E-Word-of-Mouth on health social networking sites: An opportunity for tailored health communication

Authors


Bo Liang, Department of Marketing, University of Utah, USA.

E-mail: Bo.liang@utah.edu

ABSTRACT

Because of the limitation of time and resources, health professionals are often not able to provide support for consumers via personal communication. Consumers are increasingly active in managing their own health. E-word-of-mouth (eWOM), or consumer-generated content, on health social networking sites (SNSs) provides a new opportunity for health communication tailored to consumers' personal needs. Using netnography, a well structured internet-based ethonography, we analyze discussion threads posted to an obesity support group on a popular health SNS based in the US. We classify e-group members as support seekers and providers. Our results identify two major features of eWOM on health SNSs that facilitate tailored health communication. First, eWOM on health SNSs provides the types of support (informational and emotional) that help support seekers actively cope with heath-related problems and strains. Second, eWOM on health SNSs contains massages rich with personal experiences of support providers that offer solutions for problems that support seekers face in everyday situations. As support seekers participate in SNSs, their opportunity for learning means that they can become ‘information elite’, evolving from support seekers to support providers. We discuss important implications for policymakers and marketers for learning from and participating in eWOM on health SNSs. Copyright © 2011 John Wiley & Sons, Ltd.

INTRODUCTION

Health communication is an important factor in disease prevention and health promotion. Research on health communication has shown that the most effective health communication is tailored health communication (THC). That is, message framing should be congruent with individual message receivers' needs and characteristics (e.g. health locus of control), skills (e.g. reading skill), abilities (e.g. information literacy), and motivations (Salovey and Williams-Piehota, 2004; Rimer and Kreuter, 2006; Rothman et al., 2006; Hibbard et al., 2007; Updegraff et al., 2007).

In today's healthcare environment, it is difficult for consumers to get all the support they need through communication with their doctors within a 20-minute clinic visit. Social network sites (SNSs) as virtual communities in which an individual can identify and communicate with ‘friends’ or informed others (Boyd and Ellison, 2007) are becoming a common means for health communication. Nearly half of internet users who have searched for online health information reported that they have participated in health-related communication (reading or posting messages) on SNSs (Fox and Jones, 2009). SNSs may be of value in facilitating tailored health communication.

Previous research has demonstrated the benefits of health communication for consumers via early SNSs (e.g. usenet and bulletin board systems), including psychosocial (e.g. decreasing stress Sharf, 1997; Braithwaite et al., 1999; Monnier et al., 2002; Maloney-Krichmar and Preece, 2005; Im and Chee, 2006; Donelle and Hoffman-Goetz, 2008; Eichhorn, 2008; Ginossar, 2008; Macias and McMillan, 2008) and instrumental benefits (e.g. 24-hour availability and anonymity) (Cline and Haynes, 2001). Early SNSs have limited functions for personal interaction compared with current SNSs. Thus, studies on early SNSs have primarily focused on the analysis of individual messages (posts or emails) and ignored the analysis of the interaction among users. Examination of user interactions on current SNSs can provide insights into the ways in which user benefits are generated and experienced, and the extent to which responses from ‘others’ are tailored to the questions that individual users ask.

There are a variety of platforms for electronic World-Of-Mouth (eWOM), a marketing concept defined as consumer-generated content posted on the internet, such as product review websites and SNSs (Bronner and De Hoog, 2010). Previous research on eWOM has been heavily focused on eWOM about high-tech and media products (e.g. movie and video games) on product review platforms (e.g. amazon.com; Chen and Xie, 2008). There has been little research on eWOM related to medical- and health-related products and issues on social-networking platforms. This study contributes to the fields of health communication and marketing by providing theoretical insights and empirical evidence about the major features of eWOM on health SNSs, focusing specifically on the discussion/interaction among users. In doing so, this study examines the types of social support and the content of messages transmitted via health SNSs and the extent to which these are tailored to users' individual needs.

THEORETICAL BACKGROUND AND PROPOSITIONS

We first provide some theoretical context for examining the interaction between users on health SNSs (e.g. patientslikeme.com that are specialized for health-related issues). In dealing with health-related issues, it is important to understand the strategies individuals may employ to cope with their problems and concerns. In understanding the role that health SNSs may play in such coping, it is useful to examine both the social support and the content of messages that are transmitted on these sites.

Social support tailored to coping strategies

The major activity of users on health SNSs is to post messages in a specific thread. Individual posters can ask for support in one thread and respond to a request for support in another thread. Thus, individual posters may be both support seekers and providers depending on the content of the messages they post. Two types of support can be transmitted through online communities: informational support (e.g. advice, suggestions, or useful information) and emotional support (e.g. comfort, relief, and encouragement) (Maloney-Krichmar and Preece, 2005; Donelle and Hoffman-Goetz, 2008; Ginossar, 2008).

According to Folkman and Lazarus (1980), people use two major coping strategies: problem-focused and emotion-focused coping. Problem-focused coping aims to ‘remedy or conquer the stress-producing problem itself’ (Weiten et al., 2003: 120). Problem-focused coping includes four steps: identifying the problem, finding and evaluating alternative courses of action, selecting the most appropriate course of action, and taking actions (Weiten et al., 2003: 120). We argue that health communication via SNSs can assist participants in navigating these steps. First, support seekers initiating threads about their problems are identifying a problem. Second, information about solutions to problems from support providers can assist in finding out about and evaluating alternative courses of action. After discussing problems with support providers, support seekers may have sufficient information about alternative courses of action that they have the confidence to make a choice and take action.

Emotion-focused coping is aimed at ‘reducing or managing emotional distress that is associated with (or cued by) the situation’ (Carver et al., 1989). Previous research has shown that people who get emotional support demonstrate less emotional distress (Kulik and Mahler, 1993; Krumholz et al., 1998; Alferi et al., 2001). However, people do not always accept all the emotional support they receive; they may selectively attend to emotional support from the right persons at the right time (Taylor, 2003). In daily life, people passively receive emotional support from members of close social networks, such as families and intimate friends, which may lead them to feel overburdened by the amount of advice and sympathy they receive (Taylor, 2003). By initiating threads, support seekers actively seek support from virtual ‘friends’, thus, the emotional support they receive may be from the right persons at the right time.

The theory of motivated reasoning suggests that coping strategies are shaped by individuals' motivations. Consumers' motivations create both accuracy and directional goals with regard to information search (Kunda, 1990). Individuals are motivated to acquire information that provides a reasonable solution to their health problems and support that sustains them through the process of dealing with their health problem (Mason and Scammon, 2011). For these reasons, they are likely to turn to trusted sources. SNSs may be a highly trusted source for both informational and emotional support.

Thus, our first proposition is discussed as follows:

eWOM on health SNSs facilitates tailored health communication by providing the types of support that participants need for both emotion-focused and problem-focused coping. Specifically, eWOM on SNSs provides informational support for problem-focused coping and emotional support for emotion-focused coping.

Message content tailored to learning strategies

Situated learning is a learning strategy by which knowledge and skills are learned in the contexts where they are applied in everyday situations (Lave and Wenger, 1991). Education research has shown that online social interaction can be an effective context for situated learning, where students not only learn from each other but also actively apply what has been learn to real situations (e.g. Hung, 2002 and Halverson, 2009). In the environment of health SNSs, support seekers learn health knowledge when they seek solutions to real-life problems, such as how to balance medical (e.g. taking medicine or surgeries) and nonmedical (e.g. exercising) treatments. Therefore, we argue that health communication through health SNSs facilitates situated learning.

Personal stories and experiences are one of the keystones of SNSs. ‘Patients are experts in the experience with their symptoms, and physicians are dependent on patients for this expertise’ (Bensing et al., 2000). A major feature of the content of communications on health SNSs is users' personal experiences with treatments and other issues related with specific health problems or symptoms. Participants in health SNSs have expertise based upon their personal experiences giving them credibility that may exceed that of health professionals.

Patients are also experts in their experience with doctors. Traditionally, consumers rely on WOM from family and friends and physician referrals to choose doctors (Tu and Lauer, 2008). Now, eWOM on health SNSs has become another source of information about doctors. Bensing et al.(2000) proposed two types of interactions between patients and doctors: ‘cure’ systems, which are task-focused, and ‘care’ systems, which are socioemotion-focused(as cited in Ong et al., 1995: 906). Because patients often have negative emotions, such as anxiety and fear (Marks et al., 2005), they need ‘care’ from doctors. ‘Cure’ systems can be evaluated by objective measures, such as treatment effectiveness, whereas ‘care’ systems can only be evaluated by patients' self-report of their experiences with doctors. We argue that based on personal experience with doctors, consumers are likely to discuss their doctors in terms of the ‘care’ systems they provided. Health SNSs thus can be effective channels for WOM about the care environment participants will find with individual doctors.

Further, as SNSs are online communities for people who share interests or activities; users of health SNSs may have similar backgrounds (e.g. demographics and health conditions) and thus similar personal experience. Social comparison theory (SCT) suggests that similar others are ‘the most effective support-givers’ (Thoits, 1995: 67) because of ‘empathic understanding’ (Thoits, 1986: 420) and ‘collective wisdom’ (Davison et al., 2000: 206). It has been suggested that patients who have ‘embarrassing, socially stigmatizing, or disfiguring’ illnesses, such as obesity and HIV, are especially likely to seek support from similar others (Davison et al., 2000: 213). The support transmitted via health SNSs might be highly applicable for the situations that consumers need to address for specific diseases or symptoms.

Support seekers can learn from support providers, and support providers can also learn as they read posts by other support providers. Thus, both support seekers and providers can be learners. By accumulating health knowledge during their tenure on SNSs, both support seekers and providers may evolve into ‘information elites’, the ‘agents with the most valuable knowledge’ (van Alstyne and Brynjolfsson, 1995).

Thus, our second proposition is discussed as follows:

eWOM on health SNSs facilitates tailored health communication by providing messages with personal experience with specific health problems or symptoms that participants need for situated learning.

METHODOLOGY

To examine our propositions, we conducted a single case study employing netnography, or internet-based ethnography, as the research method. Hine (2000) first called her study of the culture of online communities ‘virtual ethnography’. Subsequently, researchers have developed this methodology, giving it different names, such as ‘digital ethnography’ (Murthy, 2008). Kozinets (2002) was the first to propose procedural guidelines for the study of the interaction of users on publicly available online forums calling the method ‘netnography’. Netnography has become a widely accepted methodology in the field of consumer and marketing research. Researchers have used netnography to explore a wide range of topics, such as consumers' responses to brand advertising for video games (Nelson et al., 2004) and innovative learning in online communities (Hemetsberger and Reinhardt, 2006). Using netnography, our research proceeded in three steps: (i) preliminary analysis from which we developed a coding plan, (ii) downloading and coding online data including complete threads and related information, such as date and time of posting, and (iii) analyzing the data to explore the major features of eWOM on health SNSs.

Coding scheme

We developed coding schemes based on past studies of online communities (Maloney-Krichmar and Preece, 2005; Donelle and Hoffman-Goetz, 2008; Ginossar, 2008) as well as our preliminary analysis of 15 randomly selected threaded discussions generated by the users of a health social networking site. As Table 1 shows, we coded posts into three categories: informational support-seeking/giving, emotional support-seeking/giving, and a combination of informational and emotional support-seeking/giving posts. We defined informational support as the behavior of providing advice, guidance or information and emotional support as the behavior of showing respect, love, caring, and concern (Bloom et al., 2001). Table 2 presents the basic features of threads and posts along four dimensions including length of discussion threads, length of posts, timeliness of responses, and identities of posters. We coded themes of threads into three major categories including medical treatments, nonmedical treatments, and general discussions; further, we identified subcategories under each theme category (Table 3).

Table 1. Coding scheme for posts
CategorySubcategoryDescriptionExample
Types of help seeking1-Informational help seekingAsking for information (e.g. suggestions or comments)‘Please tell me if you have tried Alli and how it worked for you’
 2-Emotional help seekingExpressing negative emotions (e.g. embarrassment) to seek help‘I am so huge and I just wish I could die, even though I have a loving family…’
 3-Informational and emotional help seekingAsking for information and expressing negative emotions‘I'm terrified I will die of obesity related complications. I have two very small children, ages 5 years and 6 months and can't comprehend leaving them any time in the next 30 or 40 years and knowing the pain of losing my own mother as a teenager, I want to spare them but I'm just sitting doing nothing. Does anyone feel this same way…? That they will die if they can't motivate themselves?’
Types of help giving1-Informational help givingOffering information (e.g. suggestions)‘Not sure about Xenical but I do take Phentermine. I have been on it a little over a month and have lost like 18 pounds so far’
 2-Emotional help givingExpressing positive emotions (e.g. hope) to provide help‘I'm almost 500 pounds and I understand the wanting to die, everyone feels that way sometimes, but its not really how you feel. After three years of depression and all kinds of not nice stuff. I'm moving forward slowly trying to get my life back together and I really want to live. Be strong, there is people all around you that wants to help.’
 3-Informatoinal and emotional help givingOffering information and expressing positive emotions‘I lost 20 pounds by exercise within 4 months. I'm proud of myself. I think you can do the same as me. Be strong! Never give up!’
Types of responses by help seekers1-Informational responsesAsking for further information or indicating thoughts or next actions based on help providers' responses‘I'll talk with my doctors about your suggestions and then make a decision.’
 2-Emotional responsesExpressing certain emotions (e.g. gratification)‘Thank you for your suggestions.’
Table 2. Coding scheme for basic features of threads and posts
CategorySubcategoryDescription
Length of threads1-ShortThreads with no more than 5 posts
 2-MediumThreads with 6 to 15 posts
 3-LongThreads with more than 15 posts
Length of posts1-ShortPosts with no more than five lines of messages
 2-MediumPosts with 6 to 15 lines of messages
 3-LongPosts with more than 15 lines of messages
Timeliness of responses1-EarlyResponding to help seeking posts within 6 hours
 2-LateResponding to help seeking posts between 7 to 12 hours
 3-Very lateResponding o help seeking posts after 12 hours
Identities of posters0-Not surePosters did not indicate their identities.
 1-Participants self-identified as overweight or obesePosters indicated that they were suffering overweight or obesity.
 2- Families of overweight or obese individualsPosters indicated that their families (e.g. husbands/wives, sons/daughters) were suffering overweight or obesity.
 3- Friends of overweight or obese individualsPosters indicated that their friends were suffering overweight or obesity.
 4- Healthcare professionalsPosters indicated that they were healthcare professionals(e.g. physicians and therapists)
Table 3. Coding scheme for themes of threads
CategorySubcategoryDescriptionExample
Medical treatmentM1-SugeriesSurgeries for weight loss (e.g. gastric bypass surgeries)‘Looking for information from people who have actually had stomach banding/stapling/whatever.’
M2-DrugsDrugs for weight loss (e.g. Alli)Xenical ..does it work?’
Non-medical treatmentN1-Self-supportTreating overweight or obesity by oneself (e.g. exercises and writing)‘This may sound a bit of a stupid suggestion but is anyone interested in a penpal for support with dieting, talking about fears and problems, a bit of motivation on paper? Let me know’.
N2-Commercial weight-loss programsTreating overweight or obesity by commercial programs (e.g. Weight Watchers, Jenny Craig)‘has anyone here lost & maintained their weight loss with weight watchers’
General discussionG1- Weight lossDiscussions about the motivations and effective ways to lose weight‘I want to lose weight because I don't want to have people say to me “you have such a pretty face”. One thing that a fat woman doesn't want to hear is that her face is PRETTY!!!!!! Or how about “ you're beautiful on the inside”!!!! You know, I don't care that I'm Miss Universe on the inside, I'm quite happy to fly in on my broom, cause my mischief and then fly home again - as long as I'm looking sensational while doing it’. :)
G2-Health and Medical problemsDiscussions about health problems or diseases caused by overweight or obesity‘Many diseases -- diabetes, heart disease, hypertension, etc. -- blame obesity for their problems. But what if the American Medical Association were to declare that obesity itself is a disease? Most people are balk when the idea is posited that obesity is a disease. We simply don't want to view obesity as a disease. Why?’
G3- Social anxietyDiscussions about overweight or obese individuals' anxiety (e.g. depression, worry) about social situations/ interactions with others.‘I was wondering if any of you worried about what other people thought about you? that is one of my biggest things. i let what other people think bother me. if this happens to you how do you deal with it?’
G4- ParentingDiscussions about how parents help their overweight or obese children‘I have come here in search of answers. I have 3 daughters of 11,9 and 6. My 6 year old is obese according to our family doctor. She eats the same as everyone else she just doesn't like to move around as much. I try to encourage sports but it is a no go. She simply refuses. I cannot do anything. What can i do?’
G5- DoctorsDiscussions about attitudes and behaviors of doctors‘we live in a country where doctors don't even assist patients in need, such as obese patients. Reminds me that we have to be in control of our own future!’
G6- ProductsDiscussions about the demand for the products tailored to the needs of overweight or obese individuals‘It's terribly depression to go into store after store and not find anything to fit right and compliment your body. I though maybe it wouldn't be so bad to find a 16, 18 sizes but I was wrong. It is and nothing is more depression than to look at your naked body or clothed body in a store mirror’.
G7- Public policyDiscussions about the support by public policy and programs‘One of the things I'm sure we've all noticed is that eating healthy is a lot more expensive than eating poorly. The government says that we have an obesity epidemic. DUH!!! Maybe if prices for good, healthy food weren't so high, it would be easier for everybody to get their weight under control. The govt. is willing to subsidize a whole lot of other industries. How about lobbying Congress to subsidize the healthy food industry?’

Data collection

Dailystregth.org, founded by three entrepreneurs in the US in 2006, is a well-recognized health social networking site with high traffic: around 280,000 average daily visitors by April of 2011 (Compete.com, 2011). This site contains more than 500 online communities that deal with different medical conditions and life challenges. Obesity is one of the major health concerns facing US consumers: more than 20% of adults are obese, and 30% of adults are overweight (Pi-Sunyer, 2002). Thus, we selected the obesity group as the object of this research. At the time the data for this study were collected, the group had 520 members.

In assembling the data for this study, a lot of information was available on the Dailystrength website. The web page of the obesity support group displayed the titles of all the discussion threads ordered by date. Clicking on the title of each thread showed the thread page and the data used for this study including the complete text of each post, the total number of posts in each thread, the posting time of each post, and the name of individual posters. A hyperlink from each poster's name to each poster's profile page was also available. From the profile page, we collected data including the poster's gender, age, and membership time. In addition, we used Google Ad Planner (a tool for advertisers to obtain information about sites) to collect data about average education level and annual household income of the site members.

From our preliminary analysis, we found that posts in one thread are usually generated within one month of initiation of the thread. Thus, we downloaded all posts in one thread generated within one month following the date the thread was initiated. We collected data in reverse order from the most recent threads generated one month before our study started to the oldest without considering seasonality. Posts included in our dataset span the years from 2006 to 2010. We downloaded a total of 90 threads with 882 postings. Each thread contained at least three posts, so that we could conduct meaningful analysis of each thread. Data collection continued until no new insights could be developed from the data.

Data analysis

The unit of analysis used in previous work on social networking communication has been one message generated by one individual. However, because netnography is a way to understand interaction among online community members ( Kozinets, 2002: 64), the primary unit of analysis employed in this study is one discussion thread containing one initial post by a support seeker and several subsequent responding posts by support providers.

Using SPSS, we performed quantitative analysis of coded data and the data from posters' profiles. An individual poster may post more than one post in the same or different threads. Thus, for the analysis of data categories about posts (types of support giving/seeking posts, length of threads and posts, timeliness of responses, and themes of posts), each post is counted; and for the analysis of data categories about posters (identities, age, and membership time of posters), only unique posters (i.e. having the same user ID) are counted. We first computed descriptive statistics for each data category. We then conducted correlation analysis (two-way Pearson's chi-square analysis) between the types of initial posts and their responding posts (informational, emotional, or the combination of informational and emotional types). In addition, we conducted qualitative analysis of the text message data to discover themes.

FINDINGS

We proposed that eWOM on health SNSs facilitates tailored health communication. We present the findings about the characteristics and contents of threads containing support seeking and giving posts illustrating the ways in which social support was tailored to the coping and learning strategies of support seekers and providers.

Characteristics of support seeking, giving, and follow-up

Support seeking

Our dataset included support-seeking posts as follows: 25 informational posts, 5 emotional posts, and 60 posts that were a combination of informational and emotional. The combination type of support seeking was the most dominant, consistent with previous studies that suggest that when dealing with difficult health situations, most support seekers needed both informational and emotional support (e.g. Taylor, 2003).

We found that the type of support-seeking post influences the length of threads (i.e. the number of responding posts). Threads initiated by informational support-seeking posts led to the highest percentage of short-length threads (one to five posts; 27%) but the lowest percentage of long threads (more than fifteen posts; 0%). Threads initiated by emotional support-seeking posts led to the highest percentage of long threads (20%). Threads initiated by the combination type posts led to a high percentage of long (20%) and medium-length threads (6 to 15 posts; 65%) but with the lowest percentage of short-length threads (15%). Different types of support-seeking posts appear to elicit different requirements for their responding posts. Informational support-seeking posts display clear questions simply requiring answers to the questions. However, emotional support-seeking posts reveal feelings, such as embarrassment about weight and do not specify the type of help needed from responders. As a result, support providers contribute a variety of responses to this type of post. The combination type of support-seeking posts allows responders more flexibility to develop their ideas, resulting in a high percentage of long and medium threads.

Support giving

Support providers offered support that matched the request and featured precision, timeliness, credibility, and depth. Correlation analysis showed that the types of support-giving posts were significantly associated with the types of support-seeking posts, χ2 (4, N = 882) = 131.70, and p < .01. Among 234 posts responding to informational support seeking, the highest percentage (56%) were informational support giving posts; among 52 posts responding to emotional support seeking, the highest percentage (59%) were emotional support giving posts; and among 596 posts responding to the combination type of support-seeking posts, the highest percentage (55%) were combination support giving posts. These findings show that the social support from support providers matches the requests from support seekers.

Second, our data show that among support-giving posts, more than 60% were short (posts with one to five lines of message), and morethan 98% were timely (posted within 12 hours of the initial posts). Short posts appear to serve one major function, which is to solve the problems support seekers face in a timely manner. This single function leads to posts with highly concentrated and brief information. Because information exchange is the major function of eWOM, the domination of short messages by which support seekers and providers exchange information represents the efficiency of eWOM on health SNSs.

Third, we found that support providers responded with messages with credible cues. Support providers were very clear about what they knew and what they did not know with regard to the questions posted by support seekers, thus assisting support seekers in assessing the credibility of messages they received. For example, one support seeker asked about Alli, a weight-loss drug, and one support provider first advised the support seeker to avoid high-fat food instead of taking any medication (what he knew) and then expressed his concern about the treatment effects of the drug (what he did not know).

I thought of giving it a try but I know it's just a means to discourage one's overconsumption of high-fat food -- I think I can do that myself without the aid of (an expensive) medication. Plus, I'm quite concerned about its “treatment effects” -- i.e. oily diarrhea and pant soiling. I'd love to hear from anyone who has success with it though.’

Finally, support providers offered extensive advice that went beyond the question in the initial request. For example, when support providers answered questions about the effect of weight loss surgeries, in addition to surgeries, they also talked about how they changed their life style (e.g. eating habits and exercise routines) after they had surgeries. One support seeker's question can evoke others to identify more questions about the topic. Thus, support providers may extend the original question by answering questions and then asking further questions about the topic under discussion. For example, one support seeker asked about lap band surgery. One support provider replied, ‘I would like to know more, too....is it covered by health insurance plans?’

Follow-up

We also observed that support seekers gave feedback to support providers by posting follow-up posts. Our data show that over half of initial support-seeking posts had follow-up posts in the same thread discussion. Sixty-five percent of follow-up posts showed support seekers' gratitude to support providers. The remaining follow-up posts showed support seekers' further thoughts as illustrated by this example:

Initial post

Hi, My doctor recommended for me to go on Alli… Does anyone out there know anything about this product? …Tell me if you have tried it and how it worked for you…’

Responding post

I use it and yes it has those side effects but I think it helps me a lot. Give it a shot for a month.’

Follow-up post

Thank you so much everyone for your comments. It is very helpful for me. My doctor said to take my vitamins, and b3, and fish oil at night before bed so they could go into my system before my next dose of Alli. Any thoughts on this?…’

This thread demonstrates that with the help of the support provider, the support seeker engaged in problem-focused coping. In the initial post, the support seeker identified a problem, that is, how Alli could work for him. In the responding post, he received information about his problem, and the follow-up post reveals that the information was useful. This thread also shows that the support seeker combined the support received with information from another source (doctor) helping to build confidence in taking Alli.

Research has shown that overweight or obese people may have low self-esteem (Straus, 2000; Gluck et al., 2001; Wadden and Stuckard, 2004; Franklin et al., 2006.). In our data, many participants engaged in discussions about problems caused by low or fluctuating self-esteem. The following dialogue between one support seeker (initial and follow-up post) and one support provider (responding post) about the problem of body image disturbance caused by low self-esteem exemplifies an emotion-focused coping strategy:

Initial post

I am so huge and I just wish I could die, even though I have a loving family…’

Responding post

Don't give up honey. You can do this. We all have struggled with weight and other issues too. It is not going to always be easy, but you can accomplish anything you set your mind to…’

Follow-up post

Thanks so much for your encouragement and support. That is exactly what I needed to hear!’

In the initial post, the support seeker reveals her depression over her body image. In the responding post, the support provider shows sympathy and encouragement. The follow-up post demonstrates that the support seeker felt that the support was ‘exactly’ what she needed to relieve stress.

To summarize, our findings demonstrate that interactions on health SNSs facilitate tailored health communication by providing the types of support (informational and emotional) that help support seekers actively cope with heath-related problems and strains (Proposition 1).

Message content

Similar personal background

Our data show that support seekers and providers share similar life experiences. First, support seekers and providers in this community shared similar health conditions. Our data reveal that 93% of participants self-identified as overweight or obese. This commonality may be one of the primary attractions of joining an on-line support group. Second, support seekers and providers in the obesity group shared similar demographic backgrounds (e.g. age, gender, education, and income). Consistent with previous research, we found that most support seekers and providers were middle-age females (average age = 39.7 years). Because females in this segment are also likely to be the primary caregivers for their families, their engagement in communication on health SNSs will benefit not only their own health but also that of their families. Our data also show that 79% of support seekers and providers had college degrees and 69% had annual household incomes of $50,000+. Notably, individuals with low education and income as well as young males are not currently participating in the online health network world.

Message themes

Support seekers and providers posted messages with a wide range of themes relevant to their personal experiences. Within our dataset, there were a total of 11 threads on medical treatments with a focus on surgeries (e.g. gastric bypass surgeries) and drugs (e.g. Alli), 19 threads on nonmedical treatments with a focus on self-support (e.g. exercise, penpal and healthy eating) and commercial weight loss programs (e.g. Slim-Fast and WeightWatchers), and 60 threads on a diversity of issues including weight loss (e.g. the motivation for weight loss), health and medical problems (e.g. diseases caused by overweight or obesity), self-esteem (e.g. depression caused by poor self-esteem), parenting (e.g. childhood obesity), doctors (e.g. relationships between doctors and obese patients), products (e.g. comfortable clothes and airlines seat), and public policy (e.g. Medicaid) (see Figure 1). The popularity of nonmedical treatments over medical treatments that we observed is coroborated by the data provided by the obesity group website. According to Dailystrength.org (2010), the ‘top ten treatments’ discussed most by the community users included only two medical treatments (gastric bypass surgery and Alli), while the other eight were nonmedical treatments (e.g. physical exercise and WeightWatchers).

Figure 1.

Number of threads with different topics.

In using nonmedical treatments, individuals primarily rely on their own efforts to achieve their desired health goals. The popularity of self-support shows that support seekers were anxious to learn effective weight loss methods from the personal experiences of support providers. With these personal examples as guidance, support seekers were in a better position to take actions to lose weight. We argue that one big attraction of communication on health SNSs is the fact that, by posting questions, support seekers can get personalized help (especially ideas for self-support) that they could not get from doctors.

These data also reveal that there were more threads on social issues than threads on treatments, suggesting that participants seek solutions to social problems. It appears that through participation in SNSs users can learn from other consumers with similar personal experience. In the following, we present major themes demonstrating the ways in which support seekers learn from support providers to deal with problems in their everyday lives.

Situated learning

Treatments

In the obesity support group, support seekers often expressed concerns about commercial weight-loss programs (e.g. Weight Watcher). We identified three major concerns that support seekers had about packaged meals commonly used in commercial programs, including high prices, short-duration effectiveness, and lack of diversity. As a result, support providers offered advice about substitutes for diet meals. For instance, one support seeker asked about the effectiveness of Nutrisystem, a commercial dieting product. One support provider replied that he was using Hoodoba that was about 1/3 of the price of Nutrisystem. Other support providers suggested that they enjoyed cooking their own healthy meals instead of buying packaged meals.

Another support seeker asked about how to fit exercise into a busy schedule, a problem that many overweight or obese people face. Support providers offered useful advice from their own experience as illustrated in the following thread:

Initial post

I'm trying to get more exercise into my daily routine. I leave the house for work at 8:00 in the morning and I don't usually get home until late at night… Does anybody have any ideas how to fit any kind of exercise into a busy schedule?…’

Responding post

It's hard for people that work a full time job. Have you thought about getting a video. I hate getting up early, but I would rather do my exercise in the morning or I have trouble falling asleep …’

Doctors

In our data, eWOM about doctors primarily focused on two topics, when to see a doctor and how to choose a doctor. When a person experiences symptoms, she/he might not be very clear about whether to wait and see how the symptoms develop, engage in self-treatment, or go to see a doctor. Deciding when to see a doctor can be difficult for some people. Some critics argue that relying upon the advice available on SNSs may lead individuals to delay seeing a doctor. However, our data suggest that when support seekers asked about whether they should see a doctor when they developed illusive health problems, such as headache, back pain, and depression, most support providers suggested seeing a doctor. This suggests that eWOM on health SNSs may facilitate support seekers seeing a doctor in a timely manner. Whether appropriate utilization of doctors is facilitated by SNSs is an empirical question.

Care systems may be especially important for overweight and obese persons because they frequently not only have health problems but also psychological problems. People may tend to show biases toward overweight or obese persons, and healthcare professionals are not exempted (Puhl and Brownell, 2001; Schwartz et al., 2003). With the help of eWOM recommendations, individuals may increase their visits to doctors who demonstrate caring. We found that some support seekers felt uncomfortable seeing a doctor and, as a result, tried to avoid seeking health care from doctors. One help seeker initiated a thread looking for a ‘fat-friendly doctor’ in her area. She said, ‘I have been avoiding the doctor for over 12 years and I know it is very unhealthy. I am overweight and I am scared of being judged at the doctor's office.’ One support provider, although unfamiliar with the area where the support seeker lived, offered practical advice about how to get information about the ‘care’ system of doctors by checking with ‘the State Board of Disciplinary office to see if there have been any complaints filed against a doctor’, and talking to ‘people who have visited the doctors as they can tell you what their personality is like’.

Thoughts and feelings

Support seekers and providers also exchanged information about their thoughts and feelings about overweight or obesity. For example, some support seekers did not want to be regarded by others as being overweight or obese and were very sensitive to certain words as one support seeker noted:

Anyone ever wonder why the word obesity? I personally hate that word. I would rather hear chubby, pleasantly plump, overweight, even fat. Anyway I just wondered isn't that the most offensive description? I think so and was wondering what you thought’.

One support provider used his own experience to show how to deal with the term ‘obesity’:

I don't fall into the “morbid” range, but the very term “obesity” has been a hard one for me to admit to myself. I've always thought of myself as “a little heavy”--but no, I'm obese. Joining this group and “saying so” has been a major step for me in acknowledging my issues and taking the steps to do something about them?’

Continuing learning

Our dataset includes a total of 882 posts made by 410 unique users; on the average, one user made 2.15 posts. This finding illustrates the opportunity for continued learning among users who post frequently. We identified the top 10 most active users based on the number of posts they made. We found that the average time one of these users had been a member of this obesity group was nine months, suggesting that knowledge was gradually accumulated over time. As an example, one active user became a member of this group in December of 2006. Our data showed that, in 2007, this user made a support-seeking post asking for suggestions about how to keep on her diet plan, and in 2008, she made two support giving posts about the effectiveness of Alli and motivation to keeping dieting.

These findings about learning process provide evidence for our proposition 2. Users in health SNSs with shared personal experiences seek and/or provide support for problems that they face in everyday situations. By continuing to learn from other users, they can become ‘information elite’, as suggested by the transformation of their roles from support seekers to support providers.

DISCUSSION AND IMPLICATIONS

Our analysis suggests that the basic features of eWOM on health SNSs translate into important benefits for consumers facilitating health communication tailored to their coping and learning strategies. These benefits can accrue not only for active users who read and post messages but also for passive users who just read but do not post any messages. Within threaded discussions, this one-to-many communication can spread a health message broadly among a large population. Because health SNSs run 24/7, health information can spread in a timely manner through eWOM on health SNSs. In addition, as most health SNSs including Dailystrength.org provide a search function, consumers can easily find the discussion topics they are interested in by simply typing keywords into search engines.

As interest and participation in SNSs has grown, such sites provide an important additional resource for consumers. In healthcare specifically, these sites can provide support groups for consumers dealing with health problems and concerns. These sites likely will not replace interactions with healthcare professionals, family, friends, or even live support groups. However, they may provide the ‘right support from the right person at the right time’ facilitating emotional coping.

It is important that policymakers and marketers assess how they can better serve consumers using eWOM on health SNSs. The most direct way to use health SNSs is to learn from the contents of threaded discussions. As these discussions are open to the public, policymakers and marketers can have easy access to rich first-hand information about consumers' health and medical opinions and problems. Further, by forming their own groups on SNSs, policymakers and marketers can actively communicate with consumers without interrupting the existing communities. From our observation of the health-related activities on popular general SNSs (e.g. Facebook) and health SNSs (e.g. Dailystrength), the major online health-related activity of policymakers and marketers is to release information on their group pages. The US Centers for Disease Control (CDC) has formed groups on SNSs, such as Twitter, Facebook, and Dailystrength. However, the only activity that CDC engages in on their group pages is to release health information. An important opportunity exists for such organizations to engage in more interaction on SNSs.

Considering the value of the information and support available on health SNSs, policymakers and marketers should also aim to improve the quality and quantity of eWOM messages on health SNSs. One effective way is to advise consumers on how to find health SNSs matching their personal needs and how to read and post messages on their selected health SNSs.

FUTURE STUDY

The present research focused on the basic features of eWOM on health SNSs, especially on how communication among users is tailored to users' personal needs. This research has two limitations that suggest directions for future research. First, more research is needed to investigate consumer differences with regard to their social ties with online communities and interests in online activities. Consistent with Kozinet's study, we found that consumers in our study differed in the extent and manner in which they participated. Additional research is needed to investigate the characteristics of different types of participants. Studying participants' online public profiles that include personal information, such as participants' online activities, journals, personal goals, and friends, may provide insights about important differences across participant groups.

A second limitation of this research relates to the influence of eWOM on consumer decision making processes. eWOM on health SNSs has become an important health information source for consumers. This new source of information may provide new and/or different information which then may be factored into a decision. Future research is needed to investigate how information obtained through eWOM on health SNSs impacts decision-making, particularly how consumers combine information received from different sources to make health decisions.

Biographies

  • Bo Liang is a PhD student in the Department of Marketing in the David Eccles School of Business at the University of Utah. Bo holds a Master's degree in Business Administration (MBA) from Clemson University. Her research interests include health communication and social networking.

  • Debra L. Scammon, PhD, is the Emma Eccles Jones Professor of Marketing in the David Eccles School of Business at the University of Utah. Her research interests include patients' access to and use of information in decisions about their health and healthcare. Current projects include evaluation of strategies for primary care clinics to collect patient-generated information and to provide patient self-management tools through Internet portals.

Ancillary