Recruiting online for a randomised controlled trial of a web-based domestic violence intervention (I-DECIDE): Who do you get and how does it work?
Domestic violence (DV) is common and causes significant physical and psychological harm to women and their children. Health care interventions for women experiencing DV have predominantly focused on screening and referral. Yet many women may not identify as a ‘victim’ of DV, and may be reluctant to access formal services. There are well known barriers to disclosure including: fear, shame, time constraints, and concerns about confidentiality. Online interventions have the potential to overcome these barriers and provide an effective means to improve safety and wellbeing for women experiencing DV, however, it is critical that these interventions be rigorously evaluated. This presentation will report on the experience of recruiting online for the development and evaluation of an online healthy relationship tool and safety decision aid for Australian women experiencing DV (I-DECIDE). It will outline some of the challenges involved in online recruitment with this population, as well as reporting on recruitment outcomes and what strategies were successful.
Development of the I-DECIDE website was informed by four focus groups conducted with women who had experienced DV and pilot testing of the website with a small user panel. Women were recruited via student portals and paid advertisements on social media. Once the website was developed, it was evaluated through a randomised controlled trial with n=426 women aged 16-50 who had experienced fear of a partner or abusive behaviours in the previous 6 months. Baseline data collection was completed in August 2015.
Extensive online recruitment was undertaken through advertisements (both paid and unpaid) on social media, and community partnerships. Careful wording and strategies were needed to balance women’s safety with the need to recruit trial participants. Community groups did not recruit that many women, but Facebook by key organisations and celebrities worked well. Validation of women through the electoral role resulted in only a small number of exclusions. Baseline data showed a diversity of women (11% indigenous, 4% same-sex women), two thirds were highly educated and employed, however very few were from cultural and linguistic diverse populations.
Lessons for other e-health interventions are that women can be recruited safely to an online safety decision aid. However, similar to other e-health interventions recruitment attracted an educated population with some diversity of the population.