Ethics statement

The WHO’s Regional Office for Africa endorsed the study and Harvard University’s institutional research board provided an ethics exemption for this study (protocol number: IRB21-1176). The recruitment period was from March 31, 2022 to September 1, 2022. Consent was a condition of completing the written survey. The survey contained an introductory statement on the instrument itself explaining the study, stressing the completely voluntary option to complete it, and specifying that participants can decline to participate in any part of this study for any reason and can end their participation at any time. More details on the survey content are noted below.

Study design

An 18-item survey was developed and administered to directors of policy and planning and those in related positions to glean how health equity is operationalized. A survey was chosen to both provide us with a broad understanding and collect data from as many respondents as possible, given the busy nature of policymakers’ jobs, maximizing the number of individuals we could reach. M.A.’s previous research on the concept of health equity as operationalized by the WHO informed the development of survey questions. Once an initial draft of the survey was complete, J.B.B., a global and public health scholar who is well-versed in qualitative methods, reviewed all survey items.

Individual consultations were then undertaken by M.A. with six directors within national governments across the African Region. These consultations focused on sharing the primary objectives of the study, the value that this study seeks to add, and the survey questions. Feedback was collated and survey items were accordingly revised. For instance, it was suggested to indicate that directors can also include managers and other high-level policymakers to recognize differing job titles across countries and collect data on the specifics of individuals’ roles.

Distribution strategy

An invitation to complete a Qualtrics survey was disseminated by email which included a description of the study’s purpose, eligibility to participate, what participation entails, and benefits and risks to participation. The email distribution list included 481 email addresses from those who were invited to the WHO’s Fifth Health Sector Directors’ Policy and Planning Meeting for the WHO African Region [11]. The email invitation was also shared with 55 non-duplicate email addresses that were taken from an online list of Ministry of Health policy focal points.

It was not possible to discern if all email addresses on the complete distribution list were relevant to the study. For instance, translators and other similar individuals were included on the email distribution list from the WHO’s Fifth Health Sector Directors’ Policy and Planning Meeting. Additionally, some email addresses had university domains rather than government domains. However, it is possible that the owner of an email address with a university domain could have been an alumnus who became employed as a civil servant. Therefore, the entire email list was included in the survey distribution and email addresses were not excluded based on having a university-affiliated domain. Email addresses with any of the following institutional domains were excluded from the distribution list to reach policymakers as opposed to global actors: who.int; unicef.org; giz.de; unfpa.org; pasteur.sn; kemri-wellcome.org; gavi.org; gatesfoundation.org; dfid.gov.uk; and worldbank.org. Duplicate emails were excluded from the distribution list.

Eligibility to participate was further determined via a sociodemographic screening at the beginning of the survey. The sociodemographic screening gathered information about where participants work, the focus of their work, and the government level at which they work. Participants were eligible to participate if they (i) were a director of policy and/or planning or similar role, and (ii) worked in any government in the WHO Region of Africa.

Pilot testing of the survey

From the complete email distribution list, 50 email addresses were randomly selected to pilot the survey to determine if any changes to survey questions were needed before distributing the invitation to participate to the remaining email addresses. Pilot respondents’ responses appropriately answered the survey questions indicating that the questions posed were clearly interpreted. Therefore, no further changes to the survey were required.

Response rate

The invitation to participate was distributed to a total of 536 email addresses. There were 282 emails that bounced back and 254 emails that did not bounce back. Therefore, a maximum of 254 potential respondents received the email invitation. The actual number received would be lower in ways we cannot estimate, including how many were diverted by spam filters or how many went to inactive accounts. All potential respondents were sent a total of four emails at approximately one-week intervals with the invitation to participate in the survey. Subsequently, a WHO official sent one follow-up email to encourage participation in the study.

There were 42 respondents (although one respondent did not complete the survey in full) across 18 of the 47 countries within the African Region (as outlined by the WHO). The response rate was 16.5% based on the 254 potential respondent figure. Given that the list of potential respondents who received the email invitation to participate in this survey included non-directors (e.g., translators), it is important to note that 16.5% is not an accurate representation of the response rate of eligible participants. The email invitation included information on eligibility. There were no ineligible respondents who attempted to complete the survey. Had a non-eligible respondent attempted to complete the survey, they would have been excluded by the sociodemographic screening.

Survey content

The survey contained an explanatory preamble, noted that participation is voluntary, outlined the expected time commitment of participation (5-15 minutes), and pointed out the potential associated benefits (e.g., inform policy work that aims to improve global population health) and risks (e.g., a very small potential risk of being identified through participation). Great care was taken to protect participants’ anonymity, for example, by excluding country-specific details in this article. The preamble also clearly indicated that the respondents could decline to participate without consequence. At the end of the survey, respondents were invited to participate in a follow-up interview and were thanked for their time and participation. These interviews have yet to be conducted as they are planned for a subsequent study.

The survey collected demographic information and asked a series of questions focused on health equity in policy and practice. The questions for which responses were analyzed in this article related to the consideration of health equity in participants’ government departments, how participants directly incorporate health equity into their practice, particular policies or programs that address health equity, and challenges to incorporating health equity in policy and planning work.

Data analysis

Survey responses were collected in English, French, and Portuguese and were subsequently translated using DeepL software [12]. There was no set coding framework or a priori codes. E.H. and A.S. independently inductively coded the English responses and responses translated to English, guided by Braun and Clarke’s steps for thematic analysis [13–15]. E.H. used NVivo12 [16] and A.S. manually coded the survey responses in Microsoft Word. Once the coding was complete, E.H., A.S., and M.A. met to discuss all codes, noting convergences and divergences, and the relevant emergent themes. There was a majority of overlap in codes and their underlying meaning and divergences were few and easily agreed upon; when one author had a code that the other missed, consensus was consistently reached for its inclusion. Three overarching themes, which were not mutually exclusive, emerged from the data related to how respondents acted on health equity. It is important to note that these themes emerged from the data through the lens of the authors’ analysis and were not explicit approaches described by survey respondents. The themes, and subsequent sub-themes, are detailed in the results below.

How respondents sought health equity

Two main approaches for addressing health equity emerged. Most respondents mentioned the prioritization of a particular population or disease groups, thus taking a targeted approach. Some respondents alluded to their government’s policies and/or practices as ensuring services are available to all, thus taking a population-wide approach. Participants also discussed procedural justice and operationalizing population health data in seeking health equity. Although not mutually exclusive, more respondents referred to the first approach as their overall approach. These approaches in which health equity is considered in policy and practice are described below.

Challenges to incorporating health equity in policy and practice

Despite the foci of health equity efforts described above, respondents also noted challenges for further incorporating health equity. These challenges were related to four key domains: (i) understanding health equity, (ii) governance, (iii) resources, and (iv) lack of data. Notably, respondents’ perceived challenges for materializing health equity in practice were mostly within the second and third domains. Sample excerpts from respondents are noted in Table 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Key challenges with sample quotes from respondents.

https://doi.org/10.1371/journal.pgph.0004384.t001

Complexities of equity evident in approaches to health equity

The variety of themes raised in the survey responses illustrate the complexity of health equity. Inequities entail unjust differences, and justice is, of course, a multifaceted concept. To begin with, justice can be applied at several levels of analysis. For example, one can ask about the justice of a distribution of health resources, but also whether the process that led to that distribution was just. Furthermore, at any level of analysis, distinct perspectives on equity exist. Consider this point in connection with distributive justice. Some claim that a just distribution of resources prioritizes the least well-off, a view known as prioritarianism [18]. Another approach, known as sufficientarianism, insists that distributive justice requires everyone to have an adequate set of resources [19]. These two views of distributive justice need not be regarded as mutually exclusive, and some theories of health equity incorporate both [20]. Joint consideration of prioritarian and sufficientarian approaches is also suggested in some WHO documents on health equity. For example, a 2010 WHO report on addressing inequities emphasizes “the need for universal social protection systems (‎ using targeting only for populations that fall through the cracks)” [21].

The aspects of equity just noted have parallels in the survey responses. Thus, associating equity with interventions that target health resources towards especially vulnerable groups, such as people living with HIV or children of low-income families, can be interpreted as an example of prioritarian thinking. Conversely, linking equity to national efforts to ensure universal access to health care could be seen as an example of a sufficientarian approach. Procedural justice concerns are also apparent in some survey responses, such as those associating equity with stakeholder consultation. Responses about priorities being set in a top-down manner by international donors are also related to procedural justice. Again, the concern is that processes determining health priorities and resource distribution do not adequately involve the impacted populations.

Finally, the views on equity described above entail difficult judgments and decisions. To implement a prioritarian approach, it is necessary to decide who is most vulnerable. If resources are scarce, it may not be possible to both prioritize those in greatest need while ensuring that everyone has access to adequate health care. Thus, it is unsurprising that problems of deciding who should benefit from interventions were mentioned by some respondents as a barrier to equity. One goal of future work should then be to provide a framework to support policymakers in articulating and addressing such challenges.

Targeting was the predominant approach to health equity

Although respondents did not explicitly use the terminology “targeted approach” to describe how they value and mobilize health equity, interventions for “vulnerable” population groups based on a specific criterion were often equated with health equity itself. Health equity was predominantly operationalized through interventions for those with fewer means (e.g., lower socioeconomic status), specific target populations (e.g., women and children), and those with select conditions (e.g., HIV/AIDS, malaria, tuberculosis). Analysis of responses highlighted that very few characteristics (e.g., age, sex, gender, location) were discussed by respondents when describing a targeted approach, but no attention was paid to sexual orientation, race, or Indigeneity. There was also no discussion of intersectionality. Given that characteristics can interact in unique ways to further perpetuate health inequities, differing needs must be considered [22]. However, it is difficult to determine if this resulted because the questionnaire did not elicit information on intersectionality or because it was not a consideration for the responders. As such, we suggest future research explore policymakers’ views on operationalizing intersectional health equity efforts.

Resource constraints are not the only barrier to health equity

Many respondents noted that resource constraints are a barrier to health equity. Given that initiatives such as providing perinatal care or ensuring access to adequate health care for all inevitably come with a price tag, these responses are entirely understandable. However, we suggest that it would be a mistake to conclude that promoting health equity is solely a matter of securing funding for needed projects. That is because promoting equity is not only about doing more; in many cases, it is also about doing things differently, for example, through engaging in more equitable partnerships [23]. This point is especially clear in connection with procedural justice. For example, consider a system in which global health priorities are set by donor organizations without adequate input from target countries, mentioned by several respondents as an obstacle to health equity. Changing such a system may be more a matter of political will to reform organizations than acquiring new infusions of funds. Similarly, a lack of understanding of health equity and how to implement it in practice were also noted as barriers. The survey responses, then, suggest that advancing health equity is not only about finding more resources, although that is often important. Reforming top-down processes through which priorities are set and decisions are made and establishing mutual trust may also be crucial.

Competing priorities and agendas can conflict with anti-colonial approaches and have implications for health equity

Results demonstrated that external agendas (e.g., disease-focused funding such as those from non-governmental actors) may skew national priorities, which can conflict with anti-colonial approaches and reduce procedural equity. In conducting the analysis, it was apparent that many respondents drew on WHO or other United Nations (UN) discourses when discussing priorities. This included mention of the Sustainable Development Goals (SDGs) (“This will help to solve the problem in the right way and reduce injustices to achieve the SDGs”); focus on UHC which is a major priority for the WHO (e.g., “We are currently in the process of setting up universal health coverage”, “We have developed the Universal Health Coverage (UHC) roadmap and implementing it with the aim of achieving the three principles of UHC, namely, access, quality and affordable cost.”); and through historical statements promoted and regurgitated by the WHO (“Leave no one behind” and “Our goal is health for all”). This seeping of WHO and UN language perhaps points to the influence of these international organizations on countries’ policy priorities, agendas, and policy development. With recent attention paid to decolonizing global health and anti-colonial efforts more generally [24], further research is needed to assess the dynamics between international organizations and governments and the plausible influence of international organizations on national public policymaking. For instance, does earmarked funding provided by international organizations sway national policy agendas and associated actions away from national priorities?

Further, if global agendas distract attention from more pressing national issues making it on the policy agenda, this can have implications for health equity. For instance, although noncommunicable diseases (NCDs) have become the leading cause of morbidity and mortality worldwide, they still receive less political and financial attention from the global health community compared to other conditions, such as HIV/AIDS [25–27] as cited in [28]. The focus on the “big three”—HIV/AIDS, malaria, and tuberculosis—which are also prioritized by many global health organizations perhaps reveals the origins of these priorities. In cases where a government prioritizes improving health equity for populations at risk of NCDs, but donor agendas focus on the big three, national health equity efforts for NCDs may be hindered. On the other hand, increased global attention on UHC could shift the focus of national agendas, potentially advancing health equity in areas where it had not been prioritized previously.

Limitations

This study has limitations in at least four areas: response rate, missing perspectives, computer translation, and survey method.

First, in terms of the response rate, because the complete participant list from the WHO’s Fifth Health Sector Directors’ Policy and Planning Meeting included participants who were not target participants in the study (e.g., translators, observers, report writer), it is impossible to know what denominator to use to accurately determine the response rate. We expect the true response rate to be higher than the 16.5% response rate calculated. Although there is no standard figure for an average response rate, our response rate is not high enough to eliminate concerns. Evidently, there may be various factors contributing to the response rate beyond the control of the researchers, such as survey invitations being filed in spam folders.

Second, despite the survey not having the intention to be representative or generalizable, there were no responses from certain countries and a higher response rate from other countries. For example, there were six responses from the United Republic of Tanzania. Therefore, it is important for future studies to include missing perspectives to have a better understanding of the views of individuals from other countries as well.

Third, because this study was conducted in three languages, we translated French and Portuguese responses into English using computer software to facilitate analysis. Therefore, it is possible that nuances in the use of certain words or expressions were not captured. We expect this limitation to be immensely minimal, as we do not expect any desired meanings to be misinterpreted.

And lastly, there is a recognition that the survey method may not be the most appropriate approach for collecting lengthy response data. Given how surveys are administered, the survey administrator may not be able to interact with respondents in real time despite the trade-off of potentially reaching more respondents. Unlike research methods like interviews or focus groups, respondents are unable to ask questions to clarify the questions and the survey administrator is unable to request additional information if a response is unclear. Without the opportunity to have a conversation or seek and provide clarifications for either party, there may be the possibility of misinterpretations or incomplete responses. We suggest that future studies employ key informant interviews to obtain a better understanding of responses around health equity.