On November 19, 2008 we searched Facebook using the platform’s built-in search engine and the keyword breast cancer (Figure 1). We restricted our analysis to Facebook groups that were related to breast cancer, operated in English, and were publicly available to anyone with a Facebook account to view and join. Pages for individual members, organizations, events, and applications were excluded.

Two of us (JLB and MCJM) independently reviewed the resulting set of eligible groups and extracted information on the following: (1) general characteristics (eg, group name, purpose, creator, and URL), and (2) membership and user-generated content (eg, number of members, discussion posts, wall posts, photos, and videos).

We determined the purpose of each group based on a content analysis of, and in order of priority (if available), the title of the group, the description of the group, the information in the Recent News section, the discussion posts, and the wall posts. (The content analysis of the discussion and wall posts was restricted to those displayed on the main page of the group.)

We began by analyzing the content of the first 100 groups to develop a coding and classification scheme that could be applied to the entire set. This initial step led to the identification of four main types of breast cancer groups:

After independently classifying the general purpose of the groups using the above coding scheme, we resolved any differences. Next we generated a second-tier coding scheme to subclassify and more specifically describe the purpose of each group.

We also developed and independently applied a coding scheme to classify the approximate age and geographic location of the creators of the support groups. We restricted our analysis of the creators to the support groups, because we were primarily interested in the role of Facebook groups as a source of supportive care.

Lastly, we calculated descriptive statistics using SPSS version 17 (IBM Corporation, Somers, NY, USA) to summarize and compare the size (in terms of number of members) and amount of user-generated contributions of each type of group (in terms of wall posts). Most data were expressed as medians with interquartile ranges (IQRs) because the number of group members and user-generated content varied considerably and did not follow a normal distribution. We used chi-square tests to compare categorical data across groups.

This study was a component of a larger research study for which ethical approval was obtained. However, it should be noted that this study met the exclusion criteria of the (Canadian) Tri-Council Policy Statement as to what studies require review by an institutional research ethics board, because all information was publicly available.

As shown in Table 1, the majority of groups (513/620, 82.7%) were created for fundraising or awareness purposes. In total, 44.7% (277/620) were created to raise funds for breast cancer, 38.1% (236/620) to raise awareness about breast cancer and related events, 10% (61/620) to promote an external website raising funds or awareness for breast cancer through the sale of products or services, and 7% (46/620) to generate support for people affected by breast cancer. A minority of groups (9%) were classified as having an additional purpose, 34% (19/55) of which related to fundraising or support, 27% (15/55) to raising awareness, and 4% (2/55) to supporting an external website. As shown in Table 2, the three most common types of breast cancer groups on Facebook, which comprised 69% of the total sample, were (1) groups created to raise funds for a fundraising walk associated with a charitable organization in the United States or Canada (239/620, 38.5%), (2) groups raising awareness about a specific fundraising event (95/620, 15%), or (3) groups promoting the importance of breast cancer in general (94/620, 15%).

We identified a total of 1,090,397 Facebook users who were members of one or more of the 620 breast cancer groups. The awareness groups contained by far the most members (957,289, 87.8%), followed by the promote-a-site groups (64,861, 5.9%), fundraising groups (51,307, 4.7%), and support groups (16,940, 1.5%). The groups ranged in size from 1 to 772,815 members and had a median of 196.5 members (IQR 214.7). Most groups (612/620, 98.7%) contained 5000 or fewer members and 70.8% (439/620) contained 101 to 500 members. On average, the promote-a-site groups had the greatest median number of members (median 373.5, IQR 932), followed by the awareness groups (median 270, IQR 389), support groups (median 235.5, IQR 237), and fundraising groups (median 151, IQR 92) (Table 1).

A user can contribute content to a Facebook group in various ways, such as posting messages to the “wall,” news section, or discussion board, or uploading multimedia such as photos or videos. As Table 3 shows, the most frequently used communication feature was the wall. Although wall posts ranged in number from 0 to 8614, the groups contained a median of 5 wall posts (IQR 11). The majority of groups (532/620, 85.8%) had 25 wall posts or fewer. The support groups had the greatest median number of wall posts (median 16, IQR 38), followed by the awareness groups (median 6, IQR 19), promote-a-site groups (median 4, IQR 9), and fundraising groups (median 4, IQR 7). The difference in median number of wall posts across the groups was statistically significant (c² ₃= 52.0, P < .001).

Nearly half (32/65, 49%) of the support groups were created to generate support for anyone affected by breast cancer. A typical purpose statement for these types of groups was “For anyone who knows someone who has survived, is battling or has died of breast cancer. For congratulations, hope and [in] memoriam.” An additional 38% (25/65) of the support groups were established to obtain support for the creator of the group or a loved one affected by breast cancer and 12% (8/65) were created as a forum for information sharing among people participating in a fundraising walk (Table 2). Interestingly, a minority of the groups that were created “for anyone” affected by breast cancer (6/32, 19%) were initiated by individuals with an afflicted family member or friend, even though the explicit purpose of the group was not to gain support for the creator of the group or a loved one in particular. In the remaining 26 of these groups, the motivation of the group creator was not explicitly described. A small percentage of the support groups (5/65, 8%) were also serving in memoriam of a loved one who had died of breast cancer.

We also examined the creators of the support groups for anyone, oneself, or a loved one affected by breast cancer (excluding groups created as a support forum for people participating in a fundraising walk, because we were primarily interested in breast cancer-related support). All but one of the creators of the support groups (n = 57) restricted the visibility of their personal profile pages to members within their networks. However, in 47% (27/57) of the support groups the academic institution of the creator and their expected graduation date either was included on the group page itself or was available in the search result content, and in 86% (49/57) of the support groups the geographic location of the creator was also available. Of the groups with available information on the approximate age of the group creators, 56% (15/27) were college students, 37% (10/27) were high school students, and 7% (2/27) were recent college graduates. None of the support group creators appeared to be health care professionals or associated with a health care organization. Of the groups with available information on the geographic location of the support group creators, 57% (28/49) were located in the United States, 41% (20/49) in Canada, and 2% (1/49) in Australia.

The findings of this study are valuable because they provide information on the health-related use of the most widely popular social network site in existence. They indicate that Facebook groups are being used by a considerable number of people affected by breast cancer for fundraising and awareness purposes, and to a lesser extent supportive care. That being said, our findings should not be interpreted to imply that Facebook is rarely used for supportive care purposes, given that several ways to solicit or provide support on Facebook were not examined in this study, including private messages, wall posts on personal profile pages, and status updates. These findings do suggest that Facebook may play an important role in facilitating public engagement in health promotion and fundraising activities, particularly among youth.

This study has important limitations. First, we were unable to collect demographic information on 53% (30/57) of the support group creators due to their use of privacy settings. However, this finding suggests that users of Facebook not only are becoming aware of the public nature of their online activities, but also are activating the privacy measures offered. In fact, all but one of the support group creators in our sample restricted their personal Facebook profiles to their networks, whereas a study of Facebook users conducted in 2005 found that only 0.06% of college students restricted the visibility of their profiles to members within their networks [40]. Since then, significant changes made to the platform and user base of Facebook might in part explain the increased use of privacy settings by this sample, such as the launch of the NewsFeed feature, which provides updates on the activities of friends [41], the introduction of third-party-developed applications [42], and the expansion of registration to anyone.

Another related limitation was our reliance on user self-reported data (that were available on the group page itself or in the search result content) to infer the approximate age and geographic location of the support group creators. This information is possibly incorrect or fabricated. In addition, we could not determine the exact number of unique individuals affiliated with a particular type of breast cancer group on Facebook, given that a single user could be a member of multiple groups. Therefore, the total number of members affiliated with each type of breast cancer group could be inflated. At the same time, the total number of breast cancer groups identified in this study is likely only a portion of the total number of breast cancer groups on Facebook, given that we restricted our study to groups in English, while Facebook is available in more than 70 different language versions [4].

Lastly, we encountered numerous challenges while investigating the nature of breast cancer groups on Facebook that were primarily related to its limited functionality as a search tool. The search bar yields an imprecise yield (eg, “>500 groups”), the order of the search results is inconsistent and unclear, and the search is limited to the title of the group. Since the time we conducted our study the search tool has been enhanced but, to our knowledge, these specific issues have yet to be resolved. We contacted Facebook to notify them of these technical issues and obtained an encouraging response. Collaboration with platform owners would certainly facilitate future research in this area.

Further research is warranted to understand the implications of participating in health-related groups on Facebook. While other researchers have examined site activities that lead to higher levels of social capital [34,35], no known studies have examined the impact of participating in a health-related group on Facebook. It is also unknown whether general social network sites such as Facebook are as effective as disease-specific online communities in providing health-related information and support, and for whom. Given the importance of anonymity in facilitating disclosure in online breast cancer communities [13], research is warranted to examine breast cancer survivors’ perceptions of social network sites as a source of supportive care in comparison to other sources. Lastly, a better understanding is needed of the privacy implications of sharing personal health information on public social network sites, which has raised concern [25], leading some to advise against disclosing personal information on these sites [8].

Facebook groups have become a popular tool for awareness-raising, fundraising, and support-seeking related to breast cancer, attracting over one million users by the end of 2008. Given their popularity and reach, further research is warranted to explore the implications of social network sites as a health resource across various health conditions, cultures, ages, and socioeconomic groups.