From Community Analysis to Prototype: Creating an Online Matchmaker for Inflammatory Bowel Disease Patients

  • Jermain Kaminski


Since about five years, the MIT Center for Collective Intelligence, Cambridge, U.S., and the Chronic Collaborative Care Network (C3N) at Cincinnati Children’s Hospital, Cincinnati, U.S., have been working together to improve care for inflammatory bowel disease (IBD) patients by harnessing methods of computational social science. The goal of this contribution is (1) to present an approach in measuring communication patterns and sentiments within online communities of IBD patients, (2) to analyze the enablers for a better connectedness of community members, and (3) to introduce a prototype application of a collective intelligent online network for IBD patients, named “YouMeIBD”. The mobile application, developed within an interdisciplinary student class at MIT and four other universities, aims to improve the connectedness, well-being and diffusion of innovations in a community of IBD patients.


Social network analysis Sentiment analysis Coolhunting Patient community Collaborative innovation network Inflammatory bowel disease (IBD) 

1 Introduction

Innovation and care for patients become increasingly decentralized and digitalized in the medical system of the future. In fact, disease discovery, treatment and social support are already taking place online, mediated through an increasing number of self-tracking platforms and mobile applications. With more and more patients providing their data and proposing new ideas in a collaborative system, health innovation becomes a bottom-up process, reframing the current top-down logic of health innovation. This article presents an example analysis on the inner workings of a patient online community and shows how data analysis can contribute to the development of a new online patient platform aiming to improve the diffusion of innovation.

In their remarkable book “Connected”, Christakis and Fowler point out: “For thousands of years, social interactions were build solely on face-to-face communication. But technology changed this with the […] ways of communicating person-to-person at a distance.” Christakis and Fowler (2011)

Truly, new technologies enabled individuals to peer beyond an individual’s social horizon and to create new connections independently from physical distances; this holds true for patients communicating online. As Wagner’s Chronic Care Model (Wagner 2001) further suggests, the best outcomes in health care depend on decision support, effective delivery systems, information systems and active communication among health care teams and patients (cf. Gloor et al. 2011). Virtual communities may support knowledge about a disease and can improve the self-management process through a number of mechanisms such as enhanced self-care, the support of patient–physician interactions, structured information, the improvement of emotional well-being and individually perceived disease control (cf. Wellman 1992; Gallant 2003; Eysenbach et al. 2004; Elliott et al. 2007).

The analysis shown in this article concentrates on an online community of inflammatory bowel disease (IBD) patients. IBD is a chronic disease further explained in Sect. 2), affecting more than 3.8 million people in the U.S. and Europe combined. The patients, mostly diagnosed at an age between 15 and 35 (Loftus 2007; Economou and Pappas 2008; CCFA 2015), are particularly interested in encouragement through online social networks and finding health information online (Wicks et al. 2010, 2011). Indeed, the young age and Internet connectivity of the patient group provides a key asset both for the community analysis and prototype development.

The key methods described in this article emerge from the field of computational social science (Lazer et al. 2009). Computational social science became a foundation in the understanding of community dynamics and their inherent economics, which also counts for health related networks (Valente 2010; Christakis and Fowler 2011). It is a relatively new research discipline that serves to measure and understand social behavior and phenomena with the help of analyzing digital traces. The research field emerges with the increasing capacity to collect and analyze big data. One underlying core belief of computational social science is that social and economic behaviors can be found in data patterns.

Social network analysis, a sub-discipline of computational social science, focuses on measuring and visualizing the connections among people within social organizations and contributes toward understanding their communication patterns, structures and sentiments (Wellman 1988; Wasserman and Faust 1994; Portier et al. 2013). Revealing the social physics (Pentland 2014) behind transactions in health networks can be a key success factor in identifying levers that “make or break” a successful network, in particular with regard to the diffusion of innovations (Rogers 2003[1962]). In a sense, like the eye is said to be a window to the soul, network patterns might be a window to the soul of a community.

The empirical foundation of the IBD online community analysis presented in Sect. 3 is an anonymized and subsampled dataset of five public IBD-related Facebook groups, covering data from 2006 to 2011. Analyzing an online group on IBD is valuable, because it might not only mirror real-world friendship networks but also helps to identify key discussion topics. While Facebook groups might be the first stop for most patients seeking online connections, they are also alternatives to patient-oriented websites such as PatientsLikeMe ( or the IBD-focused Crohnology (, both providing excellent information on self-tracking and medication effects. That being said, no online platform can be a jack-of-all-trades. For instance, platforms with a focus on self-tracking and data may not simultaneously fulfill a patient’s desire of social interaction and personal connectedness. Admittedly, it is a big challenge to unify all of a chronic patient’s needs in just one platform.

Section 4 presents an analysis on the diffusion of innovation for the case of whipworm therapy, a potentially effective medication for IBD patients. Since 2005, experimental and clinical data sustain the idea that helminths like the whipworm might provide protection against the disease. By means of coolhunting (Gloor and Cooper 2007) the term “whipworm”, it was possible to investigate the presence and connotation of the term both in the World Wide Web and in public Facebook group discussions. The analysis was highly contributory, because it demonstrated both how the external information (www) on whipworm therapy transfers into the network (Facebook) and how the network structure within Facebook groups affected the diffusion of information.

Section 5 presents a prototype mobile application, targeting the identified issues in the diffusion of innovations within the IBD online community (Baebler et al. 2011; Fuehres et al. 2011; Kaminski et al. 2012). The mobile application “YouMeIBD” applies matching algorithms known from dating websites to improve the connectedness of patients both virtually and locally. The ultimate goal of the application is to support a better diffusion of innovation and to enable a collaborative innovation network (COIN) (Gloor and Cooper 2007).

After all, the main purpose of this contribution is to provide practitioners with information on how the analysis of an existing community can contribute towards the development of a new platform. That being said, the content focuses intentionally more on the bigger picture than the small details and metrics of our analysis.

2 About Inflammatory Bowel Disease

In 1981, IBD researcher Cooke states “Despite a view that Crohns disease is a frightening diagnosis for both doctor and patient (Gazzard et al. 1978), the long term prognosis is good” (Cooke 1981). In 2015—more than three decades of research later—it can be stated that this finding did not lose its expressiveness. The understanding of IBD is still incomplete, but advances in medical therapies and surgical techniques stepwise improved the patients’ quality of life.

Inflammatory bowel diseases encompass the two main classifications of Crohn’s disease (also known as regional enteritis or Morbus Crohn’s, named after American gastroenterologist Burrill Bernard Crohn, cf. Crohn et al. 1984) and ulcerative colitis. Both diseases are characterized by chronic intestinal inflammation of the digestive system (Isaacs et al. 2005; Ruyssers et al. 2008) and are associated with high morbidity and reduced quality of life (McLeod et al. 1991; Cohen 2002; Longobardi et al. 2003; Kappelman et al. 2007). Pathophysiological, Crohn’s disease can affect any part of the digestive system, while the most common site is the colon. Compared to ulcerative colitis, Crohn’s disease is often segmental and the rectum is commonly spared. Patients with IBD often suffer from limited gastrointestinal motility leading to symptoms such as abdominal pain, fatigue, gastro-intestinal obstruction, bleeding and weight loss, concomitant with changes in the enteric nervous system (Geboes and Collins 1998; Ruyssers et al. 2010).

An IBD like Crohn’s disease seems to run in some families. While males and females seem equally affected, Crohn’s disease may occur in people of all ages, but it is predominantly a disease of adolescents and young adults, mainly in the age between 15 and 35 (Loftus 2007; Economou and Pappas 2008; Sonnenberg 2009; Thukkani et al. 2011; CCFA 2015). In fact, 10 % of those affected are less than 18 years old (CCFA 2015). While some patients have long periods of remission, i.e. periods when they are free of symptoms, the chronic nature of the disease can induce a permanent psychological pressure for the patient (Ruyssers et al. 2010; Sajadinejad et al. 2012).

Although the causes of the disease are still subject to research, the current hypothesis claims that IBD results from an uncontrolled (auto-) immune response to the normal gut flora (Xavier and Podolsky 2007; Ruyssers et al. 2008, 2010). Both genetic and environmental factors may contribute to the damaging mucosal immune response (Fiocchi 1998; Ruyssers et al. 2008). For the etiology of the disease, it is suggested that the lack of exposure to intestinal parasites like helminths, as a result of enhanced living standards and medical conditions, modulates the development of the immune system and thereby increases the risk of immune diseases in certain populations (Weinstock et al. 2002; Lashner and Loftus 2006; Garn and Renz 2007; Ruyssers et al. 2008; Strachan 2000). This assumption is supported by epidemiological studies that show an inverse relation between the frequency of helminth colonization and the prevalence of IBD (Weinstock et al. 2002; Elliott and Weinstock 2012). Correspondingly, while the incidence of IBD has steadily been increasing in the developed world since 1950 (Lakatos 2006; Ruyssers et al. 2008; Burisch and Munkholm 2013), it has a geographic trend, repeatedly been reported in northern regions, especially in North America and Europe (Loftus 2004; Probert and Brown 2008; Thukkani et al. 2011). While Alic (2000) and Bernstein et al. (2001) find that the incidence of Crohn’s disease is highest in well-developed countries, the hygiene hypothesis states this is directly related to higher hygienic standards in respective countries (Strachan 2000; Ruyssers et al. 2008).

In terms of frequency, it is estimated that about 1.6 million U.S. Americans suffer from IBD. Compared to the U.S. population, it means that 1 out of 200 people suffer from IBD (Loftus 2004; Isaacs et al. 2005; Kappelman et al. 2007; CDC 2014). Approximately 70,000 new cases are diagnosed each year (Loftus 2007; CCFA 2015). For Europe, the number of people with IBD is estimated to be about 2.2 million (Burisch et al. 2013; Loftus 2004).

The treatment of the disease usually includes drugs, nutrition supplements, surgery or a combination of all these options (Duchmann and Zeitz 2005). By tendency, surgical treatment (removals) has been reserved for refractory disease and other complications. According to Van Assche et al. (2010), about 80 % of patients have to undergo surgery at some point. Although data regarding the health care costs of inflammatory bowel disease is limited, studies estimate the annual disease-attributable costs of IBD in the U.S. to be in a range from about USD 14 billion to USD 30 billion (Kappelman et al. 2007; Park and Bass 2011; CCFA 2015). In sample studies of U.S. children and adults with inflammatory bowel disease, researchers find that the mean treatment costs for Crohn’s disease range from about USD 5000 (Kappelman et al. 2007; CCFA 2015) to USD 19,000 (Gibson et al. 2008; Yu et al. 2008; Venu and Cohen 2011) per year, depending on the applied therapy spectrum. Observing the drug product range, Isaacs et al. (2005) estimated that the number of products under study for the treatment of IBD increased from three products and one target in 1993 to more than 30 products and more than 10 targets in 2005. According to an industry report by Datamonitor (2010) and a research summary by Dutton (2013), new drugs could potentially drive sales of IBD related drugs from USD 4.26 billion in 2011 to USD 5–7 billion in 2019.

Summarizing the circumstances that inflammatory bowel disease brings in its train, both the incentive to improve the life of patients, as well as the economic burden of IBD encourage to improve the situation. Considering the presented state of research as well as internal surveys with patient advisory councils at C3N (Baebler et al. 2011), the demand for online and computer-mediated medical intelligence in IBD is very high.

3 Analyzing the IBD Patient Community Network (2010–2011)

In preparation of the netnographic patient community analysis, the five biggest English speaking public IBD-related Facebook groups were identified. The initial screening revealed that more than 40 IBD-related Facebook groups were active that time. Similar to an individual’s Facebook profile page, Facebook group pages enable users to connect to these entities by “liking” the page or becoming a group member. Once a user liked or joined a group, he received updates on new links and discussions within the group via the Facebook News Feed.

For an initial analysis of the underlying network structure within the Facebook groups, each a sample of 500 random and anonymized members was selected. The sample represented each about 3–10 % of the network and allowed a first insight into the connectedness of group members, i.e. the friendship network. The visualization of networks was conducted with Gephi, an open-source visualization platform (, cf. Bastian et al. 2009). In terms of privacy, it should be highlighted that the analysis was exclusively focused on the group’s network and discussion board as a whole.1 The study did explicitly not encompass an analysis of single group members or individual profiles. There was no information collected that could be linked to a living individual, just as there was no interaction between the investigator and subject or any manipulative experimental setting. It was ensured to comply with data protection and security with regard to individual privacy and research ethics as approved by the Institutional Review Boards at C3N (Cincinnati Children’s Hospital).

Figure 1 exemplarily shows the network structure in one of the five Facebook groups, depicting a network with very low connectedness among members (low-density network). The figure is a sociogram, i.e. a graph that visually represents the interrelationships within a group, where actors are visualized as nodes (points) and edges (lines) that represent a connection between actors. The bigger a node, the more connections (degrees) it has. The visualization suggests only few connections with star-like networks (one central connector) as opposed to a more desirable “galaxy” of interconnected nodes. The analysis of anonymous IDs also revealed one other insight: 97.5 % of members belonged to only one of our selected Facebook groups on IBD. This result was quite surprising as one might assume that patients “leverage” the risk of missing information by joining several groups at the same time. However, it should be added that this argument is certainly limited as we only observe a subsample.
Fig. 1

Friendship network structure of a public IBD-related group with more than 24,000 members (subsample of 500 members)

With regard to connectedness, Metcalfe’s law (Metcalfe 1995) suggests that the value of a communications network is proportional to the square of the size of the network. However, networks tend to become less dense when they grow in size. This means that there is an inverse correlation between the overall connectedness (density) and the number of group members (network size). As a network grows in size, it becomes more difficult to establish stable social relationships (cf. Dunbar 1992; Wellman 2012). Correspondingly, as network density decreases, “structural holes” are likely to open and can be a source of information loss (Burt 2001). This effect is particularly important when information has to travel from highly connected members to less connected members, mostly in a network’s periphery (or vice versa).

Aside from the low-density network structure of the groups, further insights on the communication pattern and content seemed interesting. The communication and sentiment analysis was conducted with Condor, an application developed at the MIT Center for Collective Intelligence (Gloor and Zhao 2004). Condor enabled to parse public discussions in one of the selected Facebook groups, i.e. posts, likes, comments and their respective date and time, covering the timeframe from November 2006 until June 2010. The dataset was generated by requesting publicly available Facebook posts and comments in the selected group through the Facebook Graph API. We collected an aggregated total of 12,000 records of posts and comments across all five groups. All information that Condor accessed was congruent with one-to-many communication that any member of the selected public Facebook groups could potentially browse through.

Table 1 presents a categorization of the most frequent discussion terms in a major public group with more than 24,000 members.
Table 1

Public discussion topics of an IBD-related Facebook group with more than 24,000 members. By category and count of words






Social contacts


2908 crohn

2622 cause

1344 disease

598 ibd

468 colitis

2346 feel

1652 pain

1391 hope

1353 bad

1255 luck

1025 thanks

981 sick

824 hard

761 feeling

609 worse

439 sorry

250 horrible

219 happy

178 worried

1228 flare

804 effects

717 blood

682 colon

638 stomach

524 weight

307 tired

222 inflammation

1112 diagnosed

906 body

788 meds

608 drink

602 remission

513 bathroom

349 food

154 alcohol

1830 doctor

1236 remicade

933 surgery

848 drug

847 prednisone

813 hospital

609 humira

431 imuran

423 treatment

371 infusion

311 colonoscopy

273 asacol

266 pentasa

138 azathioprine

1608 people

515 doctors

454 school

419 mom

380 family

369 friends

345 guys

250 college

235 insurance

140 parents

3137 eat

985 started

982 experience

647 helps

586 tried

448 heard

384 remember

384 change

354 advice

198 inform

As Table 1 suggests, negative communication sentiments and a variety of uncertainties (cf. Emotions) dominated the discussion. In particular, the sentiment analysis revealed that experiences with doctors and nutrition (“eat”) are of high importance.

In a second step of the analysis, the co-appearance of terms by betweenness centrality was visualized (Fig. 2). Betweenness centrality is a social network measure and helps to center those terms in the center of a network visualization that are important connectors to other terms in the group discussion. Such an analysis was helpful in establishing a relationship between the use of certain words and to see how certain terms were connected.
Fig. 2

The top 9 co-appearing words by communication frequency. Facebook group with 24,000 members, 12,007 posts, by betweenness centrality

The term network indicates that “feeling” and “flare” (“flares” or “flare-ups” are one perceived disease activity) co-appeared most often within the post and comments of the analyzed Facebook group. In a broader perspective, the visualization points out that most of the discussions were mostly about the symptoms and feelings that patients experienced.

Taken together, the sentiment analysis summarized in Table 1 and Fig. 2 suggest that members of IBD-related Facebook groups
  1. 1.

    share their emotional state,

  2. 2.

    are in pain,

  3. 3.

    frequently use the bathroom and

  4. 4.

    are very worried about their surroundings, for example in school. Last but not least, they are

  5. 5.

    very open in sharing their individual best practice with food and experimentations.


The semantic analysis provided a very precise impression of an IBD patient’s challenges in daily life. Taken together, the discussion sentiments are rather of a negative nature, which potentially discourages a continuous interaction of patients. Such a communication culture may also entail contagious effects. Investigations of a large real-world social network, collected over a 20-year period, suggest that people's happiness depends on the happiness of others with whom they are connected (Fowler and Christakis 2008). Such insights can be used to precisely address and solve problems with regard to a newly developed application. For example, success stories and more positive sentiments (through a currency of “thankfulness” for sharing information) could be good incentives to visit the online community more frequently.

In a third step, the analysis focused on communication patterns and structures within the Facebook groups. The contribution to a group’s discussion can be measured with the “contribution index”, measuring the ratio of sending and receiving information (Gloor et al. 2003). The contribution index analysis yielded that there was a maximum of six members across the aggregated dataset of all five groups who dominated the content creation and communication. Such dominance can cause problems, because topics covered in discussions may be biased by the interest of a few. Only about 1 % of users significantly contribute to the discussion through new content creation, comments or likes. Roughly only about 40–50 % of these posts would receive a comment or like by other group members. More than 95 % of users in our sample did not make any contribution, as for example by posting new content, liking posts or commenting on them.

Checking for the public statistics on newly joining users among the five selected groups, it could be observed that fewer and fewer members were joining the network—in fact, two out of five groups had a decreasing number of group members toward the end of 2010. One reason for the negative linear trend might be a natural saturation, as other public and closed IBD-related Facebook groups are created. The more speculative assumption is that potential new entrants were demotivated by their first impression of the discussion boards. As negative emotions in the groups appear to take up much space, they do not provide a good incentive for patients to join and engage. Maybe, “bad news” is not the kind of news patients search for, in particular if they hold up on “hope” (see Fig. 2). Taken together, the analysis in this section uncovered a loosely coupled network with following key characteristics: There were
  1. 1.

    very few connected and active patients, there was

  2. 2.

    a rather “problem-” than “solution-”oriented discussion sentiment, and

  3. 3.

    a decreasing number of new group entrants could be observed.


The key question of the analysis was now, which impact such network structures and patterns might have on the diffusion of information, for example of a new therapeutic innovation. In the next section, the diffusion of innovation for the case of whipworm therapy will be analyzed. This therapy emerged around the year of 2005 in science and news as a new promising therapy for IBD patients.

4 Tracing the Whipworm: A Dynamic Communication Analysis

“The etiology and pathogenesis of inflammatory bowel disease has puzzled investigators for decades” and epidemiologic theories of a “hygiene hypothesis” had not been subjected to direct clinical trials with intentional inoculation with infectious agents until now.” (Cohen 2005)

Since about 2005, experimental and clinical studies sustain the idea that helminths might provide protection against IBD (Elliott et al. 2003; Weinstock et al. 2005; Summers et al. 2005; Ruyssers et al. 2008, 2010; Flowers and Hopkins 2013). The clinical studies showed that the treatment of IBD patients with “trichuris suis” (pig whipworm) ova lead to a decrease in the Crohn’s disease and ulcerative colitis disease activity index (Summers et al. 2005).

Cohen (2005) concludes that intentionally infecting patients with (porcine whip-) worms makes headlines. The idea of being infected with living parasites could be psychologically tough to accept for patients (Ruyssers et al. 2008), especially with regard to side-effects.

Unlike the human whipworm (trichuris trichiura), trichuris suis does not spread within the human body as it is not able to deal with the human pathogene structure (Cohen 2005). As such, it does not cause an ingestion of the ovary, noninvasive larvae and a colonization of the colon. Still, therapeutic human helminth colonization needs to be examined for possible adverse side effects (Summers et al. (2003) Reddy and Fried 2009).

According to Ruyssers et al. (2008), the therapy with whipworms might bear a residual risk, which is that helminths might invade other tissues in the human host and cause pathology (Ruyssers et al. 2008). To minimize the risks of a therapy with living parasites, current research focuses on the documentation and classification of immunosuppressive molecules that contribute to the protective effect of helminths (Ruyssers et al. 2008; Heylen et al. 2014). In other words, this yet novel stream of research aims to mimic the whipworm’s effect with the use of convenient and non-stigmatized pills, which were recently studied in multi-center clinical trials (Weinstock and Elliott 2013).

Being informed about the promising research studies on whipworm therapy and their exposure in magazines like The New York Times (Velasquez-Manoff 2008), a further netnographic analysis was conducted, aiming to of track the diffusion of the idea of whipworms as an IBD therapy. In particular, a quantification of scientific publications related to the alternative IBD whipworm therapy as well as a “coolhunting” analysis in the World Wide Web (Gloor and Cooper 2007) was conducted. Such an analysis enabled deeper insights into the external presence of the term “whipworm” that could potentially diffuse into the discussion of the Facebook groups. The diffusion analysis assumes two simplified models of information flow: (1) Information on innovation diffuses from research and news to doctors and from doctors to patients, (2) or directly from research and news to patients, provided information is available online.

Figure 3 indicates the number of general IBD publications and IBD publications with a focus on whipworm/helminthic therapy from 1990 until 2010.
Fig. 3

Number of publications on IBD therapy and whipworm-related IBD therapy

Using the Publish or Perish software and the Google Scholar database in February 2011, scientific papers with the terms “whipworm”, “hookworm”, “trichuris suis” or “helminth” in their title or keywords were queried. While the analysis provides certainly only a rough estimate of the worldwide research volume, the publication data at least resonates with previously mentioned research studies and the first major publication wave on whipworms, reaching its peak in 2005. As it appears, the research on whipworm therapy has been trendy since then and might suggest that such coverage should theoretically induce communications about the innovation within online patient networks.

In order to check the World Wide Web’s knowledge about whipworm, an additional Condor web crawl analysis was conducted. A web crawl analysis fetches thousands of results that a Google search query on “whipworm” returns and structures the available information according to the source, repeating terms and their connectedness.

The graph visualization in Fig. 4 is a structured snapshot of the World Wide Web’s knowledge about whipworms in February 2011. The graphic shows that Google search results on “whipworm” provided information about IBD and whipworm therapy in articles, blogs and other content. The term graph indicates that whipworm theory seemed to be a known “IBD” and “crohn” therapy alternative (“helped”, “using”, “treatment”) in the Internet. However, as the neighborhood to the terms “recently” and “new” accentuates, it was relatively a relatively new therapy at that time.
Fig. 4

Co-appearance of the term “whipworm” in top Google search results (by betweenness centrality, 2 degrees depth, top 50 Google Search results, February 2011)

Given the evident presence of helminthic therapy in research and web news, the diffusion of the topic of whipworm therapy within the selected IBD related Facebook group was now subject to investigation. For instance, following posts or comment excerpts from the selected group reflected the discussion about the whipworm therapy (paraphrased for privacy reasons):

“I found a new study on whipworms, which may help put Crohn’s in remission. It’s really interesting. Although some people are wondering how someone could swallow a worm, I would do anything to get rid of my symptoms. Tell me what are your thoughts?” (2008)

“I’m glad you gave the alternative therapies a try. The trichuris suis whipworm is something that not a lot of people know about. I’m glad that it got you back where you want to be. Rock on!” (2009)

“‘Dr. Elliott said that in Argentina, researchers found that patients with multiple sclerosis who were infected with the human whipworm had milder cases and fewer flare-ups of their disease over a period of four and a half years’ [Quote of the New York Times Article by Velasquez-Manoff (2008)—Any thoughts on that?” (2010)]

These are examples of messages, potentially created by innovator- or early-adopter-type users (Rogers 2003[1962]).

Figure 5 visualizes Facebook group members and their betweenness centrality in group discussions from June 2006 until June 2009. Betweenness centrality is a measure that explains a users’ influence on the transfer of information through the network. According to Kidane and Gloor (2007), betweenness centrality is predictive for creative activity, as the betweenness of a node qualifies to measure the extent, to which a user plays the part of an information gatekeeper, controlling the access and the flow of information (Gloor 2006). Creativity of individuals often appears with the oscillation between different groups and systems. Very often, users with high betweenness centrality are known innovator type users (Ryan and Gross 1943; Rogers 2003[1962]), opinion leaders (Katz 1957; Katz and Lazarsfeld 1957). Betweenness centrality has one important implication for a patient community in practice: By social contagion theory (Aral and Walker 2011), the most between members in the friendship network are in a position to influence their friends (or friends of friends), e.g. with the communication of new innovations such as whipworm therapy. The higher the betweenness centrality of an actor, the more central he appears in the visualization (Fig. 5). If a member posted content relating to whipworm, the respective node is highlighted in black.
Fig. 5

“Whipworm” as a discussion topic in communication of one IBD-related Facebook groups with 24,000 members, by actor betweenness centrality (11/2009)

The analysis suggests that from June 2006 until June 2010, the term “whipworm” (helminths, trichuris or worm) appears only twice (November 2006, November 2009). In both cases, the discussion on the new innovation dissolved very quickly. There was neither a considerable spread of information nor any echo through comments on the topic in the group network. While the posts about whipworms in the observed network were done by rather centrally positioned members, their information did also not diffuse into the group network’s periphery, i.e. through comments by less active or connected members. Assuming that the top influential users by betweenness centrality were identified, one should have observed a more prolonged and broader information spread of the whipworm therapy within the network and course of time.

With a deeper look into the dataset, it seems remarkable that the discussion about whipworm appears together with the terms “heard” and “doctor”. A check of the anonymized posts confirmed that patients “heard” about the “whipworm” therapy from their “doctor”. So far, the research-doctor-patient diffusion model was partially confirmed.

Complementing the previous analysis, Fig. 6 shows the betweenness centrality of different terms, including content in posts and comments in the selected Facebook group with 24,000 members from 2006 to 2010. Data shows that the whipworm therapy does not significantly appear in the discussion sentiment. Rather, classical drugs (“pentasa”, “humira”, “steroids”) take center stage, just as discussions about “pain”, “doctor” and “feelings” (cf. Table 1). The discussion about “whipworms” is located in the far periphery of the overall communication (lower left), so that the average user would very unlikely stumble upon related information in posts or comments when visiting a Facebook group.
Fig. 6

Discussion terms in the group communication of a groups with 24,000 members, by betweenness centrality (06/2006–06/2010)

It is arguable that the social environment of diffusion plays an important role also for online patient networks (Emerson 1962; Lyytinen and Damsgaard 2001). One reason for the low level of communication might be that adopting living helminths as a therapy is not compatible with the social and individual norm. It should further be taken into account that there is a relation between high hygiene standards and the occurrence of IBD. Having that knowledge might indirectly lead to the conclusion that such a patient population has a lower likelihood to cheer for an apparently unappetizing worm. Another reason might be the lack of experience in the own social network and peer structure, and the reliance on others as source of information (Midgley and Dowling 1993). One could argue that trust is particularly low in those cases, were patients are not connected with each other, as shown in Fig. 1. A certain number of trusted members within the personal network must have adopted and posted positive outcomes about helminthic therapy before new patients would jump on the bandwagon. Social proof (Cialdini 2001) is a precondition for cooperation and innovation adoption. All that being said, what seems to be at the core of all observations is that the network’s low density hinders a better diffusion of innovation. A higher density in the friendship network and less centralized discussions might certainly contribute toward a better spread of information within the network and incentivize more feedback loops.

5 Creating a Collaborative Online Patient Network (2012: Ongoing)

On the basis of the results of the patient community anamnesis, a mobile application named “YouMeIBD” is proposed. Alongside the prototype application development, undertaken by an interdisciplinary student class at MIT, the application development concentrates on three questions:
  1. 1.

    Which socio-psychological and socio-technical factors constitute connectedness? (enablers)

  2. 2.

    Which emotional factors drive patients to collaborate and form strong ties? (incentives) and

  3. 3.

    How can mechanisms be created that foster the spread of new innovations? (creativity and diffusion)

It can be concluded that following main network characteristics should be addressed with a solution designed toward helping IBD patients to better connect online:
  1. 1.

    A network’s density must be increased to better connect patients and to support the diffusion of innovations.

  2. 2.

    Information within the network must be better structured as for example in Facebook group discussions. Also, information should be differentiated with regard to their quality, for instance through a collective sorting mechanism. Emotional signals and feelings need a different communication channel than relevant information.

  3. 3.

    A network should harness both the social and the knowledge graph. Patients need to share more relevant information with each other in a trusted environment. While the cause of IBD is still puzzling researchers, more collectively structured information can be contributory to a patient’s individual solution/therapy/everyday life. In terms of co-development, a platform should channel in-formation on alternative or young medical approaches, simultaneously addressing researchers, doctors and patients. In this context, crowdfunding, for instance, has the potential to support innovative processes outside the laboratory (cf.

  4. 4.

    Patients need a place with success stories and positive emotions. IBD patients are likely to conceal their disease toward friends and might be in need of social support. It is important to think about the “currency” users deal with: As patients seek for a more positive and constructive dialog, the values of love, (intrinsic motivation) and glory (peer recognition) should be addressed as important motivators (cf. Malone et al. 2009). A virtual community, enabling to establish connections to distant others with similar interests, can potentially sustain social needs that are not locally met (Wellman 1992; Haythornthwaite 2005).

  5. 5.

    An online community should provide an opportunity to collect—self-reported—patient data on a large scale, permitting researchers to analyze aggregated patient information.

  6. 6.

    Recent studies show that patients with IBD often do not receive an optimal medical therapy (Trivedi and Keefer 2015), which is also reflected in the communication sentiments. In preparation of the platform development, the MIT Center for Collective Intelligence and the Cincinnati Children’s Hospital conducted a quick patient survey in 2011, comprising 57 responses (Baebler et al. 2011). While most patients confirmed they try to use Facebook to connect online, they expressed their need for a more efficient and trustful online platform. Symptoms, treatment options and their social situation are the key contents that patients want to discuss about in a trustworthy environment. Thus, a network must address an individual user’s need with regard to his interests, symptoms and medications.


All these points together, resulting from an extensive community analysis, are factors to be considered in the creation of an online social network for IBD patients. The following figures demonstrate the main functions of the mobile application “YouMeIBD”.

The starting page of “YouMeIBD” (Fig. 7) invites users to login with their current Facebook account (Login). Using a user’s Facebook credentials comes with the advantage that it is an easy way to catch information a user was already willing to expose. For instance, his profile picture, age, gender, location information, friends and likes. The latter are especially important as they can provide the matching algorithm with relevant data (Fig. 8). Like on online dating websites, common music, movie or website interests, group belongings and mutual friends are good factors to facilitate connections, even though they rather seem to be “soft factors” in a patient community. Since decades, research claims that common interests are a good predictor of homophily (Lazarsfeld and Merton 1954; Byrne 1971; McPherson et al. 2001) and thus promising in creating a denser network with many strong connections among patients. Probably, there is a strong correlation between the willingness to share health data and the level of trust and empathy for peers.
Fig. 7

YouMeIBD functions: Login, Me and My Friends, My Profile

Fig. 8

YouMeIBD functions: Create Question, Answer Question, Matches

Me and My Friends serves to update a user’s profile, emotional mood and further allows sending messages to friends in their network. Such friendships are named “Soulmates”. Users can see the most recent status updates by their peers and get an impression on their thoughts and emotions. Another sub-element is a profile page My Profile where users can add tags that matter to them, as e.g. their symptoms or current medication. In addition to profile and question data, these self-tags can be used to improve the matching of patients, i.e. matching a beginner in a certain therapy method with an experienced expert.

The Questions function as shown in Fig. 8 provides users with the opportunity to create questions. Correspondingly, the Answers function asks a user to rate and answer questions by other users.

Based on algorithmic matching of Facebook data, common tags and answered questions, users will finally be able to Find Soulmates who share mutual attributes and interests. For each user, a score of matching is calculated and displayed with supplementary information on the match. As electronic relationships deepen (Rheingold 1993; Kendall 2002), online connectivity often induces face-to-face contacts in the real world. The Meetings function (see Fig. 9) allows users to create events, supporting real-world connectedness.
Fig. 9

YouMeIBD functions: Meetings, Stream—Submit Link, Stream—Index

Another future functionality could be the opportunity to create virtual online “flash teams” of expert patients, doctors and researchers to collaboratively investigate certain information that might be interesting to specific patient groups (Retelny et al. 2014). Within a set timeframe, patients could start a flash team of crowdworking patients that aims to summarize scientific studies on whipworm therapy or the status of current clinical trials. Current online collaboration software such as Slack, Google Docs or Dropbox could support the coordination and collective information gathering. Intrinsically motivated and equipped with expert knowledge, patient flash teams could potentially perform better than teams in laboratory settings.

The intention of the Stream function is to channel relevant news (therapies, research, stories, blog entries). The stream serves as an information system, relaying, amplifying, and structuring information. Functionally, users can submit and upvote links that seem helpful or interesting for other patients. Contents are then ordered by their overall (upvote) score. The more relevant an information seems to other users, the more points the author of the content will receive. By doing so, YouMeIBD aims at implementing a “Hacker News for Patients”. Hacker News ( is a programming, technology and hacks focused social news feed, that applies a time- and point-oriented algorithm. The feed, that was started by the startup accelerator Y Combinator in 2007, enables expert users to submit links to the website, that can subsequently be up-voted by users of the platform. The more up-votes a website link receives within a certain amount of time, the higher the link will be ranked on the website. Each link can be clicked and commented by users of the system. Hacker News or a more product-related system called “Product Hunt” ( became popular as efficiently working hubs for information on new technologies and software. Potentially, a patient community focused on a specific disease could harness such an infrastructure to collaboratively create and filter relevant links. Such a mechanism comes with two benefits:
  1. 1.

    In case a link (information) is ranked high, it could get faster response from the community. As Malone et al. (2009) state, Create and Decide are two of the four main underlying processes in a collective intelligent system. While Create describes the collective innovation and co-creation process, Decide represents the process of selection and evaluation of given options, e.g. through collective voting mechanisms. As such, the patient crowd would be able to prioritize relevant information.

  2. 2.

    Information that otherwise would not have found its way into a patient community, could now be elevated. For instance, there could be a link to an IBD patient's findings on self-tracking content that may never have been discovered by a doctor or scientific journal. As the nature of the medical innovation system leads to years of development time, especially young patients run out of patience. Medicine is a highly regulated arena in which new drugs and therapies must undergo extensive testing and approval processes before they can be offered to the market. A recent estimate suggests that it takes 8–12 years and costs ranging from USD 0.5 billion to USD 2 billion in order to bring a new drug to the market (DiMasi et al. 2003; Adams and Van Brantner 2006; Flowers 2014; Flowers and Hopkins 2013). On the other hand, studies by Von Hippel (1986) have shown that ‘lead users’ often experience emerging needs and may prototype products and services that can satisfy these needs that are not met by the market. User innovators also draw their innovations from ‘local knowledge’; knowledge that is distinct from the types of knowledge that is generated within firms. This ‘sticky local information’ (Von Hippel 1994) that users acquire is often costly to transfer to producers or hard to find within a clinical setting.

    In the context of IBD, where therapies may require a decade or more of research and clinical trials before they overcome regulatory approval, some users may ‘side-step’ the innovation process (Flowers and Hopkins 2013, Flowers 2014) and actively diffuse their own innovation (Harhoff et al. 2003). Here, a collaborative sorting mechanism can contribute to rank information on (user) innovation from network peripheries that would otherwise not have been visible.


6 Summary and Outlook

This contribution started with an exemplary analysis on the pattern and profile of an online IBD patient community. It could be learned that coolhunting and “community snapshots” can provide very valuable information on a community’s characteristics, its problems and needs.

Besides analyzing the low-density structure of the network, a sentiment analysis provided insights into the emotionality and content of communication. This article further explained how a high network centrality and low network density hindered the diffusion of an innovation, that counts at least for the case of whipworms.

Insights from computational social science helped to design a prototype concept, addressing the identified problems. The efficiency of the proposed application prototype will be subject to future empirical investigation.

Initially targeted for the IBD community, YouMeIBD is a non-profit prototype and to be extended to other chronic disease groups. The model could equally be applied in other patient communities of chronic diseases or any disease where belonging and cohesion do not only support individual well-being but also the diffusion of information and innovation. The access to such a mobile web platform can be limited by the ability and practice to communicate online, particularly for elderly or handicapped patients. However, as Internet affinity and technological aids will accompany the next generation of patients, one can be optimistic that this group of patients will be able to access online patient platforms.

Started in 2010, five student classes contributed to the YouMeIBD project. Class by class, the project went one step further toward a real world application for IBD patients. On the open source project websites of COINs classes (cf. COINs Class 2010, 2011, 2012, 2013, 2014, 2015) and at you may follow the recent development of the application until July 2015. Everyone who is interested in continuing on our model the produced mockups and functions is invited to get in touch.


  1. 1.

    Information on an online patient community is worthy of protection, even if—like in this case—the available data is publicly exposed. With respect to privacy and the ongoing ethics discussion in the computational social sciences, the analysis of interconnectedness is limited to randomized subsamples of information. It should be underlined, that our returning information certainly has a model character, but still seems to confirm what could be learned through close collaboration with patient councils.



This contribution represents the work of a team consisting of more than 15 students, patients, clinicians, designers, researchers and developers. Prof. Dr. Peter A. Gloor at the MIT Center for Collective Intelligence and the Collaborative Chronic Care Network (C3N) team at Cincinnati Children’s Hospital initiated this research project. Hauke Führes, Linda Bäbler, Jonas Lauener and Leslie Marticke contributed to the first YouMeIBD prototype application described in this article. Since then, five interdisciplinary and international student classes contributed to the project (cf. COINs Class 2010, 2011, 2012, 2013, 2014, 2015). I like to thank the Dr. Werner Jackstädt Foundation for their generous support.


  1. Adams CP, Brantner VV (2006) Estimating the cost of new drug development: is it really $802 million? Health Aff 25(2):420–428CrossRefGoogle Scholar
  2. Alic MSM (2000) Inflammatory bowel diseases are diseases of higher socioeconomic status: dogma or reality? Am J Gastroenterol 95:3332–3333. doi: 10.1016/S0002-9270(00)02123-7 CrossRefGoogle Scholar
  3. Aral S, Walker D (2011) Creating social contagion through viral product design: a randomized trial of peer influence in networks. Manag Sci 57:1623–1639. doi: 10.1287/mnsc.1110.1421 CrossRefGoogle Scholar
  4. Baebler L, Gloor P, Kaminski J, Marticke L, Dellal G (2011) YouApp—it’s all about YOU: finding patients like myself on Facebook. In: Proceedings of the 3rd international conference on Collaborative Innovation Networks COINs 2011, Basel, Switzerland, 8–10 September 2011. Accessed 1 Apr 2015
  5. Bastian M, Heymann S, Jacomy M (2009) Gephi: an open source software for exploring and manipulating networks. Third international AAAI conference on Weblogs and Social Media. Accessed 1 May 2015
  6. Bernstein CN, Kraut A, Blanchard JF, Rawsthorne P, Yu N, Walld R (2001) The relationship between inflammatory bowel disease and socioeconomic variables. Am J Gastroenterol 96:2117–2125. doi: 10.1016/S0002-9270(01)02509-6 CrossRefGoogle Scholar
  7. Burisch J, Munkholm P (2013) Inflammatory bowel disease epidemiology. Curr Opin Gastroenterol 29:357–362. doi: 10.1097/MOG.0b013e32836229fb CrossRefGoogle Scholar
  8. Burisch J, Jess T, Martinato M, Lakatos PL (2013) The burden of inflammatory bowel disease in Europe. J Crohns Colitis 7:322–337. doi: 10.1016/j.crohns.2013.01.010 CrossRefGoogle Scholar
  9. Burt RS (2001) Structural holes versus network closure as social capital. In: Lin N, Cook KS, Burt RS (eds) Social capital theory and research. Aldine de Gruyter, New York, pp 31–56Google Scholar
  10. Byrne D (1971) The attraction paradigm. Academic, New YorkGoogle Scholar
  11. CCFA (2015) Crohn’s and Colitis Foundation of America. Fact book. Accessed 1 May 2015
  12. CDC (2014) Center for Disease Control and Prevention. Accessed 20 Nov 2014
  13. Christakis NA, Fowler JH (2011) Connected: the surprising power of our social networks and how they shape our lives. Little, Brown, Boston, p 256Google Scholar
  14. Cialdini RB (2001) The science of persuasion. Sci Am 284:76–81. doi: 10.1038/scientificamerican0201-76 CrossRefGoogle Scholar
  15. Cohen RD (2002) The quality of life in patients with Crohn’s disease. Aliment Pharmacol Ther 16:1603–1609. doi: 10.1046/j.1365-2036.2002.01323.x CrossRefGoogle Scholar
  16. Cohen RD (2005) Worms, germs and IBD: is it your mother’s fault? Gastroenterology 129:1127–1128. doi: 10.1053/j.gastro.2005.06.076, discussion 1128–9CrossRefGoogle Scholar
  17. COINs Class (2010). Accessed 1 Apr 2015
  18. COINs Class (2011). Accessed 1 Apr 2015
  19. COINs Class (2012). Accessed 1 Apr 2015
  20. COINs Class (2013). Accessed 1 Apr 2015
  21. COINs Class (2014). Accessed 1 Apr 2015
  22. COINs Class (2015). Accessed 1 Apr 2015
  23. Cooke WT (1981) Factors in the management of Crohn’s disease: a discussion paper. J R Soc Med 74:757Google Scholar
  24. Crohn BB, Ginzburg L, Oppenheimer GD (1984) Regional ileitis. A pathological and clinical entity. JAMA 251:73–79. doi: 10.1001/jama.251.1.73 CrossRefGoogle Scholar
  25. Datamonitor (2010) Commercial insight: inflammatory Bowel disease—future market prospects rely on addressing cost and safety. Research and Markets. Abstract: Accessed 20 Nov 2014
  26. DiMasi JA, Hansen RW, Grabowski HG (2003) The price of innovation: new estimates of drug development costs. J Health Econ 22:151–185. doi: 10.1016/S0167-6296(02)00126-1 CrossRefGoogle Scholar
  27. Duchmann R, Zeitz M (2005) Crohn’s disease: current pathogenetic paradigms. Mucosal Immunol 3:1265–1286CrossRefGoogle Scholar
  28. Dunbar RIM (1992) Neocortex size as a constraint on group size in primates. J Hum Evol 22:469–493. doi: 10.1016/0047-2484(92)90081-J CrossRefGoogle Scholar
  29. Dutton G (2013) Big need in IBD: drugs to slow progression. Genetic Eng Biotechnol News 33(14):14–14. doi: 10.1089/gen.33.14.05 CrossRefGoogle Scholar
  30. Economou M, Pappas G (2008) New global map of Crohn’s disease: genetic, environmental, and socioeconomic correlations. Inflamm Bowel Dis 14:709–720. doi: 10.1002/ibd.20352 CrossRefGoogle Scholar
  31. Elliott DE, Weinstock JV (2012) Helminth-host immunological interactions: prevention and control of immune-mediated diseases. Ann N Y Acad Sci 1247:83–96. doi: 10.1111/j.1749-6632.2011.06292.x CrossRefGoogle Scholar
  32. Elliott DE, Li J, Blum A, Metwali A, Qadir K, Urban JF, Weinstock JV (2003) Exposure to schistosome eggs protects mice from TNBS-induced colitis. Am J Physiol Gastrointest Liver Physiol 284:G385–G391. doi: 10.1152/ajpgi.00049.2002 CrossRefGoogle Scholar
  33. Elliott DE, Summers RW, Weinstock JV (2007) Helminths as governors of immune-mediated inflammation. Int J Parasitol 37:457–464. doi: 10.1016/j.ijpara.2006.12.009 CrossRefGoogle Scholar
  34. Emerson RM (1962) Power-dependence relations. Am Sociol Rev 27:31. doi: 10.2307/2089716 CrossRefGoogle Scholar
  35. Eysenbach G, Powell J, Englesakis M, Rizo C, Stern A (2004) Health related virtual communities and electronic support groups: systematic review of the effects of online peer to peer interactions. BMJ 328:1166. doi: 10.1136/bmj.328.7449.1166 CrossRefGoogle Scholar
  36. Fiocchi C (1998) Inflammatory cowel disease: etiology and pathogenesis. Gastroenterology 115:182–205. doi: 10.1016/S0016-5085(98)70381-6 CrossRefGoogle Scholar
  37. Fowler JH, Christakis NA (2008) Dynamic spread of happiness in a large social network: longitudinal analysis over 20 years in the Framingham Heart Study. BMJ 337:a2338CrossRefGoogle Scholar
  38. Flowers S (2014) Medical user innovation in the context of chronic disease: the creation of Helminthic therapy. Workshop on medical user innovation and medical commons, NYU School of Law, 15–17 May 2014. Working paper. Accessed 20 Nov 2014
  39. Flowers S, Hopkins M (2013) Autoimmune disease: patients self-treat with parasitic worms. Nature 493:163. doi: 10.1038/493163c CrossRefGoogle Scholar
  40. Fuehres H, Kaminski J, Gloor P, Zhang X (2011) Finding a cure for Crohn’s disease—analyzing medical conditions in online communities. In: Proceedings of the Sunbelt 2011, St. Pete’s Beach, Florida, 10–13 February 2011Google Scholar
  41. Gallant MP (2003) The influence of social support on chronic illness self-management: a review and directions for research. Health Educ Behav 30:170–195. doi: 10.1177/1090198102251030 CrossRefGoogle Scholar
  42. Garn H, Renz H (2007) Epidemiological and immunological evidence for the hygiene hypothesis. Immunobiology 212:441–452. doi: 10.1016/j.imbio.2007.03.006 CrossRefGoogle Scholar
  43. Gazzard BG, Price HL, Libby GW, Dawson AM (1978) The social toll of Crohn’s disease. Br Med J 2:1117–1119. doi: 10.1136/bmj.2.6145.1117 CrossRefGoogle Scholar
  44. Geboes K, Collins S (1998) Structural abnormalities of the nervous system in Crohn’s disease and ulcerative colitis. Neurogastroenterol Motil 10:189–202. doi: 10.1046/j.1365-2982.1998.00102.x CrossRefGoogle Scholar
  45. Gibson TB, Ng E, Ozminkowski RJ, Wang S, Burton WN, Goetzel RZ, Maclean R (2008) The direct and indirect cost burden of Crohn’s disease and ulcerative colitis. J Occup Environ Med 50:1261–1272. doi: 10.1097/JOM.0b013e318181b8ca CrossRefGoogle Scholar
  46. Gloor PA (2006) Swarm creativity: competitive advantage through collaborative innovation networks. Oxford University Press, New YorkCrossRefGoogle Scholar
  47. Gloor PA, Cooper S (2007) Coolhunting: chasing down the next big thing. AMACOM, New YorkGoogle Scholar
  48. Gloor PA, Zhao Y (2004) TeCFlow—a temporal communication flow visualizer for social network analysis. ACM CSCW Workshop on Social Networks, pp 1–4Google Scholar
  49. Gloor PA, Laubacher R, Dynes SBC, Zhao Y (2003) Visualization of communication patterns in collaborative innovation networks—analysis of some W3C working groups. In: Proceedings of the twelfth international Conference on Information and Knowledge Management—CIKM’03 56. doi: 10.1145/956873.956875
  50. Gloor PA, Grippa F, Borgert A, Colletti R, Dellal G, Margolis P, Seid M (2011) Towards growing a COIN in a medical research community. Procedia Soc Behav Sci 26. In: Proceedings COINs 2010, collaborative innovations networks conference, Savannah, GA, 7–9 October 2010. doi: 10.1016/j.sbspro.2011.10.557
  51. Harhoff D, Henkel J, Von Hippel E (2003) Profiting from voluntary information spillovers: how users benefit by freely revealing their innovations. Res Policy 32(10):1753–1769CrossRefGoogle Scholar
  52. Haythornthwaite C (2005) Social networks and internet connectivity effects. Inf Commun Soc 8:125–147. doi: 10.1080/13691180500146185 CrossRefGoogle Scholar
  53. Heylen M, Ruyssers NE, Gielis EM, Vanhomwegen E, Pelckmans PA, Moreels TG, De Man JG, De Winter BY (2014) Of worms, mice and man: an overview of experimental and clinical helminth-based therapy for inflammatory bowel disease. Pharmacol Ther 143:153–167. doi: 10.1016/j.pharmthera.2014.02.011 CrossRefGoogle Scholar
  54. Isaacs KL, Lewis JD, Sandborn WJ, Sands BE, Targan SR (2005) State of the art: IBD therapy and clinical trials in IBD. Inflamm Bowel Dis 11(Suppl 1):S3–S12. doi: 10.1097/01.MIB.0000184852.84558.b2 CrossRefGoogle Scholar
  55. Kaminski J, Gloor P, Marticke L (2012) YouMeIBD—an interactive online matchmaker for patients of IBD/Crohn’s disease, Stanford Medicine X, Stanford University, Palo Alto, 28–30 September 2012. Accessed 20 Nov 2014
  56. Kappelman MD, Rifas-Shiman SL, Kleinman K, Ollendorf D, Bousvaros A, Grand RJ, Finkelstein JA (2007) The prevalence and geographic distribution of Crohn’s disease and ulcerative colitis in the United States. Clin Gastroenterol Hepatol 5:1424–1429. doi: 10.1016/j.cgh.2007.07.012 CrossRefGoogle Scholar
  57. Katz E (1957) The two-step flow of communication: an up-to-date report on an hypothesis. Public Opin Q 21:61. doi: 10.1086/266687 CrossRefGoogle Scholar
  58. Katz E, Lazarsfeld PF (1957) Personal influence. Free Press, New YorkGoogle Scholar
  59. Kendall L (2002) Hanging out in the virtual pub: masculinities and relationships online. University of California Press, BerkeleyCrossRefGoogle Scholar
  60. Kidane YH, Gloor PA (2007) Correlating temporal communication patterns of the Eclipse open source community with performance and creativity. Comput Math Organ Theory 13:17–27. doi: 10.1007/s10588-006-9006-3 CrossRefGoogle Scholar
  61. Lakatos PL (2006) Recent trends in the epidemiology of inflammatory bowel diseases: up or down? World J Gastroenterol 12:6102–6108CrossRefGoogle Scholar
  62. Lashner BA, Loftus EV (2006) True or false? The hygiene hypothesis for Crohn’s disease. Am J Gastroenterol 101:1003–1004. doi: 10.1111/j.1572-0241.2006.00563.x CrossRefGoogle Scholar
  63. Lazarsfeld PF, Merton RK (1954) Friendship as a social process: a substantive and methodological analysis. Free Control Mod Soc 18:18–66Google Scholar
  64. Lazer D, Pentland A, Adamic L, Aral S, Barabási A-L, Brewer D, Christakis N, Contractor N, Fowler J, Gutmann M, Jebara T, King G, Macy M, Roy D, van Alstyne M (2009) Computational social science. Science 323:721–723. doi: 10.1126/science.1167742 CrossRefGoogle Scholar
  65. Loftus EV (2004) Clinical epidemiology of inflammatory bowel disease: incidence, prevalence, and environmental influences. Gastroenterology 126:1504–1517. doi: 10.1053/j.gastro.2004.01.063 CrossRefGoogle Scholar
  66. Loftus EV (2007) The burden of inflammatory bowel disease in the United States: a moving target? Clin Gastroenterol Hepatol 5:1383–1384. doi: 10.1016/j.cgh.2007.10.016 CrossRefGoogle Scholar
  67. Longobardi T, Jacobs P, Bernstein CN (2003) Work losses related to inflammatory bowel disease in the United States: results from the National Health Interview Survey. Am J Gastroenterol 98:1064–1072. doi: 10.1111/j.1572-0241.2003.07285.x Google Scholar
  68. Lyytinen K, Damsgaard J (2001) What’s wrong with the diffusion of innovation theory? The case of a complex and networked technology. Reports Aalborg University Department of Computer Science R985010 187:1–20Google Scholar
  69. Malone TW, Laubacher R, Dellarocas C (2009) Harnessing crowds : mapping the genome of collective intelligence. Elements 1:1–20. doi: 10.2139/ssrn.1381502 Google Scholar
  70. McLeod RS, Churchill DN, Lock AM, Vanderburgh S, Cohen Z (1991) Quality of life of patients with ulcerative colitis preoperatively and postoperatively. Gastroenterology 101:1307–1313CrossRefGoogle Scholar
  71. McPherson M, Smith-Lovin L, Cook JM (2001) Birds of a feather: homophily in social networks. Annu Rev Sociol 27:415–444. doi: 10.1146/annurev.soc.27.1.415 CrossRefGoogle Scholar
  72. Metcalfe B (1995) Metcalfe’s law: a network becomes more valuable as it reaches more users. InfoWorld 17:53–54Google Scholar
  73. Midgley DF, Dowling GR (1993) A longitudinal study of product form innovation: the interaction between predispositions and social messages. J Consum Res 19:611. doi: 10.1086/209326 CrossRefGoogle Scholar
  74. Park KT, Bass D (2011) Inflammatory bowel disease-attributable costs and cost-effective strategies in the United States: a review. Inflamm Bowel Dis 17:1603–1609. doi: 10.1002/ibd.21488 CrossRefGoogle Scholar
  75. Pentland A (2014) Social physics: how good ideas spread. Penguin Press, New YorkGoogle Scholar
  76. Portier K, Greer GE, Rokach L, Ofek N, Wang Y, Biyani P, Yu M, Banerjee S, Zhao K, Mitra P, Yen J (2013) Understanding topics and sentiment in an online cancer survivor community. J Natl Cancer Inst Monogr 2013:195–198. doi: 10.1093/jncimonographs/lgt025 CrossRefGoogle Scholar
  77. Probert CS, Brown M (2008) Are there any ethnic groups that are more likely to develop IBD? Inflamm Bowel Dis 14(Suppl 2):S24–S25. doi: 10.1002/ibd.20601 CrossRefGoogle Scholar
  78. Reddy A, Fried B (2009) An update on the use of helminths to treat Crohn’s and other autoimmunune diseases. Parasitol Res 104:217–221. doi: 10.1007/s00436-008-1297-5 CrossRefGoogle Scholar
  79. Retelny D, Robaszkiewicz S, To A, Lasecki W, Patel J, Rahmati N, Doshi T, Valentine M, Bernstein MS (2014) Expert crowdsourcing with flash teams. In: Proceedings of the 27th annual ACM symposium on User interface software and technology, ACM, 2014. doi: 10.1145/2642918.2647409
  80. Rheingold H (1993) The virtual community: homesteading on the electronic frontier. Addison-Wesley, ReadingGoogle Scholar
  81. Rogers EM (2003[1962]) Diffusion of innovations, 5th edn. Free Press, New YorkGoogle Scholar
  82. Ruyssers NE, De Winter BY, De Man JG, Loukas A, Herman AG, Pelckmans PA, Moreels TG (2008) Worms and the treatment of inflammatory bowel disease: are molecules the answer? Clin Dev Immunol 2008:1–7. doi: 10.1155/2008/567314 CrossRefGoogle Scholar
  83. Ruyssers NE, De Winter BY, De Man JG, Ruyssers ND, Van Gils AJ, Loukas A, Pearson MS, Weinstock JV, Pelckmans PA, Moreels TG (2010) Schistosoma mansoni proteins attenuate gastrointestinal motility disturbances during experimental colitis in mice. World J Gastroenterol 16:703–712. doi: 10.3748/wjg.v16.i6.703 CrossRefGoogle Scholar
  84. Ryan B, Gross NC (1943) The diffusion of hybrid seed corn in two Iowa communities. Rural Sociol 8:15–24Google Scholar
  85. Sajadinejad MS, Asgari K, Molavi H, Kalantari M, Adibi P (2012) Psychological issues in inflammatory bowel disease: an overview. Gastroenterol Res Pract 2012. doi: 10.1155/2012/106502
  86. Sonnenberg A (2009) Demographic characteristics of hospitalized IBD patients. Dig Dis Sci 54:2449–2455. doi: 10.1007/s10620-009-0973-3 CrossRefGoogle Scholar
  87. Strachan DP (2000) Family size, infection and atopy: the first decade of the “hygiene hypothesis”. Thorax 55(Suppl 1):S2–S10. doi: 10.1136/thorax.55.suppl_1.S2 CrossRefGoogle Scholar
  88. Summers RW, Elliott DE, Qadir K, Urban JF, Thompson R, Weinstock JV (2003) Trichuris suis seems to be safe and possibly effective in the treatment of inflammatory bowel disease. Am J Gastroenterol 98:2034–2041. doi: 10.1111/j.1572-0241.2003.07660.x CrossRefGoogle Scholar
  89. Summers RW, Elliott DE, Urban JF, Thompson R, Weinstock JV (2005) Trichuris suis therapy in Crohn’s disease. Gut 54:87–90. doi: 10.1136/gut.2004.041749 CrossRefGoogle Scholar
  90. Thukkani N, Williams JL, Sonnenberg A (2011) Epidemiologic characteristics of patients with inflammatory bowel disease undergoing colonoscopy. Inflamm Bowel Dis 17:1333–1337. doi: 10.1002/ibd.21513 CrossRefGoogle Scholar
  91. Trivedi I, Keefer L (2015) The emerging adult with inflammatory bowel disease: challenges and recommendations for the adult gastroenterologist. Gastroenterol Res Pract. doi: 10.1155/2015/260807
  92. Valente T (2010) Social networks and health: models, methods, and applications. Oxford University Press, New YorkCrossRefGoogle Scholar
  93. Van Assche G, Dignass A, Panes J, Beaugerie L, Karagiannis J, Allez M, Ochsenkühn T, Orchard T, Rogler G, Louis E, Kupcinskas L, Mantzaris G, Travis S, Stange E (2010) The second European evidence-based consensus on the diagnosis and management of Crohn’s disease: definitions and diagnosis. J Crohns Colitis 4:7–27. doi: 10.1016/j.crohns.2009.12.003 CrossRefGoogle Scholar
  94. Velasquez-Manoff M (2008) The worm turns. New York Times, June 29 2008. Accessed 20 Nov 2014
  95. Venu N, Cohen RD (2011) The new economic reality in the world of IBD. In: Cohen RD (ed) Inflammatory bowel disease: diagnosis and therapeutics. Springer, New York. doi: 10.1007/978-1-60327-433-3_19 Google Scholar
  96. Von Hippel E (1986) Lead users: a source of novel product concepts. Manag Sci 32(7):791–805CrossRefGoogle Scholar
  97. Von Hippel E (1994) “Sticky information” and the locus of problem solving: implications for innovation. Manag Sci 40:429–439. doi: 10.1287/mnsc.40.4.429 CrossRefGoogle Scholar
  98. Wagner EH (2001) Meeting the needs of chronically ill people. Br Med J 323:945–946. doi: 10.1136/bmj.323.7319.945 CrossRefGoogle Scholar
  99. Wasserman S, Faust K (1994) Social network analysis: methods and applications. Cambridge University Press, CambridgeCrossRefGoogle Scholar
  100. Weinstock JV, Elliott DE (2013) Translatability of helminth therapy in inflammatory bowel diseases. Int J Parasitol 43:245–251. doi: 10.1016/j.ijpara.2012.10.016 CrossRefGoogle Scholar
  101. Weinstock JV, Summers RW, Elliott DE, Qadir K, Urban JF, Thompson R (2002) The possible link between de-worming and the emergence of immunological disease. J Lab Clin Med 139:334–338. doi: 10.1067/mlc.2002.124343 CrossRefGoogle Scholar
  102. Weinstock JV, Summers RW, Elliott DE (2005) Role of helminths in regulating mucosal inflammation. Springer Semin Immunopathol 27:249–271. doi: 10.1007/s00281-005-0209-3 CrossRefGoogle Scholar
  103. Wellman B (1988) Structural analysis: from method and metaphor to theory and substance. In: Wellman B, Berkowitz SD (eds) Social structures: a network approach. Cambridge University Press, CambridgeGoogle Scholar
  104. Wellman B (1992) Which ties provide what kinds of support? Adv Group Process 9:207–235Google Scholar
  105. Wellman B (2012) Is Dunbar’s number up? Br J Psychol 103:174–176. doi: 10.1111/j.2044-8295.2011.02075.x CrossRefGoogle Scholar
  106. Wicks P, Massagli M, Frost J, Brownstein C, Okun S, Vaughan T, Bradley R, Heywood J (2010) Sharing health data for better outcomes on PatientsLikeMe. J Med Internet Res. doi: 10.2196/jmir.1549
  107. Wicks P, Vaughan TE, Massagli MP, Heywood J (2011) Accelerated clinical discovery using self-reported patient data collected online and a patient-matching algorithm. Nat Biotechnol 29:411–414. doi: 10.1038/nbt.1837 CrossRefGoogle Scholar
  108. Xavier RJ, Podolsky DK (2007) Unravelling the pathogenesis of inflammatory bowel disease. Nature 448:427–434. doi: 10.1038/nature06005 CrossRefGoogle Scholar
  109. Yu AP, Cabanilla LA, Wu EQ, Mulani PM, Chao J (2008) The costs of Crohn’s disease in the United States and other Western countries: a systematic review. Curr Med Res Opin 24:319–328. doi: 10.1185/030079908X260790 CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2016

Authors and Affiliations

  1. 1.Witten/Herdecke UniversityWittenGermany
  2. 2.MIT Center for Collective IntelligenceMassachusetts Institute of TechnologyCambridgeUSA

Personalised recommendations