This research priority setting process consisted of three phases and nine steps inspired by the priority setting guidelines developed by the Council on Health Research for Development (COHRED) and other globally used priority setting procedures (figure 1).18,19
Phase 1 Setting the scene and designing the process
Step 1
To assess the situation in which the priority setting takes place and clarify the need for a priority setting exercise at the intersection of healthcare and religion, we purposively sampled a core group of experts with extensive global work experience. In addition, data was used from document analyses and five focus group discussions with faith leaders and healthcare workers (N=170) in South Africa. We interviewed selected experts and used these preparatory interviews to establish a larger group of stakeholders, which was important for the priority setting.
Step 2
To specify the scope, map and engage potential key informants and users and help set the scene for the priority setting, we organised an expert meeting in Geneva with representatives from World Council of Churches, UNAIDS, IAS members and academia.
Step 3
Following the COHRED guidelines, we developed a tailored research priority setting approach, for which we combined the Delphi 20 method with elements from the CHNRI method.21 We defined our expected output as an inclusive global agenda, citing research priorities at the intersection of healthcare, HIV and religion. We aimed to bring together the expertise of scholars with in-depth knowledge of the existing research reservoir and the needs and experience-based knowledge of other stakeholders.
Phase 2 Constructing and conducting Research Priority Setting
Step 4
In this next phase, we identified key stakeholders (policymakers, healthcare providers, faith leaders, academics and HIV activists) from around the world, through purposive sampling and snowball sampling. Two researchers (MN, HS) interviewed the selected stakeholders about the need for research (N=53). Interviews lasted between 45 and 90 minutes, took place either in person (N=30) or via skype (N=23) and were audio recorded and transcribed verbatim.
Step 5
A preliminary list of research questions and themes was identified by four researchers (MN, MK, SLvE, HS) after analysing and coding all interviews. Interviews were analysed using MAXQDA and constant comparative method of analyses. 22
Step 6
The list of research questions and themes was sent out by email as part of a digital questionnaire to purposively sampled key informants representing high-burden countries. Themes and questions were ranked and potentially missing topics were identified (N=110). The process did not yield any new themes.
Step 7
Based on the questionnaire outcomes, a specified list of questions and themes was compiled by three researchers (MN, MK, SLvE). These questions and themes were presented and discussed at two expert working groups at the IAS conference in Paris (N=7) and a workshop organised by the University of Kwazulu-Natal, Collaborative for HIV and AIDS, Religion and Theology (CHART) (N=14). Experts were from different high burden countries and represented international organisations such as the World Council of Churches, IAS, UNAIDS and several universities.
Step 8
Based upon the discussions in the experts working groups, a final research priority agenda was constructed by three researchers (MN, MK, SLvE). This agenda was member checked and approved by five experts from the two working groups in step 7.
Phase 3 Making the research priority agenda work
Step 9
As a final step, we closely collaborated with key funders, researchers, HIV activists and other key stakeholders to promote the use of the research priority setting agenda. The agenda was presented at several national and international conferences, discussed with key stakeholders in planning meetings and in workshops and both a soft copy and printed booklet has been made available. In addition, collaborations with global partners such as the World Council of Churches, Christian AIDS Bureau for Southern Africa (CABSA) and others will ensure that the agenda is known and used in an international context.
Descriptive data from the questionnaires was analysed using SPSS statistics package version 25.0, USA. This study was approved by University of Cape Town Health Research Ethics Committee (Reference number: 123/2015). Written informed consent was obtained from all participants prior to the interviews, focus groups and questionnaires.