals that can be used to inform public policy, it has been impeded largely by limited end-to-end solutions for large-scale analysis of social media reports for different drugs. MATERIALS AND METHODS: We present a dataset for training and evaluation of ADE pipelines where the ADE distribution is closer to the average 'natural balance' with ADEs present in about 7% of the tweets. The deep learning architecture involves an ADE extraction pipeline with individual components for all 3 tasks. RESULTS: The system presented achieved state-of-the-art performance on comparable datasets and scored a classification performance of F1 = 0.63, span extraction performance of F1 = 0.44 and an end-to-end entity resolution performance of F1 = 0.34 on the presented dataset. DISCUSSION: The performance of the models continues to highlight multiple challenges when deploying pharmacovigilance systems that use social media data. We discuss the implications of such models in the downstream tasks of signal detection and suggest future enhancements. CONCLUSION: Mining ADEs from Twitter posts using a pipeline architecture requires the different components to be trained and tuned based on input data imbalance in order to ensure optimal performance on the end-to-end resolution task.

The reach of commercially motivated junk news on Facebook.

Commercially motivated junk news-i.e. money-driven, highly shareable clickbait with low journalistic production standards-constitutes a vast and largely unexplored news media ecosystem. Using publicly available Facebook data, we compared the reach of junk news on Facebook pages in the Netherlands to the reach of Dutch mainstream news on Facebook. During the period 2013-2017 the total number of user interactions with junk news significantly exceeded that with mainstream news. Over 5 Million of the 10 Million Dutch Facebook users have interacted with a junk news post at least once. Junk news Facebook pages also had a significantly stronger increase in the number of user interactions over time than mainstream news. Since the beginning of 2016 the average number of user interactions per junk news post has consistently exceeded the average number of user interactions per mainstream news post.

Analyzing Empowerment Processes Among Cancer Patients in an Online Community: A Text Mining Approach.

BACKGROUND: Peer-to-peer online support groups and the discussion forums in these groups can help patients by providing opportunities for increasing their empowerment. Most previous research on online empowerment and online social support uses qualitative methods or questionnaires to gain insight into the dynamics of online empowerment processes. OBJECTIVE: The overall goal of this study was to analyze the presence of the empowerment processes in the online peer-to-peer communication of people affected by cancer, using text mining techniques. Use of these relatively new methods enables us to study social processes such as empowerment on a large scale and with unsolicited data. METHODS: The sample consisted of 5534 messages in 1708 threads, written by 2071 users of a forum for cancer patients and their relatives. We labeled the posts in our sample with 2 types of labels: labels referring to empowerment processes and labels denoting psychological processes. The latter were identified using the Linguistic Inquiry and Word Count (LIWC) method. Both groups of labels were automatically assigned to posts. Automatic labeling of the empowerment processes was done by text classifiers trained on a manually labeled subsample. For the automatic labeling of the LIWC categories, we used the Dutch version of the LIWC consisting of a total of 66 word categories that are assigned to text based on occurrences of words in the text. After the automatic labeling with both types of labels, we investigated (1) the relationship between empowerment processes and the intensity of online participation, (2) the relationship between empowerment processes and the LIWC categories, and (3) the differences between patients with different types of cancer. RESULTS: The precision of the automatic labeling was 85.6%, which we considered to be sufficient for automatically labeling the complete corpus and doing further analyses on the labeled data. Overall, 62.94% (3482/5532) of the messages contained a narrative, 23.83% (1318/5532) a question, and 27.49% (1521/5532) informational support. Emotional support and references to external sources were less frequent. Users with more posts more often referred to an external source and more often provided informational support and emotional support (Kendall τ>0.2; P<.001) and less often shared narratives (Kendall τ=-0.297; P<.001). A number of LIWC categories are significant predictors for the empowerment processes: words expressing assent (ok and yes) and emotional processes (expressions of feelings) are significant positive predictors for emotional support (P=.002). The differences between patients with different types of cancer are small. CONCLUSIONS: Empowerment processes are associated with the intensity of online use. The relationship between linguistic analyses and empowerment processes indicates that empowerment processes can be identified from the occurrences of specific linguistic cues denoting psychological processes.

Predicting life expectancy with a long short-term memory recurrent neural network using electronic medical records.

BACKGROUND: Life expectancy is one of the most important factors in end-of-life decision making. Good prognostication for example helps to determine the course of treatment and helps to anticipate the procurement of health care services and facilities, or more broadly: facilitates Advance Care Planning. Advance Care Planning improves the quality of the final phase of life by stimulating doctors to explore the preferences for end-of-life care with their patients, and people close to the patients. Physicians, however, tend to overestimate life expectancy, and miss the window of opportunity to initiate Advance Care Planning. This research tests the potential of using machine learning and natural language processing techniques for predicting life expectancy from electronic medical records. METHODS: We approached the task of predicting life expectancy as a supervised machine learning task. We trained and tested a long short-term memory recurrent neural network on the medical records of deceased patients. We developed the model with a ten-fold cross-validation procedure, and evaluated its performance on a held-out set of test data. We compared the performance of a model which does not use text features (baseline model) to the performance of a model which uses features extracted from the free texts of the medical records (keyword model), and to doctors' performance on a similar task as described in scientific literature. RESULTS: Both doctors and the baseline model were correct in 20% of the cases, taking a margin of 33% around the actual life expectancy as the target. The keyword model, in comparison, attained an accuracy of 29% with its prognoses. While doctors overestimated life expectancy in 63% of the incorrect prognoses, which harms anticipation to appropriate end-of-life care, the keyword model overestimated life expectancy in only 31% of the incorrect prognoses. CONCLUSIONS: Prognostication of life expectancy is difficult for humans. Our research shows that machine learning and natural language processing techniques offer a feasible and promising approach to predicting life expectancy. The research has potential for real-life applications, such as supporting timely recognition of the right moment to start Advance Care Planning.

Online support community for adolescents and young adults (AYAs) with cancer: user statistics, evaluation, and content analysis.

PURPOSE: Peer support is an important unmet need among adolescent and young adult (AYA) cancer patients. This study was conducted to describe the use and evaluation of a Dutch secure online support community for AYA diagnosed with cancer between 18 and 35 years. METHODS: User statistics were collected with Google analytics. Community members were asked to complete questionnaires on the usefulness of the community. A content analysis using Linguistic Inquiry and Word Count was conducted. RESULTS: Between 2010 and 2017, the community received 433 AYA members (71% female; mean age at diagnosis 25.7 years; 52 Dutch hospitals represented). The mean time since diagnosis when subscribing to the community was 2.7 years (SD 4.4). Questionnaire data among 30 AYA community members indicated that the use of the community resulted in acknowledgment and advice regarding problems (56%) and the feeling of being supported (63%). Almost half of the respondents felt less lonely, 78% experienced recognition in stories of other AYA. Anonymized content analysis (n=14) showed that the majority of the online discussions encompassed emotional and cognitive expressions, and emotional support. CONCLUSION: The secure Dutch online AYA community can help AYA cancer patients to express feelings, exchange information, address peer support, and has been found helpful in coping with cancer. As AYA cancer patients often lack the option of meeting each other in person, the AYA community is helpful in establishing peer support. Its use would benefit from promotion by health care professionals.

Analysis of Content Shared in Online Cancer Communities: Systematic Review.

BACKGROUND: The content that cancer patients and their relatives (ie, posters) share in online cancer communities has been researched in various ways. In the past decade, researchers have used automated analysis methods in addition to manual coding methods. Patients, providers, researchers, and health care professionals can learn from experienced patients, provided that their experience is findable. OBJECTIVE: The aim of this study was to systematically review all relevant literature that analyzes user-generated content shared within online cancer communities. We reviewed the quality of available research and the kind of content that posters share with each other on the internet. METHODS: A computerized literature search was performed via PubMed (MEDLINE), PsycINFO (5 and 4 stars), Cochrane Central Register of Controlled Trials, and ScienceDirect. The last search was conducted in July 2017. Papers were selected if they included the following terms: (cancer patient) and (support group or health communities) and (online or internet). We selected 27 papers and then subjected them to a 14-item quality checklist independently scored by 2 investigators. RESULTS: The methodological quality of the selected studies varied: 16 were of high quality and 11 were of adequate quality. Of those 27 studies, 15 were manually coded, 7 automated, and 5 used a combination of methods. The best results can be seen in the papers that combined both analytical methods. The number of analyzed posts ranged from 200 to 1,500,000; the number of analyzed posters ranged from 75 to 90,000. The studies analyzing large numbers of posts mainly related to breast cancer, whereas those analyzing small numbers were related to other types of cancers. A total of 12 studies involved some or entirely automatic analysis of the user-generated content. All the authors referred to two main content categories: informational support and emotional support. In all, 15 studies reported only on the content, 6 studies explicitly reported on content and social aspects, and 6 studies focused on emotional changes. CONCLUSIONS: In the future, increasing amounts of user-generated content will become available on the internet. The results of content analysis, especially of the larger studies, give detailed insights into patients' concerns and worries, which can then be used to improve cancer care. To make the results of such analyses as usable as possible, automatic content analysis methods will need to be improved through interdisciplinary collaboration.