We followed good practice priority setting method, as described above. Our approach included a pre-prioritisation stage to identify potential priority topics through stakeholder engagement and review of the literature, and a prioritisation stage for consultation and finalisation of the priority topics through online surveys and consensus meetings, using specific criteria (Fig. 2). Country teams were responsible for implementing each step in their respective countries and any differences in the process due to practical or other factors within the three countries were captured as part of the documentation of the process.
Overview of priority setting approach overview
Pre-prioritization
Stakeholder identification and engagement
We engaged with two different pre-specified groups. The first were the members of the Guideline steering group set up in each GELA project country. Up to 13 individuals were identified and invited to participate from the relevant national departments or ministries of health, professional associations, country-level WHO offices, and any other individuals suggested by these bodies in each country. They were identified through existing contacts of the researcher team within national departments of health responsible for guideline development and working within newborn and child health area, with whom they had worked before. These individuals then also made suggestions of other members from other stakeholder groups such as academia, non-governmental organisations, etc., who worked in the field of newborn and child health. The steering group provided initial suggestions of priority topics, made the final decisions regarding which to prioritise, and provided general oversight and technical advice on the in-country implementation of the project.
The second stakeholder group was broader and included individuals or organisations who are involved, can affect or are affected by national decisions or actions related to priority topics in the field of newborn or child health in sub-Saharan Africa [27, 28]. These included policymakers, guideline developers, health professionals, civil society representatives, patient advocacy groups as well as WHO Afro representatives, specifically those linked to potential priority topics identified. To identify them, we carried out a stakeholder mapping exercise, which included reviewing secondary data, such as existing guidance and publications and searching the websites of ministries of health, relevant professional associations, universities, NGOs and civil society groups. The national GELA Guideline Steering group members also made suggestions. This process was guided by the stakeholder power-interest matrix where those who have the most influence, and capacity to change practice for impact were prioritised and invited [29]. Participants were invited, via email, to complete a priority setting survey.
Identifying a long list of topics
We generated an initial long list of potential priority topics through (i) reviewing existing and planned WHO guidelines on newborn and child health, from which potential topics were extracted based on existing recommendations; (ii) consulting with the GELA guideline Steering Group; and (iii) reviewing the disease burden/technical data related to newborn and child health in each country, which was identified through targeted literature searches. WHO guidelines were used as a starting point as these are prepared following rigorous methods and are intended for implementation across the various member countries. Potential topics were organised according to the disease/condition being addressed and the type of intervention (e.g., diagnosis, prevention, treatment, or rehabilitation), and were collated into a spreadsheet or word document.
Prioritization stage
Online survey
The potential topics identified in the first phase were included in online surveys with stakeholders. We first user-tested the survey among the GELA project team to ensure it was readable and understandable. We then invited all identified stakeholders, via email, to complete the survey developed using REDCap [30]. In Nigeria, the survey invitation was also circulated via WhatsApp to specific stakeholders. In the invitation emails, we also asked stakeholders to forward the email to any colleagues that may have an interest in the topic.
The landing page of the survey provided information about the study’s purpose, that it was a collaboration with the national department/ministries of health, what we were asking participants to do, and a link for more details about the GELA project, after which participants were required to provide consent before they could complete the survey. The survey asked respondents to rate the listed topics according to five criteria (Box 1) using a 6-point Likert scale (6—very critical and 1—not important at all) [22, 31]. The criteria used were identified through a survey conducted with the GELA research team in which they rated the top five criteria of 22 criteria for priority setting for guidelines identified by El Harakeh et al. [32]. An explanation for each criterion was provided in the survey. Originally, we had intended that stakeholders would rate each topic according to each criterion, but we asked instead that they consider the five criteria as a whole when rating a topic. We decided that this approach was less onerous for survey respondents and less likely to lead to poor response rates. Topics rated as being of ‘critical importance’ and ‘very critical importance’ were selected for presentation at the consensus meetings with the Steering Group. The first part of the survey also collected demographic information such as type of stakeholders, what they are primarily practicing as and for how many years, the percentage of time spent in patient care, and the type of institution they are primarily based at.
The surveys remained open for 3–4 weeks. Reminder emails were sent to those who were originally invited to participate, once in SA and Malawi, and weekly in Nigeria. As we could not track emails forwarded to others, due to the anonymity setting of the survey, it was not possible to remind those who had been invited in this manner.
Steering Group consensus meeting
Each country convened a meeting of their Guideline Steering Group to identify the final top three priority topics for guideline development. The number of topics per country was based on the number of recommendations that could be addressed in each country over the broader project period given the resources available. The moderator was a member of the research team and guided the meetings and discussions. This meeting was online in South Africa, and in-person in Malawi and Nigeria. We adopted a modified Nominal Group Technique [33] to achieve consensus, including five steps:
Step 1: the research team presented a summary of how the topics for the online survey were identified, which included consultation with the same steering group, as well as the results of the survey, including the summary of the topics that were rated as critically and very critically important, which needed further prioritisation.
Step 2: With the help of the moderator the group discussed each topic to ensure that all members understood them in the same way, and we elicited their thoughts on the ratings from the survey.
Step 3: The steering group members were asked to vote, anonymously, on the topics rated as critical or very critical in the survey using a Zoom poll (South Africa) or manually using post-its (Malawi). In Nigeria, the steering group decided to reach consensus through discussion to ensure full ownership of the resulting topics by every member of the group In South Africa only one round of voting was done as the members felt there was sufficient consensus after that. When voting, members ranked the importance of each topic on a six-point Likert scale against the same five criteria used in the online survey.
Step 4: The moderator summarised the results of the voting using bar charts to visualise the rating frequency distribution [34]. The variations for ratings observed were discussed among the group, including potential explanations before another round of voting, in cases where this took place.
Step 5: Consensus on the top three topics was reached. It was originally anticipated that at the end of this meeting three priority topics—formulated as questions in Population, Intervention, Comparator, and Outcomes (PICO) format—would have been identified for the next stages of the GELA project. However, the topics identified were broad, and further scoping of the literature and existing guidelines were needed to unpack and refine them. Several subsequent meetings with the Steering Groups were therefore needed in each country to present this scoping and finalise the three priority PICO questions per country.
Add Comment