HIV education and counselling using Facebook: a possible new approach

Almost one third of individuals infected with HIV do not enter health care until late in the course of their infection. ¹ This phenomenon, called late presentation, is characterized by individuals who discover their seropositivity when in need of highly active antiretroviral therapy or with an AIDS-defining illness. Late presentation is a major concern, since it is harmful to the infected person and to society as a whole, and it is very costly. ² Its prevalence ranges from 10% to 45% in Europe. ³ Raising awareness in the general population is crucial and the Internet may be a new way of developing educational and counselling strategies. Social networking sites, such as Facebook, Twitter and MySpace, are well-known Web 2.0 applications (i.e. web sites whose content is mostly driven by the users). Most online information services concerning STDs or HIV operate with a Web 1.0 approach (i.e. information provided by the publisher, with a unidirectional stream). Shifting to Web 2.0 social networking sites may increase the effectiveness of prevention messages.

We conducted a multi-centre, prospective communicative intervention involving seven physicians with expertise in infectious diseases (termed “researchers” in what follows) who created their profile on Facebook. It was carried out from December 2009 to May 2010. The researchers created a real profile, which included photographs, lists of interests, contact details and other personal information. They used Facebook like everyone does and had conversations about their personal lives, experiences, emotions and feelings. Chatter concerning STDs and HIV topics were collected as professional messages. The researchers spent at least four hours every week on updating their profiles. They aimed to post replies within one week.

We conducted a study in two phases: the first was based on a targeted operation model and was held on Facebook. The objective of this phase was to increase the debate about STDs/HIV. We chose the number of friends and the number of questions that were raised as measures of debate. The target audience was the whole population who use Facebook: neither limitations nor exclusion criteria were defined for the choice of friends, in an attempt to use the social network just as in everyday life. To stimulate the debate, researchers posted messages, links and videos related to STDs/HIV (i.e. video clips from the HIVisible Contest held in Cologne in 2009, or from movies and songs with an AIDS theme; references to newspapers or magazines articles; commentaries to the national campaign launched for the World AIDS day of 2009 or to the Pope's statements against condom use in Africa). The number of replies was counted by each researcher. Demographic features of friends were collected from their personal pages.

The evaluation was through outcome measures. The study was designed to quantify the performance from the user perspective and to measure only the rate of query submission. We are not aware of any previous similar work, so the parameters for measurement were extrapolated from other studies on non-web social networks. The effectiveness of recruitment was measured by the Network Index (the ratio between the number of friends and the number of researchers). ⁴ The number of questions about STDs/HIV was measured by the Performance Index, a parameter set by analogy with the Network Index, and defined as the ratio between number of messages and the number of researchers.

The second phase was based on an open operation model and was held on the official site of the Italian Society of Infectious and Tropical Diseases (SIMIT). The objective of this phase was to analyse the questions and the debates that would develop from the inputs posted by researchers. Topics that needed longer explanation or that were perceived as too personal, were directed from Facebook to specific forums on the SIMIT web site. Technical questions (e.g. concerning new drugs or drug toxicities) were directed to the Expert Online forum, where an Italian physician experienced in the matter would answer. Personal questions (e.g. about risky behaviours or condom use) were directed to a public forum, where the adoption of a nickname could guarantee anonymity. Details about how to log into these forums were explained on the personal pages of the researchers. The researchers were also the coordinators of these forums.

During the study period, the number of friends reached 625; demographic features are summarized in Table 1. However, the objective of stimulating debate was not realised. The baseline Network Index was 66 and it increased to a value of 89 at the end of the study. Even though the number of friends increased by 35%, their features reflected mainly those of the researchers, i.e. young women from northern Italy, graduates, who were frequently working in health care as doctors or nurses. Despite the large number of professional postings, professional messages were few and none was followed by a specific debate on the SIMIT forums. The Performance Index was 46 for the personal (private) messages, while it was only three for the professional messages. It was not possible to categorize these messages, because of their small number. Thus the second objective of our study was not realised either.

The disappointing outcome of the work is due to some limits inherent in Facebook and to the study design. First, Facebook communication is often a manipulation of the personal virtual image with projections of “Self”, i.e. users sometimes do not write what really happens to them, but manipulate their life experiences to seem nicer, funnier and more appealing. This could lead to unreliable discussions on matters where their image could appear fragile, as in the case of STDs/HIV. This aspect could generate rhetorical or anyway unproductive public discussions, especially in the HIV field, where individuals often deny their condition. Messages on Facebook are generally very short and fast, and not really adequate for such a complex matter.

Second, there was no indication that persons who became “friends” of the research staff subsequently participated in any HIV education. “Friending” tells us about possible pathways of engagement and not actual engagement. In addition, even though the Network Index showed a good increase over a short period of time, it was not possible to reach a larger audience. Strategies used by the researchers failed to increase user numbers. Another potential bias lies in Internet access and computer literacy.

Finally, the methods of evaluation represented a problem: a validated framework is required to draw conclusions about which networks and which applications (asynchronous consultations, information sharing between peers, colleagues and patients, social media) are successful. We chose to evaluate outcome measures and not process measures, but no standard approach has been defined.

Our study represents a cautionary tale of what not to do in such an intervention and from these concerns Facebook use might be re-thought in more efficient ways. A major question is the choice of whether to create a Facebook group on HIV/AIDS issues or not. There are other Facebook groups on HIV/AIDS issues, but they are not formed by physicians and the inability to verify the identity of the users poses problems of reliability. The study design did not include the creation of a Facebook group to provide a greater trustworthiness through the researchers' personal profiles. The HON Code of Conduct for medical web sites (HONcode) ⁵ could represent a useful tool to assess the reliability of the Internet, but it could be difficult to apply to Facebook.

Patients are increasingly using the Internet for information about medical conditions or treatments. ⁶ For physicians working in outpatient HIV clinics it is not uncommon to meet patients who look for information online, often finding confusing and contradictory answers, before discussing such matters with their doctors. Facebook could be an interesting tool to develop novel counselling strategies, but its use needs to be rethought with different approaches and investigated in detail.