Digitized Patients: Elaborative Tinkering and Knowledge Practices in the Open-source Type 1 Diabetes “Looper Community”

Introduction

How individuals know about their (chronic) diseases is an ongoing concern in the social studies of health and illness. Within different healthcare contexts, scholars of sociology and science and technology studies (STS) have argued for respecting the epistemology of how patients engage with their diseases, bodies, medication, and self-care technologies, emphasizing that patients ¹ are more than “important epistemic factors” (Falke 2018, 36) within the biomedical sphere (Epstein 1995; Pols 2013, 2014; Rabeharisoa, Moreira, and Akrich 2014). In this article, I explore patient knowledge practices in increasingly digitized, data-driven, and personalized healthcare settings by focusing on the “looper community” in Germany. Here a group of people with type 1 diabetes (T1D) develops and uses an open-source closed-loop algorithm to automate their insulin delivery and with this also improve their “chronic living” (Wahlberg et al. 2021). In their open-source endeavor, they also shed light on slow innovation cycles of commercial device manufacturers, including their data monopoly.

T1D is a chronic illness where the pancreas stops producing insulin so the affected person needs to administer it exogenously. Individuals’ illness-related knowledge practices consist of measuring their glucose levels multiple times a day and administering insulin according to those measurements. Contemporary T1D self-care is often described as a burdensome practice of repetitive data work for those affected (Jansky 2021; Forlano 2016). In order to care for themselves, people living with T1D constantly need to know how to mediate between different technical devices in, on, and with their bodies, such as insulin pumps and continuous glucose monitors, and the data these technologies generate (Kingod 2018). For individuals with T1D, creating and interpreting data about themselves is a daily necessity and an integral part of their knowledge practices (Mol 2000; Piras and Zanutto 2014; Danesi et al. 2020; Liggins 2020). Even with new digital technologies, they have to be alert and active in every moment of their lives (Kaziunas et al. 2017; Gottlieb and Cluck 2019). In completing their repetitive data work, they encounter black-box algorithms, limited openness regarding the use of the data generated and collected, as well as limited interoperability (Jansky and Langstrup 2022; Gottlieb 2021).

Dissatisfaction with existing self-care technologies foregrounds the slow pace of technological innovation in T1D, device manufacturers’ persistent nontransparency about data-handling, and the associated burdens of living with T1D. In this context, the looper community started to develop and use open-source algorithms to automate parts of the burdensome repetitive data work (Schipp et al. 2021; Lewis 2019). ² Their critique of current self-care technologies was later also reflected in the community-defining use of the hashtag #WeAreNotWaiting. Loopers are “deliberately non-compliant” (Scibilia 2017) by creating and using automated systems—which commercial providers fail to deliver—in an open-source manner. In practice, this means that loopers created an open-source control algorithm that enables an insulin pump and a continuous glucose monitoring device to communicate with each other and thus take over essential data work (Braune et al. 2021). The algorithm makes small adjustments to the insulin dosage every few minutes as it responds automatically to changing glucose concentrations registered by the continuous glucose monitoring device, intended to keep glucose levels within a predefined target range (Lewis 2019). To set up the systems, loopers need to engage medical devices such as specific insulin pumps and glucose sensors as well as smartphones or smartwatches. They can then access their data on an app, which also graphically represents what the algorithm does and what it predicts. Looping is not a frictionless endeavor and has a material component to it: some loopers tinker with the battery of their glucose sensor to prolong the battery life, others 3D print transmitters. Engaging in the looper community can be interpreted as forms of chronic self-care (Schipp et al. 2021), patient-driven innovation (Demonaco et al. 2019), and activism (Jansky and Langstrup 2022).

Building on seventeen months of ethnographic fieldwork in the German looper community in 2018 and 2019, twenty-eight in-depth interviews and an analysis of public documents, I explore empirically how new patient knowledge practices emerge within this open-source endeavor. I depart from the literature on patient knowledge and argue that the phenomenon of looping points to the need to expand established theories of patient knowledge.

How loopers know is characterized by knowledge practices described in patient knowledge literature, such as tinkering (Mol 2006), practical “know-now” (Pols 2014, 82), and “credibility struggles” (Epstein 1995). Yet, as I explain below, the observed practices are also technical, recursive, and methodological. To account for how these different forms of knowing are intertwined or are kept apart, I suggest the notion of elaborative tinkering. Patients engage in knowledge practices that amount to innovative ways to change their self-care and are characterized by both intimate practices of tinkering for immediate self-care purposes, and practices of scientistic experimentations to build a knowledge base for epistemic legitimacy, for example, by establishing a research project (O’Donnell et al. 2019). Current STS understanding of patient knowledge practices can be improved by empirical accounts of cases situated at the intersection of digitized and personalized health care, open-source innovation, and patient activism.

Theorizing How Patients Know

To make private and often invisible knowledge practices visible, to show that what patients know should be regarded as a legitimate form of knowledge “amidst the multiplicity of forms of knowledge within medical practices” (Pols 2013, 82), scholars of social sciences have theorized that patients know as an “activity of knowing in a particular situation” (Pols 2014, 82) and termed this patient knowledge, first introduced by Jeannette Pols.

The first strand of patient knowledge literature I use in my analysis emphasizes that although patients’ knowledge practices are “practical knowing in action” (Pols 2014, 75), they cannot be thought of in contrast to biomedical knowledge, because they draw from the latter. From an epistemological perspective, this way of knowing might be “messy” in the sense that these practices involve “many different techniques, values, and materials” and aim for an immediate purpose (Pols 2014, 75). Pols (2014, 79) retraces this by empirically looking at how patients were using technological aids provided by the clinic at home. With this, she foregrounds that patient knowledge practices involve coordinating and translating knowledge, technologies, and advice from different sources (Pols 2014, 75). This renders patient knowledge practices neither opposite to medical knowledge nor the same (Pols 2014, 79).

Translating different knowledges, information, and techniques from different sources and constantly adjusting, testing, and experimenting with bodies, selves, technological devices, or medication in the care context have been termed tinkering (Mol 2006; Mol, Moser, and Pols 2010; Winance 2010; Kingod 2018). This notion also gives a theoretical tool to understand patients’ practices of “attentive experimentation” (Mol 2006, 411) in order to live with their disease. Tinkering is not a linear translation process but a constant readjusting, refitting, and readapting of care practices. Tinkering has an important collective component, because it involves more than just the individual affected by the disease: it is always done in (care) collectives (Winance 2010, 102). The collective and distributed aspects of tinkering become even more evident with the emergence of patient communities on social media. These online spaces are essential sites for knowledge production (Kingod 2018, 154).

Looping has a significant health-political component to it, which extends to how people know. The second strand of literature on patient knowledge relevant for my empirical analysis concerns activist ways of knowing in the healthcare sphere. Here, patient knowledge practices are conceptualized as actively engaging in shaping biomedical knowledge. Epstein (1995), for example, reveals how “AIDS knowledge” arose out of what he calls “credibility struggles,” between different actors—activists, the media, politicians, and healthcare professionals—and therefore cannot be thought of in strictly biomedical categories. Rabeharisoa, Moreira, and Akrich (2014) coined the term “evidence-based activism” to capture forms of involvement where patient organizations engage, participate, and intervene in biomedical research practices seeking an immediate benefit for the affected. Concepts like AIDS knowledge and evidence-based activism illustrate how activists contribute and create new understandings of their disease, how their knowledge practices can influence and shape formalized biomedical knowledge, and thus account for the epistemic significance of patients within biomedical knowledge production.

These articulations of the epistemology of patients engaging with their bodies and illnesses have allowed a better understanding of the fact that patients know in relevant ways, without hierarchizing different forms of knowing in the medical sphere. I bring theoretical discussions emerged around the Quantified Self (QS) movement to the literature on patient knowledge to better grasp the methodological, innovative, and communal characteristics of loopers’ knowledge practices and their engagement with personal health data. The QS movement can be described as “a large group of avid self-trackers” (Sharon 2017, 95) and is not related to a specific health condition. Members of this movement focus on creating and interpreting data about themselves and correlate behavior with somatic aspects (Heyen 2020; Villa 2012).

Within the QS movement, the practice of gaining “self-knowledge through numbers” (Wolf 2010) involves creating and engaging with personal (health) data, in what is described as “n = 1 self-experimentation.” This points to coexistence of individualization and communality of knowledge practices that I observed in the looper community. On the one hand, “n = 1” emphasizes that it only needs one-person testing and gaining insights from an experiment. Best described by early QS proponent Kevin Kelly (2016, 241) as “you are testing the variable X against the very particular subject that is your body and mind at one point in time. Who cares whether the treatment works for anyone else? What you want to know is, how does it affect me?” (Kelly 2016, 241)

On the other hand, Jethani (2015, 39) has noted that “n = 1 quantified self-experimentation is only made meaningful in the context of a larger audience.” While the QS movement focuses on individual needs, it has a “significant social component” (Crawford, Lingel, and Karppi 2015, 484). Personal data are not just about gaining information about one’s body; instead, it “becomes a medium for connecting with others by offering a raw glimpse into one’s intimate, private life” (Sharon and Zandbergen 2017, 10). Sharing data and participating in the community and meetups are two defining aspects of the QS movement, with personal data being the language everyone can understand and relate to (Sharon 2017, 111; Pantzar and Ruckenstein 2017, 2).

STS scholarship has variously emphasized that it is not possible to make a clear-cut distinction between biomedical knowledge and what patients know, how they engage with their bodies, with self-care technologies, and with disease. Concepts including “know-now” (Pols 2014, 82), tinkering (Mol 2006), or “credibility struggles” (Epstein 1995) emphasize the practical aspects of knowing, translating, and engaging with different forms of knowledge as patients. All these concepts imply a criticism of the binary and hierarchical notion of knowing common in the biomedical sphere. This skepticism about hierarchical knowledge is a useful point of departure for this article. Yet the contexts and operations through which loopers know are very different from clinical and established healthcare settings and thus lead to new questions. Loopers’ knowledge practices are more closely related to social media communities, such as the QS movement, innovation spaces, and the open-source sphere. By bringing together these different strands of literature, we can account for the complexities of the knowledge practices that emerge when patients leave behind their provided self-care technologies and venture into an open-source innovation endeavor.

Methods and Setting

This article draws on ethnographic fieldwork carried out in the German looper community in 2018 and 2019, and the regional setting influences people’s experiences. Germany is a wealthy Western European country with universal health coverage. And while the cost of much diabetes technology is covered, Germany has also been slow in adopting a digital health infrastructure, compared to other European countries (Bertelsmann Stiftung 2019, 4). For example, Germany is only now discussing electronic patient records. In my fieldwork, I did not remain in one geographic site (Marcus 1995) but followed different actors and stories (Latour 2005, 12). I started my empirical research by participating in a hackathon, and at the same time, I met my gatekeeper, an active community member. I later joined him at a local looper meetup, in which I frequently participated thereafter. I always introduced myself as a researcher and disclosed my research interests (Spradley 1979, 58). I also conducted twenty-eight problem-centered interviews (Reiter and Witzel 2012), predominantly with people who would be considered patients, their relatives, and their healthcare professionals. Participants’ age ranged from twenty-two to eighty-one. They had different educational backgrounds, with more or less experience in technology-related subjects, although most participants had a higher education level. ³ Six participants were healthcare professionals, including two who had no T1D. To better understand public and broader negotiations about looping, I included media reports, blog posts, and statements from regulatory bodies in my analysis. Data gathering and analysis were mutually informed, following an iterative logic. Data collected early on in the analysis served as a starting point to direct the strategy of theoretical sampling (Clarke 2015, 101f).

Discussion: Elaborative Tinkering

In this article, I empirically mapped out three patient knowledge practices (technical, recursive, and methodological) that emerge when people with T1D use and develop an unauthorized automated system for their illness-related repetitive data work. The epistemologies of going beyond commercial self-care technologies and venturing into this open-source endeavor highlight an entanglement of different forms of (patient) knowledge practices that have not yet been addressed in the literature about patient knowledge.

Pols (2013, 2014), Mol (2006), and others working in their tradition note that what patients do and how they bring together different sources, techniques, and knowledges to live with a disease should be respected as knowledge among other forms of knowledge in the biomedical sphere. On the other hand, Epstein (1995) and Rabeharisoa, Moreira, and Akrich (2014) focus on how patients know in activist ways. They documented affected people self-organizing, being vocal in their demands and criticism of established healthcare structures/actors, and challenging and influencing (medical) understanding of their health condition.

These different concepts from STS are a lens to look at loopers’ knowledge practices, which have elements of tinkering, as described by Mol (2006) and others, as well as practical “know-now” (Pols 2014, 82) and can also include more activist forms of knowing with a health-political aim. Looping is firmly situated in the realm of open-source innovation. The fact that loopers are experimenting with their bodies in creative, critical, and vocal ways—stressing the importance of personal health data—brings their knowledge practices close to the QS movement. These accounts of how patients know add nuance to descriptions of patient knowledge.

Yet developing and using life-saving self-care technologies while being involved in open-source innovation and activism go beyond tinkering and know-now. With loopers, we are somewhere in between these notions of patient knowledge: loopers are demanding, they want to be recognized, they are vocal, and managed to create “actually existing alternatives” (Kelty 2008, 3) to the commercial systems. Loopers’ knowledge practices are technically advanced to a degree that is not conceived of in the current literature on patient knowledge. They know about technical details that even people trained in T1D technologies do not necessarily know. Data repair work (Pink et al. 2018); maintaining devices on, in, and with bodies; keeping data flowing; or navigating the open-source platforms are some of loopers’ common knowledge practices, even though they are not part of the established care structures of living with T1D. Loopers experiment with their bodies and self-care technologies, and they engage in innovative ways in a community outside of the clinic, on social media and in the open-source sphere. In the context of looping, knowing becomes a recursive practice of methodologically experimenting with bodies, self-care technology, algorithms, and digital infrastructures—all while staying in feedback loops with others in the community. The focus of the looper community is always directed at establishing a knowledge base and epistemic legitimacy. Looping becomes a continuous moving between deeply individual practices and inherently communal challenges that cannot be undertaken alone. It connects self-care practices with broader negotiations and tensions in the healthcare sphere regarding expertise, credibility, and knowledge legitimation. Similar to the QS movement, loopers do not see technologies and health programs provided by the “traditional” and standard healthcare regime as fitting or good enough, so they take matters into their own hands (Sharon 2017, 109). They push device manufacturers and pharmaceutical companies like Epstein’s (1995) “treatment activists.” However, with the establishment of the OPEN project, loopers are beginning to enter (clinical) academic spaces, not invited as “citizen scientists,” but by using their diverse professional backgrounds. Evidence is produced within a research project, with the possibility of formalizing what loopers know. Unlike other forms of patient activism, loopers do not stay in the biomedical and healthcare realm, instead moving in and out of healthcare/clinical and open-source innovation spaces, challenging received ideas about patients’ knowledge practices.

Loopers’ knowledge practices are complex and multifaceted and, I suggest, best understood as forms of elaborative tinkering. Downloading instructions, maintaining a technical system that immediately can improve one’s own “chronic living” (Wahlberg et al. 2021), or the engagement in online and offline peer communities for support are all practical ways of knowing and tinkering with self-care arrangements. Loopers’ knowledge practices are closely related to pushing and criticizing commercial device manufacturers and regulatory bodies, deliberate scientistic self-experimenting for “the greater good,” and being engaged in building a knowledge base for epistemic legitimacy. Loopers are engaging within the open-source innovation sphere and using and developing this automated system with the aim to improve one’s personal life with the chronic health condition while at the same time having in mind that they need to generate evidence that these automated systems are efficient.

Loopers create complex knowledge and make it explicit by sharing it with others in methodological, organized, and scientistic ways. Their emerging knowledge practices reveal loopers radically changing their own treatment, while also challenging “the political economy of health device innovation” (Jansky and Langstrup 2022, 18). If we stay only within the literature about how patients know in practical ways, or if we only see the emerging knowledge practices of the looper community through the prism of activist ways of knowing as patient, or as niche self-experimenting as patient innovators in the realm of the QS movement, we lose sight of the nuanced entanglement of these ways of knowing in practice. Focusing on elaborative tinkering captures how knowing as a patient in these spaces of innovation, activism, and digitized (chronic) self-care can and is a constantly moving between and combining different ways of knowing. Understanding that patient knowledge practices in the looper community and similar health communities are made up of different ways of knowing offers a way not to lose sight of the fact that intimate practices of tinkering with self-care arrangements and “know-now” (Pols 2014, 82) can also be entangled with activist and innovative ways of knowing, and how both might be interrelated. With an increase of digitized, personalized, and data-driven healthcare settings, where technologies do not only “travel…from the clinic to the home” (Langstrup 2005), but from open-source innovation spaces to the home (and later to the clinical context), we will see more and more that patients can know in technically advanced, recursive, and methodological and scientistic ways. Turning the gaze toward elaborative tinkering allows us to carefully explore when and how different patient knowledge practices might come together and when they are kept apart.

Conclusion

Although the case presented in this article is rather specific, from a history of high involvement in one’s care in T1D to the uniqueness of the technology and community, similar technologically advanced, recursive, and methodological ways of engaging with technologies can be observed in other healthcare settings, such as sleep apnea (Schultz 2019), epilepsy (Bogataj et al. 2021), and hearing loss (O’Kane et al. 2019). These communities might not be as organized as the looper community currently is—yet the proposed notion of elaborative tinkering may help other researchers to gain a deeper understanding of the complexities of knowing as a patient at these intersections of digitized and personalized health care, open-source innovation, and patient activism. The empirical reports above show that patients can have highly complex and sophisticated technical knowledge about an innovation that far exceeds current commercial self-care technologies. I have also illustrated that healthcare professionals can draw upon their own illness-related experiential knowledge, combined with biomedical knowledge. How this can unfold in other areas of increasingly digitized, data-driven, and personalized healthcare settings calls for more empirical exploration.