Health and medicine sociology

Social in/justice through data-driven healthcare technologies: empirical findings of early career researchers (session 2 of 2)

June 28, 2021 15:00
June 28, 2021 16:30

Laetitia Della Bianca, University of Geneva; Mélody Pralong, University of Geneva; Martina von Arx; University of Geneva


Maud Reveilhac - Lausanne University, Faculty of Social and Political Sciences, Institute of Social Sciences, Life Course and Inequality Research Centre; Arnaud Blanchard - Lausanne University, Faculty of Law, Criminal Science, and Public Administration, Institute of Private Law

Bastien Presset, University of Lausanne

Pei-hua Huang - Erasmus MC; Ki-hun Kim - TU Delft; Marina Bos-de Vos - TU Delft

Cristina Richie - Philosophy and Ethics of Technology Department, Delft University of Technology; Heike Vallery, BioMechanical Engineering Department, Delft University of Technology

Importance and have been pushed further by the SARS-CoV-2 pandemic in 2020. These digital tools are often promoted as inclusive and empowering for their users. Yet, while creating more social justice for some, they might produce other blind spots of social injustice for others. The present session invites early career scholars to present their empirical findings regarding the development and/or use of healthcare technologies (such as mobile health apps, telemedical tools, monitoring technologies, algorithmic treatments, etc.) within different matrices of power. Our aim is to gather understandings of the roles played by various actors within different networks of specific data-driven healthcare technologies. 

This paper session especially seeks to address the inclusionary and exclusionary effects of data-driven healthcare technologies. It questions to what extent such technologies blur traditional boundaries of healthcare and uncertainty in medical decision-making. Moreover, this session aims at disentangling whether related activities remain geographically and temporally bound. We, therefore, invite papers that address one or several of the following questions: 

  • How are algorithms interfering with the expertise of healthcare professionals and patients? Where and when does the so promoted anytime and anywhere healthcare take place? 
  • How do data-driven healthcare technologies reconfigure health inequalities? 
  • Are privacy concerns a privilege of the healthy? 
  • To which extent do algorithmic technologies allow for more precise and fair medical decisions? 
  • How do people adapt to or resist healthcare technologies? 
  • Which health-related issues do these technologies address and which are ignored?
  • Who has access to these technologies and the data they produce, and to whom is such access denied? 
  • Which places are configured as ‘well-suited’ for the use of such technologies, and which ones are excluded? 
  • How do these technologies reconfigure and redistribute the medical work within traditional and less traditional healthcare settings? 
  • What role do concepts such as justice and rights play in the design of data-driven healthcare technologies? 

We especially welcome contributions that draw on Science and Technology Studies, Feminist Technoscience Studies, Critical Data Studies, Sociology of Health and Medicine, and use an intersectional lens to question the role that gender, race, disability, migration, and class may play in such practices.

The framing of health-technologies on social media by major actors: prominent health issues, impact on professionals, and COVID related public concerns.

Maud Reveilhac - Lausanne University, Faculty of Social and Political Sciences, Institute of Social Sciences, Life Course and Inequality Research Centre; Arnaud Blanchard - Lausanne University, Faculty of Law, Criminal Science, and Public Administration, Institute of Private Law

Endorsing a multi-disciplinary approach at the nexus of sociology and computational social science, we focus on prominent actors in the field of healthcare technologies (HT) who are active on social media. Relying on a Twitter corpus of more than 4000 accounts (e.g., companies, governmental agencies, scholars, practitioners, and enthusiast advocates), we study how these actors shape the applications and implications of data-driving technologies for professionals and patients, as well as for developing new regulations.

The actors included in our corpus remain geographically bound to the United States, followed by Canada, and the United Kingdom, thus reflecting the situation of major HT markets offline. Their social media presence further correlates with survey measures of citizens’ Internet searches for health-related issues at the country level (OECD 2020; Eurobarometer 2014). Within European countries, there is a moderate positive correlation between the level of reliance on Internet for health-searches and the retweets ratio (above 0.4 for Internet generally and above 0.6 for social media specifically). 

Topic modelling (Blei 2012) allows us to automatically extract the most prominent themes contained in the actors’ tweets (above 7.5 mill. tweets extracted). Salient themes address patient-doctor relation, patient-centred initiatives and needs, healthcare systems, innovative solutions, big data challenges, market opportunities, and customer experience. There are, however, differences in topics prevalence between European countries and the United States, as the latter emphasise more on risk management and private funding, whereas European countries focus more on health literacy, practitioners (as opposed to scholars), and start-ups.

Word embeddings (Sahlgren 2006) further enables us to detect synonyms and associations surrounding terms of HT. We demonstrate that certain health issues – such as cancers, Alzheimer, AIDS, Diabetes mellitus, and Influenza – are perceived as more ‘well-suited’ in terms of innovation and research, which reflects the scientific interest and funds allocation to investigate these health issues (Head et al. 2020). The focus on algorithms, artificial intelligence, and machine learning is also reconfiguring the relationship between health professionals and patients toward a more preventive medicine, greater patient autonomy, and new research skills for professionals. Furthermore, these data-driven technologies raise data privacy discussions associated to professionals’ obligations (e.g., ethics, regulations, and legal issues) and patients’ concerns (e.g., cybersecurity, transparency, data portability, and consent). Finally, we show that the COVID pandemic exacerbates recent concerns related to economic and individual rights in terms of ‘liberty’ and ‘tracking’, thus acknowledging that HT based on continuous data collection and algorithmic evaluation have gained importance during the pandemic.  

Our study sheds new light on HT by showing how major actors in this field legitimise specific understanding of HT, thus complementing studies about health-related practices on social media (Lupton 2012) and the potential of social media applications for health (Koteyko et al. 2015). Gaining understanding in this discursive domain is paramount as internet is growingly important for citizens to inform themselves on health-related issues (OECD 2020) and social media platforms are essential for companies and specialists as part of their publicising efforts (Lupton 2012).

Digital self-tracking practices and health insurance: inequalities beyond promises

Bastien Presset, University of Lausanne 

Modern forms of tracking (via applications, smartwatches, etc.) are part of the ongoing digitalization of ‘health’. Drawn by promises or injunctions, many public and private health institutions have engaged in the development and promotion of so called ‘self-tracking’ technologies. Those are always presented as free and inclusive, an optimistic and deterministic discourse that often accompanies digital technologies (Wyatt, 2000). 

There have been many studies of self-tracking practices which have highlighted their role in the constitution of modern ‘digital-selves’(Lupton, 2016). Lot of emphasis has been put on the disciplinary power of ‘self-tracking’ and on the possibilities to ‘resist’ it (Nafus & Sherman, 2014). But inequalities among users have been a blind sport of this literature, despite the fact that studies on other types of mobile health technologies strongly suggests that they exist (Bol et al., 2018). This absence may be attributed to the difficulty to access large empirical sets on a single tracking technology. I offer to make up for it and to try to discern - under the myriad of promises (Audétat, 2015) that accompany them – whether ‘self-tracking’ technologies produce, re-produce or transform forms of digital inequalities among their users. 

To that end, I draw from questionnaire (n=1026) and interview (n=21) data gathered among users of a pedometer application developed by a major insurance company. I use descriptive statistics and binary logistic regression to focus on the influence of broad indicators such as age, gender, education and income on the adoption of the technology. It reveals that age, education and income all have a significant impact on adoption. I then further explore those results under the light of the interview data. Here, I focus on differences that arise in the way people make sense of numbers and use the technology in daily life. I distinguish four different types of users. I frame my results with the concept of digital divide which has been developed to highlight different forms of inequalities in the adoption of technologies (van Dijk, 2006). My results confirm that forms of digital divide are present both in the adoption and in the use of self-tracking tools. This prompt me to mitigate the – sometimes blind – solutionist belief that self-tracking practices will help solve public health issues.

Pre-Empting the Unjust: An Ethical Framework for Digital Twins for Personalised Healthcare Service

Pei-hua Huang - Erasmus MC; Ki-hun Kim - TU Delft; Marina Bos-de Vos - TU Delft

There is a growing enthusiasm for exploring the potential of digital twins for personalised healthcare. As a ‘digital replica’ of its ‘physical twin’, the digital twin offers clinicians an opportunity to simulate the efficacy of different treatment on their patient’s digital twin and tailor the treatment course based on the result. Digital twins also have great potential in early diagnosis and early prevention of lifestyle diseases. Augmented with other advanced technologies such as machine learning algorithms and wearables, a digital twin may also serve as an affordable health consultation, offering patients personalised advice regarding achieving their desired health conditions. 

Despite the interests shown by both the public and private sectors in integrating digital twins into personalised healthcare services, many scholars have raised concerns about this endeavour to digitalising healthcare. They worry that implementing digital twins in the healthcare sector might not empower patients but invite overdiagnosis, coercive healthism, and distortion of self-understanding. Some also fear that the ‘fail-fast’ IT culture might bring harmful influence to the healthcare sector, destabilising the healthcare system. These critical remarks highlight the possibility that the enthusiasm for digitalising healthcare might turn out bringing more harms than benefits. 

Currently, most of the literature focuses on either the potential benefits, or the potential harm, a digital twin might bring to a patient. Relatively few address the issue on how to help developers proactively address ethical risks when developing a digital twin for personalised healthcare.

Acknowledging the health benefits and ethical risks of digital twins might bring to individual patients, this paper aims to fill the gap in the literature and provide an ethical framework for developing digital twins. The framework offers a systematic understanding of digital twins for personalised healthcare and details a digital twin’s capability in creating desirable values for personalised medicine. The framework also serves as a framework for responsible innovation by facilitating developers of digital twins in identifying ethical risks they ought to proactively address during each designing phase. For instance, during the data collection process, poor data quality might distort the patient’s digital twin, resulting in unhelpful and even harmful health advice. Besides, algorithmic biases might occur during the information production phase. Coercive healthism could also harm a patient’s well-being when the targeted ‘value’ is not in congruence with the patient’s value system. We will use a digital twin for cardiovascular health management we are developing at TU Delft and Erasmus MC as a case study to further illustrate how to apply this framework for responsible innovation.

Section 1 presents a brief overview on the concept of digital twin and its potential to offer more personalised healthcare and enhance patient autonomy. Section 2 identifies three major phases in value creation processes: (1) data collection, (2) information production, and (3) value creation. Section 3 highlights ethical risks in each of the three phases. And finally, Section 4 presents how we apply this framework in designing a digital twin here at TU Delft and Erasmus MC as a case study.

Ethical Considerations in Rehabilitation Technology Beyond Benefit to User and Scientific Progress: Environmental Impact and Justice

Cristina Richie - Philosophy and Ethics of Technology Department, Delft University of Technology; Heike Vallery, BioMechanical Engineering Department, Delft University of Technology 

Sophisticated health care technologies, such as those used in rehabilitation like a balance-assisting gyroscopic backpack or a smart walk assist have the potential to radically improve the quality of life of those affected by motion disorders, spinal cord injury, stroke and other neurological disorders affecting gait. These medical tools are generally well received by a public willing to integrate medical health care tech into their lives. Yet, numerous ethical considerations are overlooked in the production and manufacturing of rehabilitation technology. 

This paper will first overview the uses and prevalence in rehabilitation technology, with a focus on its use and development in the Netherlands. Then, it will present the primary ethical arguments in favour of these technologies—benefit to user and scientific progress—in tension with (but not necessarily in opposition to) environmental impact and distributive justice. All health care technologies rely on resource use and produce carbon emissions, which contribute to global climate change and climate change health hazards. Ethically, there must be an appropriate balance of benefit for individuals (the technology) and global health (climate change health hazards). Second, rehabilitation technologies, which are underpinned by engineering and manufacturing, thrive on constant improvements, increasingly sophisticated programming, and technological obsolescence. These are often supported by grant-funding organizations. Rather than making technologies more affordable and widely disseminated, oftentimes the opposite is true. Despite offering only marginal benefits, cost is the same as the older versions, or higher. This increases gaps in access, while simultaneously clustering health care tech in affluent areas, thus perpetuating misdistribution of resources. Just allocation of health care tech must be embedded in the ethos of engineering, manufacturing, and funding so scientific progress can benefit everyone in society. 

The paper will conclude with a discussion of the ethical issues in rehabilitation technologies, with a focus on policy development for engineers, manufacturing companies, and grant agencies. Engineers can recognize the difference between technological developments that will be statistically significant and clinically relevant with those that are economically and environmentally ethical, and design accordingly. Manufactures must confront the reality that circular economies are often unpractical for technological progress—which relies on linear developments—and thus redesign lifecycle processes, instead of appealing to “the future” of sustainable and just health care technologies. Grant agencies, such as the Dutch Research Council (NOW) and European Union (EU) could have environmental impact and justice at the core—not peripheral—of funding decisions.

Health care technologies, such as rehabilitation technology, do not have to hinder sustainability and justice. In fact, they can support both, but only if ethical issues are highlighted and addressed. A global society cannot progress at the expense of the planet or the majority of people that inhabit it.