Crowdsourced health research studies are the nexus of three contemporary trends: citizen science (non-professionally trained individuals conducting science-related activities), crowdsourcing (use of web-based technologies to recruit project participants), and medicine 2.0 / health 2.0 (active participation of individuals in their health care particularly using web 2.0 technologies). Crowdsourced health research studies have arisen as a natural extension of the activities of online health social networks and communities. Studies may be researcher-organized or participant-organized. Professional researchers crowdsource cohorts from health social networks for the conduct of traditional studies. In contrast, participants organize their own health research studies through health social networks and health collaboration communities; these communities are created for the purpose of self-experimentation and for the investigation of shared health concerns together in groups. Before embarking on a narrative overview of crowdsourced research studies, I will first consider the definition of citizen science, crowdsourcing, medicine 2.0 / health 2.0, and crowdsourced health research studies. I will then provide an overview of crowdsourced health research studies that have been conducted by professional researchers and/or participants. Finally, I will discuss the limitations of these methods, and offer conclusions.
Citizen science is the conduct of science-related activities by individuals who have no formal training in a field specific to the topic of investigation. Citizen science practitioners may include laypersons, scientists, or professionals trained in other fields. Citizen science projects have been in existence for hundreds of years; the professional scientist is a relatively recent incarnation. One prominent example of citizen science is the National Audubon Society’s annual Christmas bird watch, in its 112th year with tens of thousands of participants in 2011 [1
]. Another high-profile project is Galaxy Zoo, where over 250,000 individuals have annotated astronomical data from the Sloan Digital Sky Survey, surprising project organizers by completing 50 million images in the first year as opposed to an anticipated 1 million [2
]. A citizen science referral and advocacy website, SciStarter [3
], listed 340 projects for participation as of January 2012 in 20 areas ranging from the environment to health. The Citizen Science Alliance [4
] is another industry group which supports citizen science and coordinates Galaxy Zoo and other astronomy-related projects.
Crowdsoucing is the practice of obtaining participants, services, ideas, or content by soliciting contributions from a large group of people, especially via the Internet [5
]. Canvasing vast numbers of individuals through an open call facilitates self-selection. In particular, there is potential for crowdsourcing to capitalize on the input of interested and fit individuals who have the best ideas and bring a diverse set of skills and backgrounds to bear on the current task. Some notable examples of successfully crowdsourced projects are discussed in The Wisdom of Crowds
] and Wikinomics
Medicine 2.0 / Health 2.0 and Participatory Medicine
Some of the early definitional discussions of medicine 2.0 / health 2.0 (which are largely used synonymously [8
]) occurred in 2008 and focused on the deployment of social media in the health context; that is, that medicine 2.0 / health 2.0 is the use of web 2.0 tools (eg, blogs, podcasts, tagging, search, wikis, video) by health care actors to improve collaboration and personalize health care [8
In 2010, a related concept, participatory medicine, was introduced to emphasize the active participation of individuals: “This new definition devised by the board of the Society of Participatory Medicine is a movement in which networked patients shift from being mere passengers to responsible drivers of their health, and in which providers encourage and value them as full partners [10
These definitions have helped to undergird the medicine 2.0 / health 2.0 and participatory medicine movement. A Pew Internet study found that 27% of US Internet users had tracked health data online and 18% had sought to locate others with similar health concerns via the Internet [11
]. At present, individuals have the opportunity to self-manage their health using web 2.0 tools, smartphone health applications, online personal health records, and health social networks. Health social networks, essentially Facebook or LinkedIn for health interest areas, are online communities where individuals may find and discuss information about conditions, symptoms, and treatments; provide and receive support; enter and monitor data; and join health studies [12
]. Health social networks exemplify the predicted progression of engagement in online communities, escalating in three stages from information-sharing, to cooperating, to participating in collaborative action [13
]. As of January 2012, some of the largest health social networks for patients are MedHelp (claiming over 12 million monthly visitors), PatientsLikeMe, DailyStrength, Tudiabetes, CureTogether, and Asthmapolis; and for physicians, Sermo, Ozmosis, and RadRounds [14
Crowdsourced Health Research Studies
The nexus of these trends—citizen science, crowdsourcing, and medicine 2.0 / health 2.0—is crowdsourced health research studies. One indication of the relative newness, growth, and interest in this area is the exponential rise in recent Internet activity in crowdsourced health research. In 2011, 1,920,000 results were returned for a Google search of the terms ‘crowdsourcing and health’; in 2010 and 2009 the comparative figure was 669,000 and 318,000 respectively. In January 2012, the term ‘crowdsourcing’ in a PubMed search yielded 16 publications, 13 of which were published in 2011.
Crowdsourced health research studies may be a blend of crowdsourcing and citizen science. In addition, these terms are often used interchangeably. This lack of precision in language is an artifact of the relative newness of these concepts that are being defined through use. Within crowdsourced studies, participants are recruited via crowdsourcing (eg, recruited online with a website or an open call to a large potential audience using Internet-related technologies). In crowdsourced studies participants could be subjects not performing any science-related activities themselves, and therefore would not be considered citizen scientists. The use of the term ‘citizen scientist’ denotes the conduct of science-related activities by participants. Examples are citizen scientists being crowdsourced to annotate astronomical data in the Galaxy Zoo projects, or when an online crowd analyzed scientific images of cells in a tuberculosis study [15
]. There may be many other permutations. Citizen scientists might be crowdsourced for idea generation and hypothesis formation, data collection, results analysis, results dissemination, and/or study funding (‘crowdfunding’).
In addition to opportunity for participant engagement, crowdsourced research may be different to traditional studies in other ways. First, crowdsourced research provides opportunity for more levels of openness and privacy, as participants decide what data to share with whom. One potential result is that there is less regulated protection of research subjects. Instead individuals take responsibility for informing themselves (possibly in consultation with physicians) about self-experimentation or study participation. Second, within crowdsourced research the rewards may accrue more directly to study participants and health communities as opposed to study funders in the more traditional model. Third, funding may come from alternative sources such as academia, industry, patient advocacy groups, research foundations, social venture capital, crowdfunding, and self-funding.
In this article, the term ‘crowdsourced health research studies’ is used to indicate that health study participants are recruited with crowdsourcing (eg, Internet-based) techniques. Study participants may or may not be acting as citizen scientists (ie, conducting science-related activities).
Participant Motivations and Expectations
Individuals have a variety of motivations for participating in crowdsourced health research studies. On a personal level, they may be drawn by natural curiosity, wanting to tinker and test hypotheses in a health interest area. Individuals taking a broader societal perspective may wish to participate in, contribute to, impact on, and at times conduct, projects that are outside the scope of traditional research. Another dimension of the broader societal perspective is how individuals may view themselves in relation to society. There is a developing notion of biocitizenry: that being a citizen scientist, and sharing personal health information, or using it as a currency for gaining access to studies could be considered acts of citizenship [16
The direct and personal connection that individuals have to health makes crowdsourced health research distinct from other crowdsourced studies. Perhaps due to this personal connection, and the ease of self-tracking and experimenting with interventions, the role of the participant is expanding more quickly in health compared to other citizen science areas. In particular, the participant is engaging not just as a provider of outsourced data collection, but also helping with data analysis, and possibly the design and conduct of studies. Participant expectations of their engagement in crowdsourced studies are also different as they make demands on study organizers to return study data, provide interpretive personalized recommendations, and want the ability to communicate with other study participants [18
Types and Methods of Crowdsourced Health Research Studies
Crowdsourced health research studies can be researcher-organized or participant-organized. Researcher-organized studies are typically traditional studies organized by institutionally-trained researchers using crowdsourced health social network cohorts or crowdsourced data as the input or research focus; for example, studies organized by PatientsLikeMe [19
] and 23andMe [20
]. Participant-organized studies are usually designed and operated by citizen scientists; for example, those conducted by PatientsLikeMe patients, DIYgenomics citizen scientists [21
], and Quantified Self individual experimenters [22
The research methods available in crowdsourced health research studies are detailed in .
Research methods employed to date in crowdsourced health research studies.
Principal Aim of This Study
The principal aim of this analysis is to provide an overview of crowdsourced health research studies. The narrative will characterize the nature of current activity and highlight differences between crowdsourcing and traditional methodologies.