The Unintended Consequences of Social Media in Healthcare: New Problems and New Solutions.

OBJECTIVES: Social media is increasingly being used in conjunction with health information technology (health IT). The objective of this paper is to identify some of the undesirable outcomes that arise from this integration and to suggest solutions to these problems. METHODOLOGY: After a discussion with experts to elicit the topics that should be included in the survey, we performed a narrative review based on recent literature and interviewed multidisciplinary experts from different areas. In each case, we identified and analyzed the unintended effects of social media in health IT. RESULTS: Each analyzed topic provided a different set of unintended consequences. Most relevant consequences include lack of privacy with ethical and legal issues, patient confusion in disease management, poor information accuracy in crowdsourcing, unclear responsibilities, misleading and biased information in the prevention and detection of epidemics, and demotivation in gamified health solutions with social components. CONCLUSIONS: Using social media in healthcare offers several benefits, but it is not exempt of potential problems, and not all of these problems have clear solutions. We recommend careful design of digital systems in order to minimize patient's feelings of demotivation and frustration and we recommend following specific guidelines that should be created by all stakeholders in the healthcare ecosystem.

Ethical Issues of Social Media Usage in Healthcare.

OBJECTIVE: Social media, web and mobile technologies are increasingly used in healthcare and directly support patientcentered care. Patients benefit from disease self-management tools, contact to others, and closer monitoring. Researchers study drug efficiency, or recruit patients for clinical studies via these technologies. However, low communication barriers in socialmedia, limited privacy and security issues lead to problems from an ethical perspective. This paper summarizes the ethical issues to be considered when social media is exploited in healthcare contexts. METHODS: Starting from our experiences in social-media research, we collected ethical issues for selected social-media use cases in the context of patient-centered care. Results were enriched by collecting and analyzing relevant literature and were discussed and interpreted by members of the IMIA Social Media Working Group. RESULTS: Most relevant issues in social-media applications are confidence and privacy that need to be carefully preserved. The patient-physician relationship can suffer from the new information gain on both sides since private information of both healthcare provider and consumer may be accessible through the Internet. Physicians need to ensure they keep the borders between private and professional intact. Beyond, preserving patient anonymity when citing Internet content is crucial for research studies. CONCLUSION: Exploiting medical social-media in healthcare applications requires a careful reflection of roles and responsibilities. Availability of data and information can be useful in many settings, but the abuse of data needs to be prevented. Preserving privacy and confidentiality of online users is a main issue, as well as providing means for patients or Internet users to express concerns on data usage.

Surveillance and Outbreak Response Management System (SORMAS) to support the control of the Ebola virus disease outbreak in West Africa.

In the context of controlling the current outbreak of Ebola virus disease (EVD), the World Health Organization claimed that 'critical determinant of epidemic size appears to be the speed of implementation of rigorous control measures', i.e. immediate follow-up of contact persons during 21 days after exposure, isolation and treatment of cases, decontamination, and safe burials. We developed the Surveillance and Outbreak Response Management System (SORMAS) to improve efficiency and timeliness of these measures. We used the Design Thinking methodology to systematically analyse experiences from field workers and the Ebola Emergency Operations Centre (EOC) after successful control of the EVD outbreak in Nigeria. We developed a process model with seven personas representing the procedures of EVD outbreak control. The SORMAS system architecture combines latest In-Memory Database (IMDB) technology via SAP HANA (in-memory, relational database management system), enabling interactive data analyses, and established SAP cloud tools, such as SAP Afaria (a mobile device management software). The user interface consists of specific front-ends for smartphones and tablet devices, which are independent from physical configurations. SORMAS allows real-time, bidirectional information exchange between field workers and the EOC, ensures supervision of contact follow-up, automated status reports, and GPS tracking. SORMAS may become a platform for outbreak management and improved routine surveillance of any infectious disease. Furthermore, the SORMAS process model may serve as framework for EVD outbreak modeling.

How to exploit twitter for public health monitoring?

OBJECTIVES: Detecting hints to public health threats as early as possible is crucial to prevent harm from the population. However, many disease surveillance strategies rely upon data whose collection requires explicit reporting (data transmitted from hospitals, laboratories or physicians). Collecting reports takes time so that the reaction time grows. Moreover, context information on individual cases is often lost in the collection process. This paper describes a system that tries to address these limitations by processing social media for identifying information on public health threats. The primary objective is to study the usefulness of the approach for supporting the monitoring of a population's health status. METHODS: The developed system works in three main steps: Data from Twitter, blogs, and forums as well as from TV and radio channels are continuously collected and filtered by means of keyword lists. Sentences of relevant texts are classified relevant or irrelevant using a binary classifier based on support vector machines. By means of statistical methods known from biosurveillance, the relevant sentences are further analyzed and signals are generated automatically when unexpected behavior is detected. From the generated signals a subset is selected for presentation to a user by matching with user queries or profiles. In a set of evaluation experiments, public health experts assessed the generated signals with respect to correctness and relevancy. In particular, it was assessed how many relevant and irrelevant signals are generated during a specific time period. RESULTS: The experiments show that the system provides information on health events identified in social media. Signals are mainly generated from Twitter messages posted by news agencies. Personal tweets, i.e. tweets from persons observing some symptoms, only play a minor role for signal generation given a limited volume of relevant messages. Relevant signals referring to real world outbreaks were generated by the system and monitored by epidemiologists for example during the European football championship. But, the number of relevant signals among generated signals is still very small: The different experiments yielded a proportion between 5 and 20% of signals regarded as "relevant" by the users. Vaccination or education campaigns communicated via Twitter as well as use of medical terms in other contexts than for outbreak reporting led to the generation of irrelevant signals. CONCLUSIONS: The aggregation of information into signals results in a reduction of monitoring effort compared to other existing systems. Against expectations, only few messages are of personal nature, reporting on personal symptoms. Instead, media reports are distributed over social media channels. Despite the high percentage of irrelevant signals generated by the system, the users reported that the effort in monitoring aggregated information in form of signals is less demanding than monitoring huge social-media data streams manually. It remains for the future to develop strategies for reducing false alarms.

Web science in medicine and healthcare.

OBJECTIVES: Medical social-media provide a new source of information within information gaining contexts. Facts, experiences, opinions or information on behaviour can be found in the medical web and could support a broad range of applications. The intention of this Focus Theme is to bring the existing research together and to show the possibilities, challenges and technologies for Web Science in medicine and healthcare. METHODS: This editorial provides an overview on the landscape of medical social-media and their possibilities in supporting healthcare. Further, it summarizes the three papers included in this Focus Theme. RESULTS AND CONCLUSIONS: The three papers of this Focus Theme consider different aspects of Web Science in medicine which are 1) detection of drug interactions from social media, 2) inferring community structures from online forums and 3) improving access to online videos through assignment of SNOMED CT terms. All three papers show the potential of medical social-media in supporting health information gathering processes from the web. However, several issues still need to be addressed in future: Methods are necessary for identifying high quality information from the medical web as well as for processing the language that is used by social media users to report about their symptoms, diseases and other health issues.

An architecture for diversity-aware search for medical web content.

OBJECTIVES: The Web provides a huge source of information, also on medical and health-related issues. In particular the content of medical social media data can be diverse due to the background of an author, the source or the topic. Diversity in this context means that a document covers different aspects of a topic or a topic is described in different ways. In this paper, we introduce an approach that allows to consider the diverse aspects of a search query when providing retrieval results to a user. METHODS: We introduce a system architecture for a diversity-aware search engine that allows retrieving medical information from the web. The diversity of retrieval results is assessed by calculating diversity measures that rely upon semantic information derived from a mapping to concepts of a medical terminology. Considering these measures, the result set is diversified by ranking more diverse texts higher. RESULTS: The methods and system architecture are implemented in a retrieval engine for medical web content. The diversity measures reflect the diversity of aspects considered in a text and its type of information content. They are used for result presentation, filtering and ranking. In a user evaluation we assess the user satisfaction with an ordering of retrieval results that considers the diversity measures. CONCLUSIONS: It is shown through the evaluation that diversity-aware retrieval considering diversity measures in ranking could increase the user satisfaction with retrieval results.

Semantic structuring of and information extraction from medical documents using the UMLS.

OBJECTIVES: This paper introduces SeReMeD (Semantic Representation of Medical Documents), a method for automatically generating knowledge representations from natural language documents. The suitability of the Unified Medical Language System (UMLS) as domain knowledge for this method is analyzed. METHODS: SeReMeD combines existing language engineering methods and semantic transformation rules for mapping syntactic information to semantic roles. In this way, the relevant content of medical documents is mapped to semantic structures. In order to extract specific data, these semantic structures are searched for concepts and semantic roles. A study is carried out that uses SeReMeD to detect specific data in medical narratives such as documented diagnoses or procedures. RESULTS: The system is tested on chest X-ray reports. In first evaluations of the system's performance, the generation of semantic structures achieves a correctness of 80%, whereas the extraction of documented findings obtains values of 93% precision and 83% recall. CONCLUSIONS: The results suggest that the methods described here can be used to accurately extract data from medical narratives, although there is also some potential for improving the results. The proposed methods provide two main benefits. By using existing language engineering methods, the effort required to construct a medical information extraction system is reduced. It is also possible to change the domain knowledge and therefore to create a more (or less) specialized system, capable of handling various medical sub-domains.

The effect of nicotine on perceptual, ocular motor, postural, and vegetative functions at rest and in motion.

Nicotine has wellknown, unpleasant side effects, e.g., transient dizziness, nausea, and nicotine-induced nystagmus (NIN). To investigate factors influencing these effects, we addressed three questions: (1) Is the intensity of dizziness, nausea, NIN, and unsteadiness dependent on nicotine dosage? (2) Does the intensity of perceptual, ocular motor, vegetative effects, and postural imbalance correlate? (3) Do visual or vestibular motion stimuli produce and/or aggravate distressing dizziness and nausea? Sixty healthy non-smokers or occasional smokers participated; 40 were tested once before and six times after application of a nicotine nasal spray in doses of 1 mg or 2 mg with or without motion stimulation; 20 received a placebo nasal spray. Plasma nicotine concentrations were significantly related to nicotine dosage. Dizziness, nausea, NIN, and unsteadiness also depended on the nicotine dosage (p < 0.01).Nicotine blood concentration was a better predictor for the temporal dependence of nystagmus than nicotine dosage. Dizziness correlated highly with nausea (R = 0.63, p < 0.001). The degree of nicotine-induced nausea significantly correlated with postural imbalance. The time course of postural sway differed according to nicotine dosage and gender: for women, there was no clear relationship between sway magnitude and nicotine dosage, while men showed increased sway with higher dosage. Motion stimulation increased nicotine-induced dizziness and nausea, but did not significantly influence NIN or postural imbalance. Our data support the view that all measured adverse effects reflect dose-dependent nicotine-induced vestibular dysfunction. Additional motion stimulation aggravates dizziness and nausea, i.e., nicotine increases sensitivity to motion sickness.