1
Abstract
Social media is increasingly used by healthcare users and providers to connect and communicate with each other. This may be beneficial to both parties, but it could also bring threats for healthcare providers and disrupt the healthcare system. Despite the fact that this is an important area requiring a deeper understanding, currently there is no clear overview of social media enabled interactions in healthcare. In order to attain a better grasp of the potential benefits and threats social media could have for healthcare providers and healthcare in general, a first step is to categorize the types of online interactions. In this context, we study six contrasting categories of social media and identify six archetypical interactions in the healthcare domain, thus providing a foundation for future research on the disruptiveness of social media in healthcare.
Key words: healthcare users, healthcare providers, blogs, social networking sites, collaborative projects, content communities, purposive sampling, content analysis
Research Master Thesis Edin Smailhodzic (S2072335)
Faculty of Economics and Business Administration Supervisors:
Prof dr. Albert Boonstra Dr. David Langley
2
Introduction
Traditionally, communication between patients and physicians entailed face to face interaction as the primary form of information exchange. However, the emergence of the Internet transformed the way in which patients look for health related information and share their experiences. Easier access to medical knowledge helps the patients to take responsibility and get more involved in management of their own health (Eysenbach, 2008). According to a recent Pew Research study, in 2012, 72% of the Internet users in the USA searched for health information online (Fox & Duggan, 2013). Users resorted to online sources for a variety of reasons either for themselves or their family members and friends.
The rise of social media enabled healthcare users to move from one-to-one to one-to many and many-to-many communication (Hawn, 2009). As such, the healthcare users’ utilization of social media gained in importance and prevalence (Ziebland & Wyke, 2012; Zhao, Ha & Widdows, 2013) as it enables them to access health related information, interpret the information and contribute their own experiences, bringing benefit both to them and to others (Adams, 2010). Resultantly, healthcare users are able to find the information they require and feel supported (Ziebland & Wyke, 2012), they increase their knowledge and exchange advice with others (Antheunis, Tates & Nieboer, 2013), feel empowered in the sense that their meaning, competence, and self-determination to manage their own health increases (Johnston, Worrell, Di Gangi & Wasko, 2013) and make better informed decisions (Wicks et al., 2010). Limited research done thus far provided merely an overview of social media in the healthcare sector (Ziebland & Wyke, 2012; Anthenuis et al., 2013; Chou, Prestin, Lyons, & Wen, 2013; Hamm et al., 2013). However, we still know little about the social media focusing specifically on health and further research is encouraged (Faraj, von Krogh, Lakhani, & Monteiro, 2014). Furthermore, to the best of our knowledge, there is no typology of the social media enabled interactions in healthcare, i.e. health related communication or exchange via social media with potential to affect many-to-many interlocutors.
3 Virtual communication reduces time or space limitations and facilitates the establishment of communities and communication such never existed before. Healthcare users can share information and feelings as part of communities targeting specific medical conditions. An example of such a case is dLife which focuses on diabetes. In this new way, social media enable healthcare users to receive information from other users (Eysenbach, 2008) but, at the same, also make it possible for healthcare providers to reach more patients and easily and quickly provide information on various diseases (Hawn, 2009). This is quite a change from traditional face to face communication in the healthcare, wherein healthcare users would interact with healthcare providers in real time and at their offices. The increasing use of social media may impact also on the offline interactions between healthcare users and providers with the former now being able to make better informed choices and easily tap into the information and experiences of others with healthcare providers (KPMG, 2011). Another potential benefit could be one’s own health management, especially in the case of chronic diseases. On the other hand, by replacing offline with online interactions, healthcare providers could employ this way of reaching healthcare users irrespective of their location to offer new solutions and thus increase their market share. However, the benefits and opportunities aside, social media may also pose a threat and create problems as the change in interaction with healthcare providers could potentially result in altering the way the providers fulfil their role in the healthcare system (Andersen, Medaglia & Henriksen 2012). Therefore should healthcare providers acknowledge the power the social media may have in this respect and respond proactively to such changes, they may reap the benefits from the shift. Therefore, both a better understanding and the way to categorize social media enabled interactions in the healthcare warrant academic attention.
4 Scientifically, this research contributes to a deeper insight into social media in the healthcare domain. As such, it adds to the literature on information systems with a particular focus on healthcare by addressing how social media features are utilized by users, which has been identified as an important topic for future research (Aral, Dellarocas & Godes, 2013). Furthermore, the research also contributes to a better understanding of how knowledge is shared within online communities, especially those focusing on health, as a topic on which further research is encouraged (Faraj, von Krogh, Lakhani, & Monteiro, 2014). In particular, it illustrates the different types of interactions between healthcare users and healthcare providers that are enabled by social media and what they bring about. The proposed typology may reduce the complexity of our understanding of social media in the healthcare sector. It provides structured overview of the social media enabled interactions in the healthcare domain that could facilitate any future research aiming to explore the interactions further. It complements a general typology of social media (Kaplan & Haenlein, 2010) by proposing archetypes of interactions specific to healthcare. It links the archetypes to the different categories of social media from this general typology. It can serve as a first step to understanding the effects of social media on healthcare providers and healthcare system.
5 This introduction is followed by a review of the literature on social media and social media in the healthcare domain. After that we describe our methods, discuss our results, and provide a conclusion.
Theoretical Background
In order to facilitate a better comprehension of the role of social media in the healthcare domain, we present the social media, social media typologies and the use of social media in relation to healthcare, indicate the research gap and state our research questions.
Social media
6 Social Media Typologies
For the time being, there is only one typology of different social media categories, proposed by Kaplan and Haenlein (2010), and it is based on two dimensions, namely social presence/media richness and self-presentation/disclosure. Social presence refers to the type of contact that can be achieved while media richness indicates amount of information that can be shared. On the other hand, self-presentation/disclosure refers to revelation of personal information to impress others. Along these lines, Kaplan and Haenlein (2010) divide social media applications into six categories as shown in Figure 1.
Figure 1. Social Media Typology (Kaplan & Haenlein, 2010)
7 Social Media in Healthcare
The last couple of years are characterized by increasing research into social media in the healthcare domain. The research indicates that social media is used by both healthcare providers (Van de Belt, Berben, Samsom, Engelen & Schoonhoven, 2012) and users (Greene, Choudhry, Kilabuk & Shrank, 2011) and covers many different topics, such as the features of health groups (Bender et al., 2012) and specific diseases (Shaw & Johnson, 2011). The past studies focus on the social networking sites like Facebook (Greene et al., 2011), Twitter (Thackeray, Neiger, Burton & Thackeray, 2013) and those built exclusively for healthcare users to share their experiences (Wicks et al., 2010) or review healthcare providers (Reimann & Strech, 2010). Additionally, the research covers blogs (Shah & Robinson, 2011) and content communities such as Youtube (Prybutok, 2013). In the next section, we focus on the utilization of social media by healthcare users and the interactions this kind of utilization has involved. In doing so, we discuss several empirical papers and focus on the recent literature reviews that reflect the state of art in this field.
Social media use by healthcare users and providers
8 Several recent papers shed more light by integrating current findings from the literature (Ziebland & Wyke, 2012; Chou et al., 2013; Hamm et al., 2013; Moorhead et al., 2013). As such, this research stream attempts to conduct more broad (scoping) or detailed literature reviews and indicate the types of interactions taking place via social media in the healthcare field. Ziebland and Wyke (2012) identify seven domains as result of online patient experience, namely finding information, feeling supported, maintaining relationships with others, affecting behavior, experiencing health services, learning to tell the story, and visualizing disease. To identify these, they combine a literature review and a public user panel on how the people’s sharing of experiences online may affect their health. Although the literature review had been complemented by the public user panel, it remains unclear from the description of the methods who were the members of the panel and if this comprised solely of healthcare providers or also included healthcare users. Whereas some of the concepts they outline were already known from the past literature, Ziebland and Wyke (2012) suggest that learning to tell the story and visualize disease are less present but important features for healthcare users. This is significant as hearing about how others experience their health related issues and illustrating this through pictures and videos constitutes a powerful form of communication. In refining the results of the study, they also propose a results matrix illustrating different Internet applications and their use. While the study provides a good overview of how sharing and communicating with others can bring benefits to healthcare users, Ziebland and Wyke (2012) do not make the distinction between social media and other applications. Furthermore, the study looks at this from the perspective of source of information and does not consider with whom the interactions take place. However, it is important to know types of actors to be able to categorize these interactions and investigate their effects on the changing user’s relationship with the healthcare system.
9 and education, underlining that the education bit was more often based on sharing personal advice rather than official medical knowledge and guidelines.
Hamm et al. (2013) conducted a broad review to examine the use of social media by healthcare providers and users. Drawing on a sample of more than two hundred studies, the authors provide a review on the social media tools and interactions within social media. The review reveals over 60% of the studies examined discussion forums, followed by social networking sites and only a few focus on blogs. Whereas discussion forums were employed predominantly to facilitate self-care in relation to different conditions such as diabetes and cancer, the other types of social media platforms were used to share personal experiences. Although social media have been reported to be used for a variety of health conditions, most studies addressed the groups that concerned lifestyle and weight loss topics. One of the limitations of the review is that it fails to clearly explain the difference in the utilization by healthcare providers as compared to the utilization by healthcare users. In particular, they do not indicate who the actors in social media platforms are and how they interact with each other.
11
Table 1. Literature Review
Paper Main findings Comments
SM Categories
Actors Effects on healthcare Antheunis, Tates and
Nieboer (2013)
Healthcare users get social support, increase knowledge and exchange advice. Healthcare providers promote themselves and interact with other providers.
Yes Yes/No No
Andersen, Medaglia and Henriksen (2012)
Ziebland and Wyke (2012)
Chou et al. (2013)
Hamm et al. (2013)
Moorhead et al. (2013)
Healthcare users interact with providers to ask questions and make appointments.
Healthcare users find information, feel supported, maintain relationships with others, affect behaviour, experience health services, learn to tell the story, and visualize disease. Social media is used for health promotion and education.
Social media is used for self-care (diabetes and cancer) and to share personal experiences about weight loss and lifestyle.
Healthcare users share health related information and receive emotional support. They interact with each other and with providers.
12 Research Gap and Research Questions
Though the considered reviews are useful and provide greater insight, difficulties arise when an attempt is made to better understand the contents and type of social media interactions, which is important to be able to map the different types and understand their potential effects on the offline interactions. Furthermore, studies often focus on one category of social media only, in particular the social networking sites (Moorhead et al., 2013). Whereas emotional support and sharing experiences may be present in social networking sites, the question arises if the same types of use and what types of interactions are evident in other categories of social media. Moreover, the literature remains unclear on the types of actors who participate in the social media enabled interactions in the healthcare domain. In particular, the past research and the related reviews fail to look into how and for what purposes the healthcare users interact between each other and for what with the healthcare providers. It also remains unclear how similar or different is the use of social media in healthcare across different categories of social media. It is important to understand these differences and be able to map social media enabled interactions in healthcare. Bornkessel, Furberg and Lefebvre (2014) suggest that the of social media should lead to so called networked model in which both healthcare providers and users are utilizing social media for the benefit of mutual communication and better healthcare. On the other hand, Hwang & Christensen (2008) suggest that the social media enabled user networks offer a platform for transferring knowledge from providers to users, which, in turn, may even disrupt healthcare system. Different types of social media enabled interactions may have different impact on the behavior of healthcare users and providers and ultimately impact on the healthcare. Therefore, we contribute to future research on social media in the healthcare by providing more clarity on types of interactions across different categories of social media and the way they can be categorized. In this way, we extend earlier research by clearly showing types of interactions, participating actors, similarities and differences across different social media categories and proposing a typology. In doing so, we pose the following research questions:
1a) What types of interactions are enabled by social media among healthcare users and between healthcare users and healthcare providers?
1b) How and why do the interactions differ among different categories of social media?
13
Methodology
Research Design
14
15 Case Selection
16
Table 2. List of Cases Blogs:
Provider initiated User initiated
General Harvard medical blog E-patient Dave
Condition specific James Hamblin Diabetes mine
Social networking sites:
Provider initiated User initiated General Boston Children’s Hospital-Facebook Medhelp
Condition specific Dr. Eisenberg-Twitter Lose weight Jo! - Facebook Content communities:
Provider initiated User initiated
General Everydayhealth BeautifulBrwnBabyDol
Condition specific Endobariatric (Dr. Alvarez) Patient power (cancer) Collaborative projects:
Provider initiated User initiated
General WikiDoc Natural health wiki
Condition specific Street Medic Wikia
Virtual game worlds:
Condition specific Gameforscience: Vitalis Island STBBI Clinic
Virtual social worlds:
17
No. SM Type Case Link Description
1 Blog Harvard medical blog http://www.health.harvard.edu/blog/ Harvard medical school blog. Many health topics, news and views.
2 Blog James Hamblin, physician http://www.theatlantic.com/james-hamblin/ Preventive medicine (topics like gluten, anxiety, and caffeine)
3 Blog E-patient Dave http://www.epatientdave.com/ Cancer survivor, international keynote speaker. 4 Blog Diabetes mine http://www.diabetesmine.com/ A blog about diabetes from patient perspectives. 5 Social networking site Boston Children’s Hospital https://www.facebook.com/BostonChildrensHos
pital
US hospital specialized for children health care. 6 Social networking site Dr. Eisenberg Twitter https://twitter.com/drseisenberg Oncologist with more than 20.000 of followers. 7 Social networking site Medhelp http://www.medhelp.org/forums/list One of the largest online health communities. 8 Social networking site Lose weight Jo! - Facebook https://www.facebook.com/groups/1542980679
17196/
Group to exchange advice on losing weight. 9 Content communities Everydayhealth https://www.youtube.com/user/EverydayHealth Youtube channel of leading provider of digital
and wellness solutions.
10 Content communities Endobariatric (Dr. Alvarez) https://www.youtube.com/user/Endobariatric Youtube channel of popular plastic surgeon. 11 Content communities BeautifulBrwnBabyDol https://www.youtube.com/user/BeautifulBrwnB
abyDol
Youtube channel with topics on lifestyle etc. 12 Content communities Patient Power https://www.youtube.com/user/patientpower Youtube channel of Patient Power (on cancer)
community started by two cancer patients.. 13 Collaborative projects WikiDoc http://www.wikidoc.org/index.php/Main_Page One of largest medical wikis.
14 Collaborative projects Natural Health Wiki http://naturalhealth.wikia.com/wiki/Natural_He alth_Wiki#
Different information about health including topics on vitamins, food supplements etc. 15 Collaborative projects Street Medic Wikia http://medic.wikia.com/wiki/Main_Page Information resource focused on medical issues
treatment protocols regarding street medic. 16 Virtual game worlds Gameforscience: Vitalis Island
STBBI Clinic
http://www.gameforscience.com/vitalis/stbbi/ Virtual game world for education of teenagers about sexually transmitted infections.
17 Virtual social worlds Lil Angels Maternity Hospital http://www.imvu.com/rooms/little-angels-maternity-hospital-58
Maternity hospital offering virtual services. 18 Virtual social worlds Krystal’s Therapy Office
http://www.imvu.com/rooms/krystals-therapy-office
Counselling therapist offering health treatment in virtual environment.
19 Virtual social worlds First Health Hospital http://nl.imvu.com/rooms/first-health-hospital-ny?save_locale=1
General hospital offering virtual treatments. 20 Virtual social worlds G.Y.M. Body & Fitness
http://www.imvu.com/rooms/gym-body-fitness-9
18 We observed and collected the data from our cases by selecting posts and/or comments from the blogs, social networking sites, content communities and collaborative projects. Within each of these categories, we collected 400 posts and/or comments. In doing so, we attempted to follow the principle of 10 posts x 10 comments to maximize variation in the data. We were not able to collect the content for the virtual game worlds and virtual social worlds. This is due to the fact that the there was no content available that participants exchanged that was public. Therefore, we observed these categories. In order to do so, we registered and created an avatar in both the STBBI clinic of Vitalis Island – Game for science and the IMVU virtual social world. We used these avatars to browse through the analyzed platforms and make observations. We observed the virtual rooms and possibilities that users can perform with their avatars.
Data Analysis
19 The next step was to analyze the data with the coding scheme. Parts or entire posts and comments were selected and coded. In this way, we arrived at 1727 quotes subsequently assigned with a thematic code and a code for the nature of interactions. Furthermore, in order to assess if the participating actors were healthcare users or providers, we checked their social media profiles and made a deduction based on the content of posts and comments. Once all the data was coded, two researchers (supervisors) independently coded a sample of data containing 84 quotes. We then analyzed inter-rater reliability applying the approach for reliability of qualitative data proposed by Rust and Cooil (1994). The inter-rater reliability was calculated for the nature of interactions, namely social-emotional versus instrumental (0.91), general versus personal (0.90), and expressing versus asking (1.00). Given the fact that the Rust and Cooil (1994) approach is limited to five categories, for the nine thematic codes, we calculated inter-rater reliability for the five most frequent types of interactions in this sample, which led us to reduce it to 80 quotes. The thematic codes in question were health condition, online social grooming, lifestyle, non-health related and product recommendation/promotion. Inter-rater reliability was 0.98. Inter-rater reliability is considered as reliable when above 0.7 (Neuendorf, 2002), suggesting that our coding scheme is robust.
20 Code
No.
Code name Definition/Explanation Example
1 Health condition Conditions that indicate lack of health due to illness, injury or other physical or mental conditions. As such, it can include acute or chronic health conditions.
Testing for colorectal cancer: There are several ways to detect small, precancerous growths called polyps and early stage cancers before they spread from the colon to other parts of the body: Colonoscopy: This is considered to be the best test. A doctor inspects the entire rectum and colon using a flexible tube with a light and camera at its tip. If a cancerous or precancerous growth is seen, it can be removed.
2 Healthcare provider Focusing on physicians or/and hospitals, insurance companies and other types of healthcare providers.
A GOOD doctor will always listen to a patient !!! make sure you have a good doctor :)
The dentist even asked me to pay $45 extra because I supposedly did not bite well the first time so he had to use another mold (this was when he took the old to make permanent crown, and not that he did a 2nd crown)
A big 'shame on you' to a doctor who says your 'mild prescription' doesn't warrant glasses
3 Health product Discusses certain health products that can be drugs, health insurance packages or dietary
products/substances, and their effects.
for newborns I like the Huggies because they have the Cut out for umbilical cord. But when they get older I like PamperS. I buy Huggies sensitive wipes.
Actually Benadryl does a pretty decent job at shutting down a cough.
When a bad cough due to pneumonia or bronchitis is keeping me up at night, I'll take the codeine please. It's better than anything else.
4 Health policy/procedures Concerning government healthcare policy and discussions about the decisions, initiatives, plans and procedures aimed at achieving the society’s healthcare goals.
A better model for #Health Information Exchanges? #HealthBlog
http://aka.ms/Lygbpz #HealthITNews #Health_IT pic.twitter.com/KABQT5OaCW @drseisenberg Should we be troubled at how many are profiting in the healthcare "industrial complex" who actually provide no healthcare?
5 Suggestion for offline contact
Explicit question/suggestion for offline contact with healthcare provider.
You need to work through this with your own doctor and/or a counselor-mental health professional, not sit at home letting the internet feed your OCD driven paranoia.
21 Please check this out with your gynaecologist
6 Product
recommendation/promotion
Recommends or promotes certain products or services to others.
@drseisenberg Hi Dr. Eisenberg. We just published an important new cancer book. Can we send you a free copy of: http://www.amazon.com/Radical-
Remission-Surviving-Cancer- Against/dp/0062268759/ref=tmm_hrd_title_0?ie=UTF8&qid=1395699746&sr=8-1 … ?
Have you tried Skinnimaker Diet? (do a google search for it) It is a quick way to lose weight fast.
7 Lifestyle Concerns a style of living not
necessarily related to a disease. Often relates to the topics such as exercising, eating and weight loss, skin/hair treatments, etc.
See... I was gonna skip packing my own food and buy something that I shouldn't be eating for lunch today
I have been exercising for years and have always had toned arms on the thin side. I still have well-toned shoulders, hips and buttocks at 68 and feel great.
8 Online social grooming Acknowledges the importance of a person and his/her personal
experiences. It can be a simple ‘Thank you’ or emotional or inspirational type of content.
This video brought tears to my eyes! Keep on following your dreams, Aaron! Lots of prayers for this very special baby and parents and for the wonderful care given by all at BCH.
Thank you . Great video. Way to go, Meb
9 Non-health After you lost all your weight what did u do with the clothes that no longer fitted u? Im from naples too!
22
Results
In line with the research questions, the results part is divided into two sections. In the first section, we describe the results and answer our first research question. Precisely speaking, we answer the question: 1a) What types of interactions are enabled by social media among healthcare users and between healthcare users and healthcare providers? and 1b) How and why do the interactions differ among different categories of social media? We start by providing a general review of the interactions identified in the four categories in which we have analyzed the content, namely blogs, social networking sites (SNS), collaborative projects, and content communities. In this part, we briefly describe the identified topics and the nature of interactions, in general and per categories. We illustrate the results through Table 5. We go on to discuss how and why the interactions are different or similar in relation to the general picture and from one another. In this part, we also discuss the social virtual worlds and virtual games in which we were not able to analyze the content but explored them by using of avatars. That answers the second part of our first research question. It is important to analyze the impact of different social media categories on types of interactions and participating actors. It can provide us with an understanding of topics and types of interactions that are specific and suitable for certain categories, but maybe not for other. Furthermore, it can reveal us what categories of social media are used by different actors and for what purpose. In comparing individual categories against general picture, we exclude the category that we compare from overall results as inclusion of the category itself skews the overall results. With this, we answer both parts of our first research question. We then move onto discussing and proposing a typology for the social media enabled interactions in healthcare. In doing so, we rely on the results that have been arrived at in relation to our first research question to answer the second research question: How can we categorize the social media enabled interactions in healthcare and place them in a typology? In this section, we propose six archetypical interactions, and propose dimensions on which we can place these archetypes into a typology.
General overview of topics and nature of interactions
23 section, these were social-emotional versus instrumental way of communication, general discussions versus personal (own) health condition and asking versus answering questions. We continue by presenting the identified topics in general and per category of social media. Nine thematic topics in total are used, all listed and defined in the methods section. 1727 quotes in total are identified across the four categories in which we have analyzed the content. The most frequently discussed topic was “health condition” with 31%, followed by “social grooming” and “lifestyle” topics with 20% and 16%, respectively. On the other end of the spectrum as the least frequently discussed topic was suggesting offline contact with healthcare provider with the share of 1%, followed by the topics of product promotion and health policy both with the share of 4%. Overall, the topics were discussed predominantly in an instrumental way (68%) or a social-emotional way (32%). 59% of these discussions were general, while 41% were related to particular individuals’ specific conditions. Whereas in 93% of posts the tone was expressing, asking was relatively low with 7%. The messages were posted mostly by healthcare users with 81% and healthcare providers with the share of 19%. Table 5 provides an overview of the results.
Table 5. Review of Topics and Nature of Interactions
Topic Overall Blogs SNS Collaborative Content communities
Health condition 31% 22% 23% 45% 33%
Healthcare provider 2% 2% 4% - 1%
Health product 14% 20% 8% 27% 2%
Health policy/procedures 4% 8% 2% 8% -
Suggestion for offline contact 1% - 3% - - Product recommendation/promotion 4% 5% 2% - 9% Lifestyle 16% 12% 13% 20% 18%
Online social grooming 20% 21% 36% - 23%
Non-health 8% 10% 9% - 14%
Nature of interactions Overall Blogs SNS Collaborative Content communities
Social-emotional 32% 38% 47% - 44% Instrumental 68% 62% 53% 100% 56% General 59% 67% 35% 100% 35% Personal 41% 33% 65% - 65% Asking 7% 6% 8% - 12% Expressing 93% 94% 92% 100% 88%
Actor Overall Blogs SNS Collaborative Content communities
Healthcare user 81% 93% 71% 78% 83%
24 How and why blogs differ from the general picture
The blog related results for the nature of interactions are consistent with the overall ones in terms of asking versus expressing and general versus personal, however, we note that for the share of interactions about general health topics is somewhat higher with 67% as compared to the average of other categories with 57%. Similarly, the interactions about health products and health policy appeared more frequently in the blogs. To be precise, the interactions about health policy take up 8% and about health products 20% in the blog related as compared to the average of other categories where the shares are 3% and 12%, respectively. The share difference as far as the interactions about health products are concerned was due to a user initiated social media platform focusing on diabetes. In their posts, the users discuss different health products for managing diabetes, which explains the higher share of this topic in the interactions we analyzed. The difference pertaining to the topic of health policy can also be explained by a single user initiated social media platform covering different issues from patient engagement to health information exchange, as reflected by the following quote:
We all pretty much know that the ONC invested over 600 million into health information exchange and guess what it failed. Isn’t the blue button just their attempt to get patients to carry the torch for them? (Anonymous healthcare user, E-Patient blog)
Another finding related to the use of blogs that could explain the high share of general and instrumental type of interactions is the educational nature of posts in the blogs analyzed. This stems, in particular, from the instrumental way in which the information is presented and directed towards the wider public. It is interesting that both user and provider initiated blogs present information in this fashion.
25 How and why SNS differ from the general picture
We observed that the interactions in SNS were personal and social-emotional to a higher degree than indicated by the overall results. Precisely speaking, the shares of social-emotional and personal interactions in SNS were 47% and 65%, respectively, as compared to 27% and 33% shown by the average of other categories. Another notable difference in this category was the exceptionally high share of online social grooming. The high share of interactions involving discussions on personal health conditions could be explained by the user initiated social medial platforms, especially Medhelp. Although user-initiated, Medhelp established cooperation with a number of healthcare providers who participate in the platform to answer specific questions of healthcare users. Hence, it includes both very specific descriptions of personal health conditions and questions to healthcare providers, as illustrated by the following quote:
Could there be an underlying cause for ALL of these conditions that my doctors are just not connecting the dots on?? A lot of other things I have noticed and learned to live with is, excessively dry skin, dry brittle hair, hair loss, insomnia, easily fatigued, join pain in my hips, back, and legs… (Healthcare user MrsTxMoose, Medhelp)
Moreover, the social-emotional sphere was more prominently featured than in the overall results, which can be attributed to a high share of social grooming across cases within this category. However, different types of social grooming were observed here and range from expressing simple thanks to providers for answering health related questions to healthcare users giving emotional and inspirational support to one another. Emotional support was expressed to other healthcare users whose family members, namely children, were coping with health conditions, while inspirational support was provided in a humorous fashion, by sharing jokes and inspiring people to live more healthily or cope with their diseases.
26 ask direct and precise questions. On the other hand, the limited length of messages and prevailing interactions on specific conditions make SNS less suitable for educating wider public.
How and why content communities differ from the general picture
This category of social media exhibits divergent pattern in terms of the general versus personal nature of interactions. Whereas the overall results from other categories indicate the share of interactions about personal health conditions of 33%, the percentage with content communities is significantly higher and amounts to 65%. Content communities also show a somewhat higher share of asking questions with 12% as compared to the average of other categories with 5%. Furthermore, they show a high share of product promotion with 9% as compared to the average of other categories with 2%. On the other hand, there are fewer discussions about health products with the share of only 2% as compared to the average of other categories with 19%.
More discussions over personal health conditions occurred across cases within this category. In one provider initiated platform, the healthcare provider posted a number of videos explaining different weight loss surgical procedures. The posts generated many responses comprising both the expressions of personal experiences and opinions as well as questions. The questions that the healthcare users posed to the healthcare provider were specific and concerned the details of particular health procedures and treatments of specific conditions. This also explains a bit higher share of asking questions within this category. Whereas in this case the expressing and asking was, in fact, the healthcare users’ response to the provider, in another case focusing on lifestyle a healthcare user posted videos about her own activities which others discussed by either commenting on the user’s lifestyle or asking the related questions. This is illustrated by the following quotes:
Are there any "side effects" that you might know of for transgender people? I am a transgender man, and am currently on testosterone (have been for almost 4 years)- (Healthcare user Suarezi, Endobariatric)
27 Interestingly, product promotion featured rather prominently in this category, taking form of videos from both providers and users but also comments in which the users promoted health related products or just spammed the discussion with non-health related products. On the other hand, there were few discussions about health products without the promotion aspect to them. This is because the health products were discussed in a promoting/recommending manner as described above and illustrated by the following quote:
Have you tried Skinnimaker Diet? (do a google search for it) It is a quick way to lose weight fast (Healthcare user Rakesh Gurung, BeautifulBrwnBabyDol)
Although being used for many discussions on personal health conditions this category of social media is also suitable for general presentations of health and lifestyle topics, in particular by healthcare providers. Additionally, healthcare providers could use it in an educational manner to describe certain health procedures such was the case in the weight loss surgery or the example involving educational videos on cancer treatment. Furthermore, this category of social media is particularly suitable for presenting lifestyle topics such as exercising and eating routines as it enables the most realistic illustrations of how to do particular exercises or prepare meals.
How and why collaborative projects differ from the general picture
28 For example, in one of the user initiated collaborative projects only three healthcare users interacted amongst themselves. The principal reasons for this is that collaborative health projects represent communities where people can share knowledge on medical issues that may be highly specialized and, hence, require advanced medical knowledge to participate in the interaction. Therefore, this category of social media is suitable for the health related interactions that are general, expressing or instrumental in nature, created by smaller groups of health experts who can be either healthcare users or providers and directed towards the public.
How and why virtual game worlds differ from the general picture
The game we studied was different compared to the overall results in the sense that it focuses on the interactions between healthcare users, specifically teenagers on a single health related topic. On other hand, it also has healthcare professionals provide medical knowledge to healthcare users, but in a simulated environment. As such, it does not provide as direct a link to healthcare providers as the other categories can. Given that the game users can add one another as friends within the game and communicate directly, the game, therefore, also enables the interactions of different nature wherein the healthcare users engage with each other. Additionally, it has an educational character and is, thus, suitable for targeting specific groups of healthcare users on particular health conditions, such had been the case here, but also lifestyle topics, which can motivate and improve prevention or/and treatments. By contrast, it seems less fitting for discussions on topics such as healthcare providers, medications or health products. However, as we were not able to analyze the actual interactions between the players, we cannot make any definitive conclusions on the matter. It is possible that the healthcare users interact through the game and discuss their personal experiences on the topics. There was no publicly available content and the communication between users is possible with their active participation.
How and why virtual social worlds differ from the general picture
29 not only able to interact with healthcare users but also virtually simulate their premises and services. One of the provider initiated platforms presented their clinic in the virtual social world and offered offline services should the users be satisfied. The category is suitable for these types of interactions, especially in relation to particular health conditions like those requiring psychological counselling, as is offered by one of the providers, where physical presence is not needed to provide treatment.
How and why social media categories are different or similar from another
In this section, we shall discuss how social media categories are similar or different from one another. We rely on the percentage differences identified on the topics and nature of interactions as provided in the Table 5. Collaborative projects were found to be different from all other social media categories including virtual game worlds and virtual social worlds. Whereas the other categories enable and exhibit different types of interactions, collaborative projects focus on the instrumental, general and expressing only. As noted it the earlier elaboration on this category, it stands out from the others also in the sense that it involves fewer participants. As opposed to the other categories, participation in expert medical collaborative projects may require advanced knowledge and an effort from the participants. On the other hand, it is similar to the provider-user interactions in blogs that are also characterized by the instrumental and general nature of interactions which serve to educate the wider public on health related topics. A similar pattern is reflected in the nature of interactions regarding social-emotional versus instrumental across all categories except for collaborative projects.
30 evident in the cases of communication with healthcare providers wherein they provided the answers to health condition-specific questions but, by contrast, not found with blogs. Although we were unable to observe the contents of the virtual social worlds, they reflect a similar pattern as they also enable direct and specific communication with healthcare providers about health conditions, as well as discussions on personal health related issues. This is because healthcare providers are present in their virtual rooms and can communicate with users.
As far as the topics of the studied interactions are concerned, discussing health conditions was prevalent in collaborative projects and content communities. Moreover, in terms of the health conditions, the communication in all categories of social media that we analyzed was focused on chronic conditions, diabetes and cancer in particular. Suggestion for offline contact with healthcare provider was extremely poorly represented or not represented at all across the categories, its highest share being that of 3% in the SNS interactions. Interestingly, suggestion for offline contact was highly specific to one particular case of a user initiated SNS with a heavy involvement of healthcare providers. In the case, this type of interaction concerning personal health conditions was very appropriate as the interactions reflect deep discussions and trust towards recommendations of healthcare providers. Another striking finding was that a similar share of interactions regarding product promotion was calculated for both content communities and blogs. However, whereas with blogs the product promotion was an integral part of the discussions on certain health conditions, in content communities it mostly took the form of spam messages disrupting the interactions between participants and, at times, being off-topic. A high share of social grooming in SNS sets them apart from the other categories and was present in both user-provider and user-to-user interactions. The type of social grooming, however, was no different than that found in other categories and included providing social and emotional support, as well as attempted inspiration to others.
Typology of social media enabled interactions
31 distinguishable from others. In this way, archetypical interactions were identified representing the typical examples of the particular types of the social media enabled interactions in healthcare. Hence, we applied the contrastive categorization as we defined our archetypes in terms of types that emphasize the contrast with the competing options. In doing so, we analyzed them on the basis of topics, actors and nature of interactions as presented in the Table 5. First of all, we briefly describe the archetypical interactions and illustrate them with examples, concluding with the Table 6 summarizing main characteristic of these archetypes. In the Table 6, we also indicate categories of social media in which these archetypical interactions are present. Furthermore, we categorize archetypes in 2x2 typology as shown in the Figure 3. We categorize archetypes along two dimensions, namely control and scope of interactions. By control, we refer to way of communication that is present amongst actors. In this way, we differentiate between formal and informal control. In formal, actors address each other in official and formal way with strong hierarchy in the communication. On the other hand, informal is characterized by free and equal participation of all actors who freely express themselves and communicate with others. As such, it reflects communication in which hierarchy is not so strong in the relationship between actors and they freely address each other. Our second dimension is scope of interactions. In this dimension, we distinguish between broad and deep interactions. Broad interactions are characterized by discussing issue in a general way and do not touch upon details of the topics discussed by neither of the parties involved in the communication. Deep interactions focus on discussing certain topics in very detailed way.
Grooming fellow sufferers
32
Post (user): “This is so me! Sharing photo: "I work out because it is good for me. Also, because I like to eat. A lot. (Healthcare user Natalie Anne Powell, Lose Weight Jo)
Comment (user): Yep I'm with ya sista!! :) (Healthcare user Jo Martin, Lose Weight Jo) ;
Personal health condition resolving
This archetypical interaction is initiated by healthcare users and directed towards healthcare providers. The healthcare users describe their symptoms in detail and pose very specific questions to the providers. The interactions of this type always focus on personal health conditions. What is characteristic of this type of interaction is that the communication between the healthcare users and providers goes on until the healthcare users are satisfied with its outcome. Therefore, they are deep and have corrective character. Even though the interaction is mainly about the exchange between a particular healthcare user and healthcare provider, other healthcare users also engage and provide specific comments/suggestions to the healthcare user who had initiated it. These interactions involve both expressing and asking done in instrumental way. This type of communication was especially associated with a particular SNS studied but also present in content communities and social virtual worlds. A classic example of the archetypical interaction is provided below:
Post (user): ….now ive been experiencing heart palpitations , weakness , fatigue , dizziness , and very rarely shortness of breath , loss of appetite im afraid…. why do I have all these other symptoms? (Healthcare user jakethasnake, Medhelp)
Comment (provider): I might recommend checking your pulse whenever you have these spells to see whether you have a fast or irregular heart rate….(Healthcare provider CCFHeartMD27, Medhelp)
Comment (user): Palpitations are very often benign and your other symptoms could be caused by a host of benign things, including anxiety (Healthcare user marlinspike, Medhelp)
Teaching users with occasional debating
33 showing appreciation for the content. In addition, the healthcare users provide their opinions on the subject and debate among themselves and with the healthcare providers. The interactions of this sort focus on a range of medical conditions and lifestyles but most frequently pertain to the symptoms and management of chronic diseases. The providers always address healthcare users in an instrumental way, attempting to address health topics in a general way. Therefore, interactions remain broad and are directed towards both corrective and preventive dealing with one’s own health. What characterizes this type of interaction is that the healthcare providers seldom engage in further discussions/interactions with the healthcare users and remain distant. This archetype may be found with different social media categories but is the most predominant in blogs and content communities. Here is an example of the type of interaction:
Provider (post): ….That name is dermatographia urticaria, usually just called dermatographia or dermographism (literally "writing on the skin"). It’s a type of "trauma-induced urticaria," but the trauma in this case can be almost nothing—basically an allergic reaction to minimal physical stimulation….. (Healthcare provider Dr. James Hamblin, James Hamblin blog)
User (comment): …Very interesting article and it is kinda interesting that these allergy related textures can be created on skin with mild scratching… (Healthcare user tstev, James Hamblin blog)
Specialized and expert knowledge building
34
Provider (post):. …For these problems to be diagnosed as ADHD, they must be out of the normal range for a child's age and development….(Healthcare provider, Dr. C. Michael Gibson, WikiDoc)
User (comment): …[[ADHD]] is a long-term, chronic condition. If it is not treated appropriately, ADHD may lead to… (Healthcare user Kalsang Dolma, WikiDoc)
Lifestyle guru advisors
This archetypical interaction is of educational character reflected through the detailed showing of how to do, for example, certain exercises and how the exercises would affect one’s weight and healthy lifestyle. It reflects communication amongst healthcare users as well as between users and providers. The healthcare users attempt to guide others and promote healthy lifestyle by provoking discussion about it. In doing so, they talk and present their own experiences as well as general topics. Also, they both express and ask questions on the topics. The healthcare providers promote healthy lifestyle through fitness and food related topics. Both users and providers do it in a broad way and focus on topics that are directed towards prevention of one’s own health. The interactions of this type entail both the social-emotional and instrumental way of expressing. They are found in SNS and content communities. Here’s an example of the archetype:
Post (user): https://www.youtube.com/watch?v=TFssp4kG_8M – video showing exercises for arms. (Healthcare provider Everydayhelth, Everydayhealth)
Comment (user): Thank you Holly! Yes, this is helping me to tone my arms (Healthcare user dylan rantucci, Everyday health)
Comment (user): …I have been exercising for years and have always had toned arms on the thin side…. (Healthcare user patc135, Everyday health)
Reviewing healthcare products and providers
35 with just a mention of certain provider or product and sometimes they go into details discussing certain product. What is characteristic of this type of interactions is that it is commonly embedded in discussions on health conditions. As such, they reflect corrective character in dealing with one’s own health. This type of interaction is found with SNS, blogs and content communities.
Comment (user):… It was done at Stanford by Dr. Steinberg also. They were great there…(Healthcare user fleuve8, Patient Power)
Comment (user):… You can tell that Dr Jeff In Oregon is very concerned and focused on helping his patients…(Healthcare user Daniel Pearson, Patient Power)
Table 6. Summary of Archetypical Interactions Name Grooming fellow
sufferers Personal health condition resolving Teaching users with occasional debating Specialized and expert knowledge building Lifestyle guru advisors Reviewing healthcare products and providers Short description Focused on personal conditions or
lifestyles and the persons who receive social and emotional support from others. Focused on personal health conditions and getting advice from healthcare providers. Focused on teaching healthcare users on the range of health conditions. Focused on building knowledge on different medical issues. Focused on guiding others and promoting healthy lifestyles. Focused on reviewing healthcare products and providers.
Actors User-to-user Provider-to-user; user-to-user Provider-to- user; user-to-user Provider-to-user; user-to-user Provider-to-user; User-to-user User-to-user Nature of interactions Social-emotional and personal. Instrumental and personal. Instrumental and general. Instrumental and general. Both social-emotional and instrumental, personal. Personal and general, as well social-emotional and instrumental. SM categories SNS, content communities and blogs SNS and content communities Blogs and content communities Collaborative projects SNS and content communities SNS, content communities and blogs Purpose of interactions Corrective/ preventive Corrective Preventive/ corrective
36
Figure 3. Typology of Social Media Enabled Interactions in Healthcare
Scop
e
of I
nt
er
act
ion
s
Type of Control
Informal Formal Broad Deep Grooming fellow sufferers Life style guru advisors Reviewing healthcare products/providersTeaching users with occasional debating
Personal health condition resolving
37
Discussion
In this section, we discuss our findings in view of earlier literature. Precisely put, we first consider how our findings relate to it in terms of the types and actors in the interactions across different categories of social media by discussing each of our archetypes and how they compare to past literature. Subsequently, we discuss how our archetypical interactions and typology relate to the general typology of social media (Kaplan & Haenlein, 2010).
38 users often use social media to share their experiences about weight loss and lifestyle, as is reflected by the archetypical interaction “lifestyle guru advisors”. Whereas Hamm et al. (2013) did not specifically refer to particular social media categories we contribute to the literature by providing more clarity in this respect. Finally, while Ziebland and Wyke (2012) suggest that people mainly use specialized health platforms to discuss their healthcare providers, our archetypical interaction “reviewing healthcare products and providers” does not appear to be necessarily limited to a particular category of social media but, rather, spans across SNS, blogs and content communities.
In regards to the typology of Kaplan and Haenlein (2010), we find that the archetypical interactions characterized by the user-to-user interactions and the social-emotional sphere to be prevalent in social networking sites and content communities On the other hand the interactions involving knowledge building and healthcare providers teaching healthcare users are characterized by the instrumental way of communication. These types of interactions were more prevalent in collaborative projects and blogs. Such a trend in interactions reflects the Kaplan and Haenlein categorization (2010), in particular in relation to the dimension of social presence/media richness. Consequently, it follows that the social-emotional sphere of interactions is dominant in the social media with greater social presence/media richness. Despite this, our findings reveal that the archetypical interactions may be taking place in different categories of social media.
Theoretical implications
39 we categorize the interactions and place them in a typology, which was not done in any of the mentioned past literature. This facilitates better understanding and provides a foundation for future research of social media enabled interactions in healthcare. We also add to the general typology of social media (Kaplan & Haenlein, 2010). Although some archetypical interactions can be traced back to specific social media categories, we also see different archetypical interactions spanning across different social media categories, at least in the healthcare context. Hence, our results contribute to further exploration and possibly refining the typology of social media categories according to the types of interactions.
Practical implications
Social media interactions in the field of healthcare also have practical implications. This holds especially in relation to healthcare providers and policy makers who can use our findings and the typology of the social media enabled interactions to provide a better care and communication with healthcare users. They can learn from our findings on how the healthcare users are using different categories of social media and what types of interactions are suitable for every one of these. In this way, they can understand the needs of healthcare users better and adjust their own use of social media in communicating with them. Furthermore, they could get a better understanding of the content discussed by healthcare users and how this may affect healthcare providers. Hence, the research could assist healthcare providers in refining their own strategies for the use of social media in healthcare. Similarly, having greater insight and comprehension of the online behavior of healthcare users could serve as the basis for the relevant policy makers to better target different public health interventions. In particular, they can understand what types of interactions transpire in which categories and target their interventions accordingly.
Limitations and future research
40 their generalizability is limited. Therefore, the future research could include a larger sample, including mobile application, and test our typology in a quantitative manner. Secondly, in categorizing the social media interactions, we focused on the ones between healthcare users and healthcare providers, not covering those among healthcare providers. Consequently, the future research might enrich our typology by looking into the interactions that transpire between healthcare providers only. Finally, we were unable to find out how the use of social media for interacting with other healthcare users and providers affects the users’ relationship and offline interactions with healthcare providers. Christensen and Hwang (2008) as well as Hwan (2009) suggest that the social media enabled networks may impact on the healthcare system as we know it. Therefore, by clearly identifying and categorizing the social media enabled interactions in healthcare, we provoke a discussion on the potential disruption via social media. Although not viewed as either products or services, social media are changing the interactions between healthcare users and providers, which could ultimately lead to interfering with healthcare in the ways different from those envisaged by the theory of disruptive innovation. The disruptive innovation theory is focused on new products, services and business models (Christensen, 1997; Christensen & Raynor, 2003) and we incite further research into the evolving types of interactions via social media as a potential novel disruptive force in the healthcare domain.
Conclusion
41
References
Adams, S. A. (2010). Revisiting the online health information reliability debate in the wake of “web 2.0”: An inter-disciplinary literature and website review. International Journal of Medical Informatics, 79(6), 391-400.
Andersen, K. N., Medaglia, R., & Henriksen, H. Z. (2012). Social media in public health care: Impact domain propositions. Government Information Quarterly, 29(4), 462-469.
Andersson, A., Elg, M., Idvall, E., & Perseius, K. (2011). Five types of practice-based improvement ideas in health care services: An empirically defined typology. Quality Management in Health Care, 20(2), 122-130.
Antheunis, M., L., Tates, K., & Nieboer, T., E. (2013). Patients' and health professionals' use of social media in health care: Motives, barriers and expectations. Patient Education & Counseling, 92(3), 426-431.
Aral, S., Dellarocas, C., & Godes, D. (2013). Social media and business transformation: A framework for research. Information Systems Research, 24(1), 3-13.
Arrow, K. J. (1963). Uncertainty and the welfare economics of medical care. American Economic Review, 53(5), 941.
Bishop, J. (2007). Increasing participation in online communities: A framework for human– computer interaction. Computers in human behavior, 23(4), 1881-1893.
Chou, WS., Prestin, A., Lyons, C., & Wen, K. (2013). Web 2.0 for health promotion: Reviewing the current evidence. American Journal of Public Health, 103(1), 9-18.
Christensen, C. M. (1997). The innovator’s dilemma. Boston, MA: Harvard Business School Press.
Christensen, C.M., Raynor, M.E. (2003) The innovator’s solution, creating and sustaining successful growth. Boston, MA: Harvard Business School Press.
Cram, W. A., & Brohman, M. K. (2013). Controlling information systems development: A new typology for an evolving field. Information Systems Journal, 23(2), 137-154.
Eysenbach, G. (2008). Medicine 2.0: Social networking, collaboration, participation, apomediation, and openness. Journal of Medical Internet Research, 10(3), 22-22.
Faraj, S., von Krogh, G., Lakhani, K., & Monteiro, E. (Eds.). (2014). Collaboration and Value Creation in Online Communities (Special Issue). Calls for Papers. Information Systems Research. Retrieved from: http://pubsonline.informs.org/page/isre/calls-for-papers
42 Glaser, B., & Strauss, A. (1967). The discovery of grounded theory. London: Weidenfield &
Nicolson.
Greene, J. A., Choudhry, N. K., Kilabuk, E., & Shrank, W. H. (2011). Online social networking by patients with diabetes: A qualitative evaluation of communication with facebook. Journal of General Internal Medicine, 26(3), 287-292.
Hamm, M. P., Chisholm, A., Shulhan, J., Milne, A., Scott, S. D., Given, L. M., & Hartling, L. (2013). Social media use among patients and caregivers: A scoping review. British Medical Journal Open, 3(5).
Hawn, C. (2009). Take two aspirin and tweet me in the morning: how Twitter, Facebook, and other social media are reshaping health care. Health affairs, 28(2), 361-368.
Hwang, J., & Christensen, C. M. (2008). Disruptive innovation in health care delivery: a framework for business-model innovation. Health Affairs, 27(5), 1329-1335.
Johnson, G. J., & Ambrose, P. J. (2006). Neo-tribes: The power and potential of online communities in health care. Communications of the ACM, 49(1), 107-113.
Johnston, Allen C., James L. Worrell, Paul M. Di Gangi, & Molly Wasko. (2013). Online health communities: An assessment of the influence of participation on patient empowerment outcomes. Information Technology & People, 26(2), 213-235.
Kaplan, A. M., & Haenlein, M. (2010). Users of the world, unite! the challenges and opportunities of social media. Business Horizons, 53(1), 59-68.
KPMG International. (2011). Increasing importance of social media in healthcare. Retrieved from:
http://www.kpmg.com/Global/en/IssuesAndInsights/ArticlesPublications/Issues-monitor-healthcare/Documents/im-healthcare-october-2011.pdf
Mays, G. P., Scutchfield, F. D., Bhandari, M. W., & Smith, S. A. (2010). Understanding the organization of public health delivery systems: An empirical typology. Milbank Quarterly, 88(1), 81-111.
Maxwell, J. A. (2005). Qualitative research design: An interactive approach. Thousand Oaks, CA: Sage Publications.
Miller, A. R., & Tucker, C. (2013). Active social media management: the case of health care. Information Systems Research, 24(1), 52-70.
Moorhead, S. A., Hazlett, D. E., Harrison, L., Carroll, J. K., Irwin, A., & Hoving, C. (2013). A new dimension of health care: Systematic review of the uses, benefits, and limitations of social media for health communication. Journal of Medical Internet Research, 15(4), e85-e85.
