How can digital health support the needs of the worst off?

In order to support these prioritarian intuitions and facilitate practical reasoning in programme design, development and implementation of digital health, we offer a set of three principles. The principles were developed for two main reasons. First, there is a need to offer an alternative to the dominant approach to digital health in LMICs of first focusing on those who are ‘easiest to reach’, as distinct from the ‘seldom-heard’. The needs of the most marginalised are often the last to be addressed; their voice is lost to those in a more privileged positions and those who are valued as early adopters. Second, one of the key failings of a utilitarian positioning to digital health (or any innovation) is that it does not sufficiently recognise ‘the complex, networked and learning-intensive features of technology’,41 or the social and contextual meaning of technology42 and is overly concerned with scale.30

If such an understanding is not urgently integrated into programmes, structural neglect and digital inequalities could result from the use of digital health in low resource settings. Our response has been to develop the principles as a useful reasoning tool by which necessary consideration is given to how digital health programmes can be weighted in favour of the worst off, especially during the early stages of their design and implementation. These principles should be viewed as evolving and are intended as a baseline for discussion. The primary aim is to support the processes of working through how to embed a prioritarian ethical framing when using digital health and technology in development more broadly. However, while we focus on digital health, we believe the prioritarian principles could have wider applicability to other domains of global health and development.

Principle 1: identify what constitutes disadvantage so the social value of digital health can be maximised for the worst off

Barsdorf and Millum discuss how the social value of a project can be determined by its benefit to marginalised populations: ‘the more disadvantaged the beneficiaries of a research project, the higher its social value’.43 However, in order to determine social value, we need to first identify what constitutes disadvantage. Prioritarianism relies on the ability to determine who is the worst off so that differences in distribution can be accounted for in the allocation of resources (money, time, technology and so on). Sharp and Millum offer one way to determine who are the worst off.44 One core component of their argument, which is relevant to our discussion here, is their review of the different approaches to conceptualising disadvantage and their objects of value (moral goods, advantages or ‘currencies’ in their terminology). They acknowledge that developing ‘an operational measure’ of disadvantage is a significant challenge because of the ‘the long-running philosophical disagreement about the currency of advantage’.44 That is, in addition to what the distribution rules should be, there is also profound disagreement about the valuable good to be distributed: ‘[r]esourcists, welfarists and proponents of the capabilities approach all characterise advantage in different ways’.44 Sharpe and Millum reach the interesting conclusion that there are identifiable common constituents of disadvantage across the different moral goods:

…there is considerable consensus about a core set of indicators that track advantage, however defined. For instance, it is widely agreed that education, minimal levels of material wellbeing, mental and physical health, and the protection of certain civil and political rights are all very important to ensuring the opportunity for a flourishing life. Thus, for practical purposes, those who lack access to education, have a low material standard of living, face severe health problems, and whose rights are not protected, are very likely to belong among the worst off.44

Their discussion is helpful in so far as it at least begins to identify how to distinguish who is worst off. Moreover, the review of various approaches to distributive justice shows that there are substantial and active bodies of research helping to identify the worst off on which prioritarianism in digital health can build. The principle itself is agnostic about the form of disadvantage. That is, the worst off do not have to be worst off in terms of the holdings of the good to be distributed. They can be worst off in other forms of disadvantage.

Principle 2: the design and implementation of digital health should address the needs of the worst off through their participation

Once the worst off have been identified, and/or social gradient in disadvantage determined, the second principle focuses on determining how best digital health can address their needs. This requires developing a deep understanding of the relationship between technology and the intersectional disadvantage the worst off face on an everyday basis. Technologists have not done a great job alleviating complex disadvantages that have produced healthy scepticism regarding the potential impact of technology on the lives of the worst off.45 Principle 2 asserts that digital health can and should address the needs of the worst off, but only when technology design and development involves the worst off in the design process in a meaningful way.46 The principle makes sure that technological developments addressing their needs are identified, addressed and integrated from the start. This principle draws on substantive research in social justice (eg, Farmer’s ‘preferential option for the poor’,32 Fraser’s conceptualisation of ‘parity of participation’,47 with the relatively small but growing community of Human Computer Interaction for Development (HCI4D) researchers and practitioners48 and on research in various forms of Participatory Action Research (PAR).46 Through a process that reflects the principle, the form of digital health intervention can be specified iteratively in increasingly higher fidelity and directly related to its function of benefiting the worst off.

A helpful illustrative example comes from our research on the co-development of a mobile-based training programme for community health workers (CHWs) in Kenya starting in 2013. To address the needs of the worst off in their community—in this case, children with disabilities and their mothers—smartphones were chosen as the technology, rather than the far more available low-end phones. The participation of CHWs in designing the intervention was critical to this technology decision. Their stated need to be able to assess the health and social care needs of children in their community required an app with smartphone-enabled functionalities (rather than a low-end phone). Thus, by taking a priority view, we rejected a utilitarian, ‘top-down’ approach to technology design and implementation that would have meant not working to overcome the intersectional disadvantage faced by the children with disabilities and their mothers and instead would have focused on an existing priority area.49 We emphasise that the rationale behind using smartphones as part of the programme intervention was not utilitarian or maximising reach for participants. Instead, the focus was on a training and decision-making app that was co-designed to directly meet the needs of the CHWs and the mothers of children with disabilities. Simply using the most available option of low-end phones would not have offered the best digital health intervention possible, given their limited functionality.

This reasoning has parallels with Farmer’s and Gutierrez’s concept of accompaniment. Accompaniment for them is listening to the poor and the oppressed and walking with them, ‘thinking, living, working and reflecting with the people who know most about poverty’.50 In terms of a programme design in digital health, the ways in which technology can be embedded to support outreach (through improved communication and peer decision-making, eg) is critical and remains under-researched. This principle supports this process through the exploration of the following questions as part of any design process:

• How can marginalised groups have a role in the design and roll-out of new technologies?

• How can an equitable environment for these developments be intentionally promoted?

• How can marginalised groups be supported to identify culturally relevant and contextually appropriate interventions?

• Are the most disadvantaged within the disadvantaged participating and voicing their needs?

Principle 3: be ethically reflective about our (global) ethical obligations as (global) digital health researchers and practitioners

Identifying the most disadvantaged (Principle 1) and prioritising the voices and needs of the marginalised in the design, development implementation and evaluation of digital health technologies (Principle 2) leads to the third ethical principle of assessing our obligations as researchers and practitioners. The hard truth of global and digital health is that those with power (eg, funders, practitioners and researchers) must provide adequate resources if they are serious about aligning themselves with prioritarianism when addressing the needs of the worst off. This implies that those with power must go beyond their usual day-to-day interactions with people who are marginalised to become advocates for them.32 Instead of relying on an ethical stance that only describes and manages social inequality (which has merits in itself), the priority view transforms those with power into advocates for social change by effectively tackling the social and political structures that produce and perpetuate inequality. That is, a multitude of intersecting political, social, cultural and economic processes and forces often conspire to constrain the agency of those people who are most disadvantaged.32 Raising their well-being is not a passive or minor effort. It is an effort that militates against the injustice of status-quo prevailing inequalities of advantage or well-being.

Let us take an example from our previous work investigating existing research on the supports available for abused children with disabilities in East Africa49 . In order to motivate the need to help those at the bottom first, from a research perspective, it was necessary to map the types of evidence that exist on this topic. Our scoping review and evidence map documented the coverage, patterns and gaps in existing research on the abuse of children with disabilities in East Africa. In doing this, we showed that the abuse of children with disabilities—a highly marginalised group—was a topic grossly neglected that must be prioritised. This research provided the necessary evidence to underpin and develop a strategic agenda in this area to better determine what future research in order to undertake to address this pressing lacuna. In some ways, this involved us assessing our obligations as researchers. One key question we faced was how can this be achieved in a pragmatic way that allows us to promote the voices of the worst off? To address this question, we drew on the research of Sarradon-Eck et al on caring on the margins of the healthcare system.51 Through their work, they emphasise the necessity to go beyond immediate improvements to healthcare for the worst off and change the environment to support them to regain their human rights, their access to social protection, and ultimately, their autonomy and social identity as participating members of society. They point out the onus is on projects to reach out to the worst off, not on those in need of care to seek it.

Principle 3 can be applied at the systems level, where we are often dealing with people who are not only marginalised but also highly stigmatised (eg, mothers of children with disabilities). Sarradon-Eck et al point out how this may lead some doctors and nurses in the formal health system to view the marginalised with ‘a form of symbolic contagion from the stigma’ arising from their situation.29 This is important to take into consideration when addressing the social factors that impact on the integration of digital health programmes aimed at the marginalised into the wider health system. Reflecting on our role as researchers and practitioners, this brings up the absolute need to attend to issues of power, representation and culture.30 Asking ethically reflective questions of how project interactions may have created misunderstandings, how to ensure all participants have the opportunity to take part and recognising the ways in which interactions can be ‘subtly oppressive’ and potentially reinforcing power, particularly in cases where technology is new to participants. Providing opportunities to elicit feedback to make adjustments to any instructional design, including knowledge co-construction can be a critically important means of resisting this.