Digital cohorts within the social mediome to circumvent conventional research challenges?

June 21, 2017|GI and Hepatology News

Anand Kulanthaivel, PhD;Rachel Fogel, MD;Josette Jones, PhD;Craig Lammert, MD

We are becoming comfortable with the concept of a sharing economy, where resources are shared among many individuals using online forums. Whether activities involve sharing rides (Uber, Lyft, and others), accommodations (Airbnb), or information (social media), underlying attributes include reduced transactional costs, enhanced information transparency, dynamic feedback, and socialization of opportunity. As health care systems realize that they are changing from direct-to-business to a direct-to-customer model, their ability to connect directly with individuals will become a foundational strategy.

This month’s column introduces us to social media as a research tool. Information derived from social media sites can be harvested for critical clinical information (the Centers for Disease and Control and Prevention tracks the spread of influenza using social media analytic tools), research data (patient preferences), and as a recruitment method for clinical studies. Kulanthaivel and colleagues have described their experiences and literature review to help us imagine new ways to collect data at markedly reduced transaction costs (compared to a formal clinical trial). While there are many cautions about the use of social media in your practice or research, we are only beginning to understand its potential.

John I. Allen, MD, MBA, AGAF

Editor in Chief

Medical knowledge, culminating from the collection and translation of patient data, is the primary objective of the clinical research paradigm. The successful conduct of this traditional model has become even more challenging with expansion of costs and a dwindling research infrastructure. Beyond systemic issues, conventional research methods are burdened further by minimal patient engagement, inadequate staffing, and geographic limitations to recruitment. Clinical research also has failed to keep pace with patient demands, and the limited scope of well-funded, disease-specific investigations have left many patients feeling disenfranchised. Social media venues may represent a viable option to surpass these current and evolving barriers when used as an adjunctive approach to traditional clinical investigation.

The term social media (SM) most commonly refers to relatively public, Internet-based communication platforms that enable users to consume and disseminate information. The most popular SM venues currently include Facebook, Twitter, YouTube, and independent online forums (Table 1). These digital platforms support sharing multiple forms of media including text, images, and videos between users who interact within a wide realm of medical groups and genres (e.g., specific diseases, symptoms, and so forth). This collective mediome¹ is a relatively untapped resource for clinical study, but research applications using SM methodology have begun to produce real study benefits in an array of diseases. Effective implementation of this technology by interested investigators will require an in-depth working knowledge of digital venues beyond their own online social presence. A firm grasp of these applications can enable contact with previously out-of-reach study participants, promote patient engagement and disease investment, and cultivate a community of interacting patients and researchers. This data-rich resource already has facilitated various aspects of biomedical studies, including dissemination of epidemiologic surveys,² direct recruitment into clinical trials,³ collection of biologic samples,⁴ and extraction of patient-provided data, all within SM platforms.⁵

Advantages and pitfalls in social media research

SM is a new frontier containing a wide spectrum of clinical and qualitative data from connected users (patients). Collection and examination of either individuals’ or groups’ SM information use can provide insight into qualitative life experiences, just as analysis of biologic samples can enable dissection of genetic disease underpinnings. This mediome is analogous to the human genome, both in content and utility.¹ Analyzing data streams from SM for interpersonal interactions, message content, and even frequency can provide digital investigators with volumes of information that otherwise would remain unattainable.

There are many advantages to scientific interrogation of the social mediome, specifically because applications within SM have no physical bounds, encourage information exchange among stakeholders, and work in real time. Patient access to clinical studies and individual investment can limit both conventional and unconventional approaches to research. However, SM far exceeds the geographic limitations determined by the location of patients and academic systems, thus expanding the available recruitment population dramatically (Table 1). Patient-to-patient communication is facilitated by the format of most SM venues (Facebook and other Internet forums), thus creating an enriched collection of disease testimonies, symptom discussions, and treatment effects. In fact, patients frequently use SM to form online support groups to share experiences with similarly afflicted patients and families. These groups and their documented communications are valuable because qualitative patient data can provide a high resolution of variable patient metrics to investigators.⁵ Finally, data collection from SM can occur continuously in real time and with little cost. Facebook, Twitter, and YouTube are free to use, although online Internet forums may incur small monetary investments (typically $15–$50 per month). Because of study heterogeneity, it remains challenging to compare costs between a SM-based research study and a similar traditional clinic-based approach. However, historically, costs incurred to SM research pioneers have been dramatically lower than cost estimates of conventional approaches in the clinic.⁶

Several limitations and potential risks of SM for medical research should be addressed, including the possible compromise of privacy and confidentiality, the use and dissemination of medical advice and information, potential demographic biases, and a required trust of the investigator by patients. Many of these challenges can be similar to traditional methods, however, as in the conventional model, careful management can drastically reduce unwanted study issues.

The risk of Health Insurance Portability and Accountability Act violations must be considered seriously in the context of patient–researcher interactions on SM. Because of the relatively public nature of these venues, patient confidentiality may be at risk if patients choose to divulge personal medical information. However, if proper protective measures are taken to ensure that the venue is secure (e.g., a private or closed group on Facebook or a by-invitation-only online Internet forum), and the researcher vets all patients who request entrance into the group, this risk may be minimized. Moreover, to further reduce any legal liability, the researcher should not provide any medical advice to patients who participate in a SM study. The drive to provide medical direction in study patients with clinical need may be strong because collaborative relationships between investigator and patients are likely to form. Furthermore, digital access to investigators on SM commonly becomes easy for patients. Safe approaches to communication could include redirecting patients to consult with their own doctor for advice, unbiased dissemination of disease-specific educational materials, or depiction of only institutional review board–approved study materials.^7,8

An investigator-driven interactive community (e.g., Facebook group) may bolster patient involvement in SM studies and help facilitate disease-specific research. However, because most SM venues facilitate patient–patient interactions, misleading or incorrect medical information may be spread quickly between patients and could be misconstrued as official medical advice.⁹ To mitigate this, a researcher or trusted study personnel must actively moderate the digital venue to prevent the spread of counterproductive information.⁷ Although it is not possible to completely eliminate the sharing of unverified information, regular moderation will reduce the potentially negative impact of such sharing.

The perception that only younger populations use SM may appear to be a significant limitation for its implementation in clinical research. However, this limitation is rapidly becoming less significant because recent studies have shown that the use of SM has become increasingly common among older adults. As of 2014, more than half of the US adult population used Facebook, including 73% and 63% of Internet-using adults ages 30–49 and 50–64 years, respectively.¹⁰ SM may not be suitable for all diseases, however, there is likely significant demographic overlap for many disease populations.

Finally, it is imperative for researchers to gain the trust of patients on SM to effectively use these venues for research purposes. Because patient–researcher interaction does not occur face-to-face on these platforms, gaining the trust of patients may be more difficult than it would be in a clinical setting. Thus, patient–patient and patient–researcher communications within SM platforms must be cultivated carefully to instill participant confidence in the research being performed on their behalf. One of the authors (C.L.) has established an SM educational model for this exchange.⁴ Specifically, he provides patients with a distillation of current field research by posting updates in a research-specific Facebook group and on Twitter. This model not only empowers patients with disease education, it also solidifies the importance of patient investment in disease-specific research. Furthermore, invested patients bring ideas to research, take a more educated and proactive role in their care team, and, ultimately, return to seek more study involvement.

A number of studies have shown SM methods to be an effective means of collecting data and improving quality of care for patients. One randomized controlled trial found that the use of SM to disseminate instructional information to patients alongside the traditional educational pamphlet increased patients’ quality of bowel preparation for colonoscopies.¹¹ Another study successfully used the Crohn’s and Colitis Foundation of America Partners Internet Cohort of more than 14,000 patients to examine factors associated with fiber consumption in inflammatory bowel disease and whether fiber was associated with disease flares.²In addition, several studies have assessed the roles of mobile applications, remote health sensors, and telemedicine in research and patient care and have found that these tools are effective at providing more complete care in real time and with decreased costs.¹² Riaz and Atreja¹³ noted that the most significant barrier to the use of these techniques in research and patient care is provider acceptability, in addition to the need for strict Health Insurance Portability and Accountability Act compliance to ensure patient confidentiality. Keeping these limitations in mind, the aforementioned studies lend significant support to the effective use of SM as adjuncts to traditional clinical investigation.