Challenges of an online randomized controlled trial
Carrying out this online RCT poses various challenges, particularly in terms of recruitment, ethics, and data collection, including participant follow-up over an extended period .
First, to attract and retain the online participation of PLHIV interested in the study, the team sought the help of web design experts. This help led to the development of a brand image (concept of branding) by an artistic director, who created a logo that would carry the image of the study and be easily recognizable by our target group. This branding set the tone (in terms of color and visuals) for our website pages, which were created to be visually attractive to participants. More specifically, upon reaching the first page of the website, participants will listen to testimonials by a few PLHIV about their own experiences with medication intake and with browsing the web in search of health-related information. Aside from being an interesting and dynamic visual tool, testimonials have been used to allow potential participants to identify with the intervention, and to humanize and personify the web-based study. These testimonials also have the potential to strengthen the branding of the intervention. Then, in a short video clip, the research coordinator and a project spokesperson will present information on the study, including its aim and inclusion criteria, and explain what participation entails.
Various communication strategies will be used to promote the study among PLHIV. First, traditional methods of dissemination will be used, such as publishing advertisements or short articles in magazines aimed at PLHIV, and distributing information brochures and posters to the clinics and community organizations targeted. Second, an internet-based approach will also be used, consisting of posting a hyperlink to our study on the websites of organizations (medical clinics, association for PLHIV, community groups) frequently visited by PLHIV. In this regard, Birnbaum  suggested that such organizations should run advertisements in their electronic newsletters. A list of targeted internet media will be compiled in collaboration with key informants working with PLHIV in community organizations. There are other methods of online recruitment, such as reaching target groups through chatrooms, sending e-mails to pre-established lists of participants, or posting banners on popular websites [94, 95]; however, we will not use these latter strategies in our study. According to Noseck et al. , even if posting advertising banners is a popular way of promoting a study, it remains a costly means of recruitment and ‘is not the most effective at netting traffic’ (, p. 167).
One of the risks that remain difficult to control or prevent in a virtual environment is having the same person participate more than once in a given study [93, 95, 96]. Although the incidence of this happening has been shown to be small, at 3%, [93, 97], certain strategies can be used to detect or prevent it and reduce the risk of its occurrence, including warning participants to register only once, asking for an email address and a password at registration, and filtering data that seem to be identical[93, 95, 96].
The ethical aspects of online studies have been discussed by various authors [98–101]. The topic has received special attention because of the online research environment in itself, where there’s no direct interaction between researchers and participants [95, 96]. Among other things, consent to participate can be presented in the form of a statement to be read and checked off before clicking on a button labeled ‘I accept’. A button labeled ‘withdraw’ or ‘quit the study’  will also be accessible to allow participants the freedom to drop out of the study at any time. As proposed by Hewson  and Nosek et al. , a ‘debriefing page’ and contact information for a resource person in the event of questions or problems will also be provided. The information must be presented in such a way as to properly inform participants but without overwhelming them with data, which could cause difficulties and discourage subjects from participating.
The security and confidentiality of electronic data in online studies are serious concerns that necessitate the implementation of stringent measures. The data collected through the online questionnaires will be stored on a secure server located at the research center where the research work is carried out. Only system administrators will have access to the data on the server. To protect the identity of participants, only an email address and a pseudonym will be requested from participants, as this information is necessary to ensure follow-up and disseminate results, as pointed out by Michalak and Szabo . Using a firewall, encrypting the data, protecting passwords, and separating the sociodemographic data from the experimental data (linking them through a code) reduces the risk of an unauthorized third party accessing confidential content [95, 101]. A database has been planned to support the subsequent direct transfer of the data to analysis software (the SPSS package; SPSS Inc., Chicago, IL, USA) in order to facilitate data analysis.
As emphasized by Bull et al. , online longitudinal evaluations entail following up participants and completing various measures. According to Bull et al. and Bennett and Glasgow , online studies have a lower retention rate than that of traditional studies, and attrition rates of 40% are not uncommon. However, Schubart et al. and Bull et al.  highlighted a number of elements that foster a high retention rate in online studies, including the availability of a health professional to provide feedback, the dynamic qualities of the study website, and the inclusion of a clinician coach, such as a nurse, in the intervention. All of these will be considered in this study.
For the purpose of encouraging participants to continue with the study, reminder e-mails (maximum of three to four) will be sent out at 7-day intervals prior to measurement. Bull and colleagues  noted that e-mails alone were not sufficient to communicate with participants, and that a variety of communication modes should be used. They suggested that participants should be asked to provide other contact details such as home and mobile telephone numbers, and a home address. According to these authors, a balance must be struck between guaranteeing confidentiality and ensuring follow-up. Because the technology we plan to use is complex and can present many problems and bugs, we plan to assign a resource person (technical support) to help users in this regard. According to Paul and colleagues , this constitutes a key source of support if the research budget allows it.
A research project that unfolds entirely online necessitates that all parties share an understanding of the specific terminology used by the various experts involved and of the role played by the professionals in each of the project’s teams (computer programming, multimedia, clinical). It is necessary to ‘think virtual’, that is, to make sure that procedures applied in the real world are operational online. Transposing research procedures to an entirely virtual research intervention entails anticipating the entire sequence of participation and follow-up of the target persons. Making this shift is not easy. Indeed, the absence of a common language between research, clinical, and web media teams, as well as a mutual misunderstanding of the needs of each group (researchers versus tech support teams, for instance), calls for a great deal of clarification and numerous meetings to make the technical operations required by the project possible. This collaboration between researchers from clinical disciplines (nursing and medicine), experts in behavioral and educational sciences, and information technology and media experts is crucial to the development of innovative solutions to supplying and delivering services to those who need them.
However, without the contribution of community partners and other clinical collaborators, such a study could not be carried out easily. Their support and expertise are invaluable at all levels, particularly in planning and developing the different stages of the online study (for example, using testimonials was our partners’ idea). Our partners will also be involved in promoting the study, recruiting participants, and ultimately transferring and implementing the virtual intervention.