Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Precision medicine and digital phenotyping: Digital medicine's way from more data to better health:

Precision medicine and digital phenotyping: Digital medicine's way from more data to better health: Precision medicine and digital phenotyping are two prominent data-based approaches within digital medicine. While pre- cision medicine historically used primarily genetic data to find targeted treatment options, digital phenotyping relies on the usage of big data deriving from digital devices such as smartphones, wearables and other connected devices. This paper first focusses on the aspect of data type to explore differences and similarities between precision medicine and digital phenotyping. It outlines different ways of data collection and production and the consequences thereof. Second, it shows how these sorts of data influence dominant beliefs in the field: The field of precision medicine relying on the dominant understanding of ‘genetic determinism’ imported from genetics, digital phenotyping building on the logic of ‘data fundamentalism’. In the end, the analysis shows how digital data informs potentials as well as challenges of precision medicine and digital phenotyping: a better health care for (some) individuals connected with individualisation and respon- sibilisation for all, with a prognosed shift from reactive to preventive medicine. Additionally, data-based approaches might facilitate epistemological and ontological redirections for the whole field of medicine that will also affect knowledge pro- duction and a reassessment of the value of different types of knowledge (quantifiable vs. non-quantifiable) with all its con- sequences. Institutionally, it might lead to shifts in distribution of power to experts in big data related technologies, i.e. private companies. Keywords Personalised medicine, genetic determinism, data fundamentalism, data-driven medicine, big data, digital health This article is a part of special theme on Digital Phenotyping. To see a full list of all articles in this special theme, please click here: https://journals.sagepub.com/page/bds/collections/digitalphenotyping only solve major healthcare challenges, but could shift the Introduction whole medical field from responsive medicine to predictive Precision medicine (PM) and digital phenotyping (DP) are medicine (Hood et al., 2012). DP seeks to revolutionise buzzwords in medicine. The promise they contain is already mental health by collecting continuous ‘real life data’ expressed in their naming. PM aims to offer a precise and from digital devices of everyday use such as smartphones targeted treatment for an individual or a group of people and wearables. With this additional data, knowledge gaps based on clinical data. DP promises to categorise people in mental health diagnosis could be closed without burden- into clinically relevant categories for the treatment of ing patients with further tasks (Huckvale et al., 2019). PM mental diseases, with the help of digital data. Both promise revolutionary shifts in medicine and healthcare. Both approaches are data-driven, have been developed University of Tübingen, Center for Gender and Diversity Research only in recent decades and hold out the prospect of chan- (ZGD), Tübingen, Germany ging the field of medicine on the basis of digitisable data. Corresponding author: They are both said to transform medicine as a scientific Renate Baumgartner, University of Tübingen, Center for Gender and field and as a healthcare discipline. PM claims to offer indi- Diversity Research (ZGD), Brunnenstr. 30, D-72074 Tübingen, Germany. vidualised treatment for specific diseases which could not Email: renate.baumgartner@uni-tuebingen.de Creative Commons CC BY: This article is distributed under the terms of the Creative Commons Attribution 4.0 License (https:// creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access page (https://us.sagepub.com/en-us/nam/open-access-at-sage). 2 Big Data & Society and DP, being data-based, require specific statistical and might look like an odd choice, since DP is practically a computational expertise, which results in new stakeholders part of PM, while PM is recently described as an approach entering the field of medicine. to collect as much varied data as possible to compose a Data today is said to be the new gold. Many publications patient’s health map. Historically, however, big parts of deal with the datafication of our lives as well as with the dis- PM predominantly deal with genetic and genomic data, tinct influence of data not only being collected and aggre- while DP is till date firmly grounded in mental health gated but also analysed by algorithmic means (Lupton, issues. Also, different stakeholders seem to concentrate 2018b; Ruckenstein and Schüll, 2017; van Dijck, 2014). on different types of data or different disease types Technical breakthroughs around data saving, transfer and (Mindstrong, 2021; Prainsack, 2017). Contrasting PM analysis within the last 20 years have made it possible based on genetics with DP dealing with mental health that approaches like PM and DP have a technical basis to conditions, we see the similarities as well as the differ- even be thought of and developed. Datafication and an ences within the working logic in a field that tends to introduction of algorithmic tools such as artificial intelli- mental illnesses versus a field that predominantly gence (AI) have taken place in different fields including deals with illnesses experts ground in the realm of the politics, economy, and law, but significantly also in health body. For this endeavour, I will focus on the following and the field of medicine (Topol, 2019). examples: breast cancer prediction, diagnosis, and treat- Digitalisation in medicine already started in the 1970s, ment as an example of PM based on genetics, and the then by the name of expert systems which were used to digital platform Mindstrong which claims to treat mental get diagnosis and treatment recommendations (Metaxiotis health issues such as depression, bipolar disorder and and Samouilidis, 2000). Over the last 15 years datafication, PTSD basedonDP. or ‘the conversion of qualitative aspects of life into quanti- fied data’, has picked up (Ruckenstein and Schüll, 2017: Precision medicine 261). The process began with more and more patient infor- mation finding its way from paper records to electronic PM is a medical approach that is best known for aiming to patient files, including the involvement of new techniques offer a specifically targeted treatment regime to a person or with digital output, e.g. in radiology. However, when a certain group of people, ideally leading to a ‘personalised’ people recently started using smartphones and wearables, treatment for each person, which in the end could declare engaging in digital communities, thereby producing also common disease labels obsolete (Ferryman and Pitcan, health-related data that can be collected, shared and ana- 2018; Prainsack, 2017). Other terms used for this approach lysed, datafication picked up momentum (Ruckenstein in medicine are ‘personalised medicine’ or ‘stratified med- and Schüll, 2017). Thus, both technical breakthroughs icine’. The terms have appeared and been used in different and sociocultural developments such as wearable usage historical, geographical and disciplinary contexts. I will use and sharing health-related experiences on social media, the term ‘precision medicine’ because it is the more recent e.g. as part of the quantified self-movement, made it possi- term and for the sake of simplicity (Erikainen and Chan, ble for data-driven methods to also win ground in medicine 2019; Ferryman and Pitcan, 2018). Since 1990, PM has and healthcare (Lupton, 2018b; Ruckenstein and Schüll, also been a buzzword and the closely related field of 2017). genomic science has been able to acquire considerable All these processes have resulted in PM and DP not just funding. An example is the human genome project (HGP) remaining a vision but becoming reality and constituting or the ‘all of us’ project in the United States (Cooper and two of the most spoken about data-driven approaches in Paneth, 2020; Ferryman and Pitcan, 2018), both with enor- medicine till date. Still, the two approaches have had his- mous expenditures. The ‘all of us’ project will reach total torically distinct starting points. They use different types costs of about 1 billion US$ once finalised (Cooper and of data which is based on different ways of data produc- Paneth, 2020). PM represents an interdisciplinary endeav- tion. This paper will explore what PM and DP has in our combining biology, medicine, informatics, computer common and what separates them based on the type of science, mathematics and statistics (Erikainen and Chan, data they use. In the end, it will show how this data-driven 2019). Computer science, mathematics and statistics func- aspect of PM and DP results in similar logics within both tion to store and analyse data. The collaboration between approaches, whether in dealing with mental health condi- human molecular genetics with modern computer science tions or diseases based in the realm of the body. First, both since the 1950s has been a thriving force to establish PM approaches will be analysed according to the sort of data (Cooper and Paneth, 2020; Ferryman and Pitcan, 2018). they use. In the second part, I will outline the logic The most prominent desiderata of PM for the field of med- behind the main arguments in the field, also based on icine are disease understanding and prediction with a strong these types of data. In the end, I will conclude with the focus on cancer, aiming to find a suitable treatment with challenges PM and DP might pose due to their role as low incidence of adverse drug reactions (Cooper and digital data-based approaches. Comparing PM and DP Paneth, 2020; Prainsack, 2017). Examples are targeted Baumgartner 3 breast cancer treatment or ‘personalised’ antiretroviral treat- (Barnett et al., 2018; Dagum, 2019; Jain et al., 2015). There ment for HIV-positive individuals based on treatment opti- is a lot of ongoing research trying to tackle different needs, misation tools (Kumari et al., 2017; Low et al., 2018). Data e.g. the RADAR-CNS project from European Union aiming about genetic predisposition, lifestyle information, clinical to develop an open-access platform around mobile health data, etc., all of which has to be ‘structured, digital, quanti- data, and initiatives to provide mental health treatment by fied, and computable data’ is combined creating a ‘unique smartphones to low-income populations. However, much thumbprint’ of a person to inform their diagnosis and treat- of the research is still in an experimental stage (Melcher ment, also called ‘personal health maps’ (Prainsack, 2017: et al., 2020; Ranjan et al., 2019). DP is currently used 3–4). The best-known examples till date, however, mostly in the realm of mental illnesses. However, thinking involve genetic and genomic data. of DP as part of a PM that gathers all sorts of data to compose a personal health map, it could be used for all types of diseases (Huckvale et al., 2019; Prainsack, 2017). Digital phenotyping As far as PM aims at incorporating all sorts of data Psychiatry got more interested in PM since struggling for a for diagnosis and personalised treatments, DP seems to be long time under the lack of objective markers for mental the part of a personalisation which Rüppel describes health conditions that science finds are a key basis for being ‘increasingly rearticulated as ‘Big Data’ project’ correct diagnosis and optimal treatment. Under the name (Prainsack, 2017; Rüppel, 2019: 593). DP and PM are of ‘personalized psychiatry’, and other similar labels scien- both data-driven approaches aimed at understanding tists aimed to identify genomic or molecular biomarkers of diseases and offering targeted treatments, only the data mental health conditions to be able to offer targeted treat- they use can vary from biological, genetic and genomic ments to individuals Rüppel (2019: 593) describes being data to sensor-collected data from digital devices such as ‘increasingly rearticulated as ‘Big Data’ project’ smartphones and wearables (DP). (Prainsack, 2017). DP is an advancement of this search for biomarkers that includes data collected by digital Different sorts of data – different devices with sensors used in everyday life in its assessment consequences? to identify ‘digital biomarkers’ (Dagum and Montag, 2019: 14). DP goals are finding ‘(digital) diagnostic markers’ for This section will focus on an analysis of the distinct sorts mental health conditions by correlating a collection of of data used in PM and DP, based on the cases of breast sensor data and self-reported data and (thereby) monitoring cancer treatment for genetics-based PM as an example of and predicting mental health statuses based on these bio- disease grounded in the biological and the platform markers, and behavioural phenotypes (Birk and Samuel, Mindstrong for DP as an example of mental health. In the 2020: 1874; Dagum and Montag, 2019). The idea of DP following, I will point out how the type of data influences or the digital phenotype was raised by two groups of scien- the way the dominant beliefs within the fields of PM and tists, Jain and colleagues and Torous and colleagues at the DP work. same time, both referring to the usage of large amounts of digital data to find patterns in human behaviours and Type of data and data collection traits, linking those to ‘disease phenotypes’ (Birk and Samuel, 2020: 1876). The assumption is that mental Ferryman and Pitcan (2018: 3) define PM based on their health conditions show themselves in digital traces of a ethnographic research as ‘the effort to collect, integrate, person and ‘behaviour-expressed symptoms’ can thus be and analyze multiple sources of genetic and non-genetic identified as ‘behavioural phenotypes’ (Dagum and data, harnessing methods of big data analysis and Montag, 2019: 14; Garcia-Ceja et al., 2018). DP aims to machine learning, in order to develop insights about gain data from ‘naturalistic settings in-situ, leveraging the health and disease that are tailored to the individual.’ The actual real-world’ and was only possible once digital data involved is genetic data and a variety of other data devices such as smartphones and wearables were ubiqui- including clinical data and lifestyle data (Prainsack, 2017). tously available (Birk and Samuel, 2020: 1876). One method of data accumulation within PM is to start Examples of DP are the assessment of the lacking efficacy with data collection just for the means of having a large of lithium for individuals with ALS in slowing down data pool. The HGP would be an example of such an disease progression, assessed through the analysis of approach. Human DNA data began to be collected with online disease communities (Jain et al., 2015). Other exam- the promise that it would inform solutions to health ples are the analysis of insomnia-related tweets, and moni- issues. The project started in 1990 with the goal of toring and prediction of (clinically) relevant outcomes in mapping the entire sequence of human DNA. The main people with mood disorders such as major depression, rationale behind it was to find new insights into human bipolar disorder, and schizophrenia based on sensory mea- health and disease, which helped to secure a huge amount surements but also on peoples’ participation in social media of funding for this project (Ferryman and Pitcan, 2018). 4 Big Data & Society The other approach is to conduct genome and sample ana- data can be used to predict depressive symptom severity lysis within pathology for specific targets. The examples on and might therefore serve as a biomarker for depression. In which I will focus involve the identification of specific summary, DP is based on the digital measurement, collec- genetic characteristics in the genome of women with a tion, analysis and interpretation of enormously varied activ- family history of breast cancer or the pathological sample ities to enhance understanding and treatment of mental of breast cancer patients, to decide upon targeted forms of health conditions. Related data is collected continuously, therapy. People with inherited BRCA1 and BRCA2 gene live and in situ within ‘real-life settings’ (Birk and Samuel, mutations have an increased risk of developing breast 2020). ovarian cancers and should follow therefore more rigid pre- ventive measures. Oncogenic human epidermal growth Data producers factor receptor (HER2) positive breast cancer patients profit from treatment with trastuzumab and lapatinib that Within the genetic part of PM, the processes of data produc- specifically target HER2, the receptor regulating ‘cell tion take place in highly regulated and controlled environ- growth, proliferation and differentiation’ (Low et al., ments. Following the example of pharmaceutical labs, 2018: 502). Additionally, genomic research in breast those facilities need to prove that they follow specific pro- cancer is still ongoing to identify other targets. In general, tocols even to be granted permission to produce the data. this genetic data is analysed, i.e. ‘produced’, at specific The adherence to the protocols has to be assured through times by specific experts in health institutes or external labs. different measures such as quality control, quality assurance Conversely, DP works with digital data collected and audits (National Human Genome Research Institute, through the usage of smartphones, computers, tablets n.d.; U.S. Government, 2021). The facilities are frequently and wearables, such as ‘FitBit’ or the ‘Smartwatch’,and audited regarding their adherence to these protocols. other devices with digital sensors. A variety of hardware Additionally, special devices used by trained expert person- and software sensors come into play collecting data such nel, e.g. lab technicians, are required to obtain the data. The as inertial sensors measuring walking speed, physiologi- data can only be produced by people with specialised train- cal sensors measuring heart rate and ambient sensors mea- ing (lab technicians) who have access to certain facilities suring temperature (all hardware sensors); but also (labs). However, it is not just laborious but also expensive software sensors that measure internet activity, social to produce this data. The necessary facilities and lab equip- media presence, typing speed, etc. (Birk and Samuel, ment are costly and so it is to follow protocols. Large teams 2020; Garcia-Ceja et al., 2018). DP can be divided into of trained personnel are needed and have to be paid to behavioural phenotyping and digital biomarkers (Dagum develop and adhere to protocols and assure their quality, and Montag, 2019). Behaviour phenotyping uses digital and are able to hold the approval status for the whole facility information such as location, physical activity, mood, (U.S. Government, 2021). As a result of the speech patterns, typing speed and call activity, but also resource-intensive and high-priced nature of data produc- social media usage and search terms to search for tion, somebody has to bear the costs and the expenditures ‘behaviour-expressed symptoms’ (Dagum and Montag, are sometimes regulated and tried to be kept at a 2019: 14; Garcia-Ceja et al., 2018). It uses, e.g. passive minimum (Low et al., 2018). In the example of breast sensing data of GPS location and call logs as a proxy cancer, testing of the genetic properties of a patient and for behaviourally phenotyping loneliness (Birk and their sample is done once (Cedars-Sinai Blog, 2019). To Samuel, 2020). Digital biomarkers aim at measuring summarise, the production of genetic data is initiated ‘trait and state changes in neuropathology that can be indi- actively, conducted at specific time points, and is a labor- cative of disease risk, disease onset, disease progression or ious and expensive process. recovery’ (Dagum and Montag, 2019: 14). Mindstrong, For DP, the collected data is produced and shared by e.g. claims to have identified ‘a set of digital biomarkers users of digital devices, whose live and local data can be from human-smartphone interactions that correlate collected continuously. The ways of collection are enor- highly with select cognitive measures, mood state, and mously varied, depending on the digital device and brain connectivity’ that can be related to depression, sensors used (Birk and Samuel, 2020; Garcia-Ceja et al., anxiety, negative, positive affect and other mental condi- 2018). Often this data is collected as a by-product of prac- tions (Mindstrong Health, 2021). Assessments tices, such as typing speed, click behaviour and not collect- Mindstrong uses, e.g. in the ongoing AURORA study ing personal data, as in the example of Mindstrong. on PTSD are ‘continuous-time accelerometry data, key- Compared to the actively initiated aspect in PM, this data stroke characteristics, time and duration of phone calls, is provided actively (when self-reported) as well as pas- time and character length of text messages, text words/ sively (e.g. when cell phone usage is monitored) by the symbols used, time and number of emails, smartphone users. Prainsack (2017: 21) calls the users of the devices screen time, and intermittent GPS data’ (McLean et al., ‘prosumers’, a combination of producers of data and consu- 2020: 5). Saeb et al. (2016) claim mobile phone location mers of information. Most of this data falls under the term Baumgartner 5 big data, which has been described with properties as could be explained with the help of genetics. This ‘volume, velocity, variety, exhaustive in scope, resolution, mindset is called ‘genetic determinism’ (Peters, 2012). relational and flexible’ (Kitchin, 2014: 1; Rüppel, 2019). Cooper and Paneth (2020) point out that in genetically This means that it entails huge amounts of data with a based PM, genetic determinism prevails. Weiss (2017) high granularity. The aspect of continuous real-life data col- also calls this logic ‘mendelian fundamentalism’, which is lection is named as one big advancement of DP compared an even more drastic description. Both are based on the to other diagnostic tools for mental health conditions assumption that living beings are their DNA, or differently (Dagum, 2019). Big data is simultaneously described as put ‘Genes R Us’, and it is thus possible to find solutions precise, because it is produced close to a person and in against illnesses knowing about the genetics behind real life, and as messy or unclean because it might have (Peters, 2012: 10). Also, if all information and logic of been collected for other purposes and in an uncontrolled the DNA are revealed, it would be possible to understand environment with sensors that might not be equipped to how the body (and the mind) work (Peters, 2012). collect data in research quality (Huckvale et al., 2019; Similarly Weiner et al. (2017: 989) speak of a ‘genetic Kitchin, 2014; Torous et al., 2016). There is a whole indus- imaginary’ being invoked where the biological is seen as try dealing (in the double sense) with health-related data key for understanding diseases and a molecular understand- called the data industry (Cosgrove et al., 2020; Prainsack, ing of diseases is supposed to lead to a new type of medi- 2017). These companies offer services around data cine. Also this logic has its roots in genetics’‘molecular saving, collecting, cleaning, trading and analysis (Kitchin, vision of life’ (Weiner et al., 2017: 999). The imaginary 2014). The producer of the data (or ‘prosumer’) is every ‘is being continuously remade and rearticulated’ and gets individual using digital technologies like smartphones and actualised as new hopes are being constructed based on wearables, or other connected devices with sensors new biotechnologies such as gene editing, and the develop- (Garcia-Ceja et al., 2018; Prainsack, 2017). Additionally, ment of biological drugs (Milne, 2020: 103). Ultimately, any person who interacts on an online platform such as the imaginary works for the persistence of the cultural Facebook and Twitter or is simply conducting an online power of genetics. However, in focussing on technological search provides data for this corpus of data. inventions, it is obscuring the social determinants of health The focus on the different types of data used so far in PM and illness (Milne, 2020). based on genetic data and DP reveals the following differ- Erikainen and Chan (2019: 320) showed how the choice ences between PM and DP: (1) the type of data being lab- of rebranding ‘personalised medicine’ that happened in the generated data versus sensor-derived data from ubiquitous U.S. policy context fell on ‘precision medicine’ also available digital devices; (2) data being collected rarely because the adjective ‘precision’ has an ‘ethically neutral versus moment by moment, live and in situ; and (3) the pro- or even positive’ connotation. I suggest that another essen- ducers being professional labs versus users of digital tial argument within the field is the very aspect of ‘preci- devices. This next section will now show how these charac- sion’, which as a characteristic not only seems to speak teristics of data influence the logic within each of the two for the possibility of choosing a targeted treatment but sur- approaches. rounds genetic data like an aura. It is the conception that genetic data is precise: first because of the effort and the ways how it is measured, second because it is conceptua- Logic within the field lised as being the basis of all living beings and third Drawing on Annemarie Mol’s ‘logic of care’, I will use the having the image of being all-encompassing and everything term ‘logic’ for a persisting rational in the field or a style we have to know to base our decisions on; the last two argu- that seems to be appropriate (Mol, 2008: 1). One could ments being firmly rooted in genetic determinism and rear- say the logic within the field is also a certain dominant ticulating the genetic imaginary (Keller, 2002; Peters, 2012; rational existing in the field which is not questioned. This Weiner et al., 2017). It seems like these aspects give the logic within the field seems to have different nuances whole field more credibility, and as if the truth is to be based on the type of data used in the field. Historically, found in the details, and the closer we look, the closer we genetic or genomic data was used for PM, which was first get to understanding it (Erikainen and Chan, 2019; Keller, expanded to clinical data in general and resulted nowadays 2002; Weiss, 2017). The hope to find more molecular in the usage of a whole variety of data that is helpful when markers for breast cancer through genomic profiling is composing a personal health map (Prainsack, 2017). As I one example of this logic (Low et al., 2018). This argument have pointed out, the most prominent examples of PM till fits very well into the scientific worlds of medical science date involve genetic or genomic data. With the establish- and biology, which operate under positivist assumptions. ment of genetics as a key discipline within biology, a It is the mechanistic assumption that understanding a certain dominant mindset was also established, driven by process detailed enough will reveal the truth behind it. the assumption that every question regarding the functions Based on this assumption, analysing the problem more pre- of living organisms including human health and disease cisely will lead to better suitable treatment options simply 6 Big Data & Society because the premises were more precise and thus closer to Huckvale et al., 2019; Jain et al., 2015: 463). The Research the truth. Hopman (2020: 425) in her study on forensic Domain Criteria Initiative (RDoC) in the U.S. aiming at DNA identifies the logic of accuracy within “the search establishing a new classification system based on ‘basic of the uniqueness of the individual”. In a search for the science’ for mental health conditions is an institutional genetic uniqueness of people, this logic results in the accu- step in this direction (Rüppel, 2019: 571). The continuous, mulation of data. It functions also as a ‘logic of expansion’ in situ and live monitoring of individuals’ behavioural that constantly searches for new genetic ‘“territories” to activities collected as big data are stylised as the missing map’ (Hopman, 2020: 428–429). In this constant search piece in understanding mental health because they can be for more data, it permanently has to attract more money. collected continuously and close to real life (Huckvale The logic resembles very much the search for the ‘unique et al., 2019; Jain et al., 2015). At the same time, current thumbprint’ to compile personal health maps Prainsack diagnostic practices for mental illnesses relying on self- (2017: 3) speaks about within PM. Also, PM is constantly reporting of symptoms are framed as not conclusive aiming to attract more money, which will seem to be well enough for diagnosis (Huckvale et al., 2019). One quest invested as ‘precision’ implies PM will make medicine of DP is the search for digital biomarkers of a field fru- ‘more effective, and thus also cheaper’ (Prainsack, 2017: strated by a long unsuccessful search for biological 79). markers (Birk and Samuel, 2020; Brietzke et al., 2019). A Several points of criticism can be raised to counter this variety of data is used, aimed at understanding mental logic. First, experts in genomic medicine reject that ‘the health diseases better. This is framed as having the potential more we learn about the genome, the more distant it not only to guide new ways of measurement and treatment seems to be from a role as a causative agent in most wide- but also to change the classification of diseases and thera- spread diseases’ and instead acknowledge the role of peutic measures (Burnett, 2015; Huckvale et al., 2019; genomic medicine as ‘a way to do science, not medicine’ Jain et al., 2015). This is interesting when looking at differ- (Cooper and Paneth, 2020: 67). Medicine based on ent conceptualisations of big data. Within some fields like genetic knowledge might be helpful for diseases firmly data science, big data is still regarded as unclean and grounded in genetics, however, not for other diseases. messy and there exists a whole industry that focuses on Thus, for multifactorial diseases, other measures such as practices of ‘cleaning’ and preparing big data for analysis. public health seem to be more effective, even if they may Stakeholders who are pro-DP mark common practices of not have the connotation of ‘precision’ and ultimate truth disease identification like interviews between patients and (Cooper and Paneth, 2020; Weiss, 2017). Second, critical doctors as subjective or at least not sufficient and at the social science perspectives would challenge the idea that same time see digital device collected data as the new a solution can only be found based on mechanistic princi- Holy Grail. Thus, it seems as if lived experiences (or ples. As we have seen in the genetic imaginary argument, digital traces of everyday practices), once they are collected they privilege technical solutions over taking into account through digital devices which are able to provide them con- systemic implications and social determinants of health. tinuously, in situ and live, gain in worth also because it was Those principles may be important for reasoning in the a digital device which gathered them and because they are field. However, decisions in the field of medicine seem to collected close to the individual in ‘real-life settings’ (Mau, be far more complex, entail much more information and 2017; Quinn, 2021). Self-tracking from digital devices is rely on manifold practices as studies, e.g. in medical anthro- framed as providing trustworthy data in contrast to the indi- pology, show (Kim et al., 2018; Spinnewijn et al., 2020). vidual body’s perception which is marked as untrustworthy For DP, the logic behind the argument is distinct but or at least not reliable enough to be the sole ground on arrives at a similar conclusion. Generally for DP, the fol- which diagnosis should be based upon (Lupton, 2015). lowing goals are expressed: early disease detection and – I propose that this way of conceptualising data and gues- surveillance, identifying and incentivising healthy behav- sing it would change a whole field can be termed as ‘data iour, developing new, more targeted interventions and treat- fundamentalism’, a term coined by Crawford (2013). Data ment strategies (Jain et al., 2015). Additionally, fundamentalism is defined by ‘the notion that correlation stakeholders claim that DP will be ‘providing a more com- always indicates causation, and that massive data sets and prehensive and nuanced view of the experience of illness, predictive analytics always reflect objective truth’ [because] an individual’s interaction with digital technolo- (Crawford, 2013). Additionally, digital traces are concep- gies affects the full spectrum of human disease from diag- tualised as (digital) ‘biomarkers’ to use a common nosis, to treatment, to chronic disease management’ (Jain concept known in science. Accordingly, Mindstrong’s et al., 2015: 462). Several experts within DP expand their homepage indicates ‘[T]o identify the digital phenotyping hopes towards a description of how this additional knowl- features that could be clinically useful, Mindstrong used edge gained through big data analysis will also lead to powerful machine learning methods to show that specific changes in classification and diagnosis and treatment of dis- digital features correlate with cognitive function, clinical eases ‘in ways that matter most to patients’ (Burnett, 2015; symptoms, and measures of brain activity in a range of Baumgartner 7 clinical studies (Mindstrong Health, 2021).’ Thus, the case information can indeed be labelled as ‘(digital) extended of DP seems to be a combination of data fundamentalism phenotype’ because it may actually show health-related and biologisation of digital traces. It correlates digital fea- conditions of a person and additionally bears traces of tures to states of health and illness (‘digital biomarkers’) how the user and other people retroact with it. His which have big data as a prerequisite and thus, comes to framing strengthens the aspect that it is also ‘a collective the conclusion that those ‘reflect objective truth’. creation, involved in social feedback loops’. Lupton However, so far no digital biomarker merits the scientific claims that the ‘human-data assemblages’ connect body definition of a biomarker, i.e. ‘a characteristic that is objec- and data, but that people make their data, as well as data, tively measured and evaluated as an indicator of normal makes people (Lupton, 2016, 2018b: 1). Still, these ‘data biologic processes, pathogenic processes, or pharmacologic bodies’ are ‘lively’ in that they are ‘unstable and generative’ responses to a therapeutic intervention’ (Dagum, 2019: 14). and ‘lead a life of their own’ (Lupton, 2018b: 2; Mager and Different scientists working on DP come to the conclusion Mayer, 2019: 95, 98–99). These data doubles ‘are not inno- that digital biomarkers are objective, and ‘reliable informa- cent or innocuous virtual fictions […]. They have ethics, tion on each individual’s behaviour’ (Brietzke et al., 2019: politics’ and are thus, ‘no longer “doubles” […] but they 223; Huckvale et al., 2019; Jain et al., 2015). Insel, e.g. are intrinsically interwoven with bodily practices and bio- compares the possibility of monitoring brain function by politics’ (Lyon, 2003: 27; Mager and Mayer, 2019: 99). DP to a ‘continuous glucose monitor in the world of dia- Companies like Mindstrong might have less interest in betes’ (Metz, 2018). These statements show how medical tracing back data to one particular individual but more in experts in the field stylise DP to provide real and reliable aggregating data about a collective to gather more data information about peoples’ behaviour that can be correlated for digital biomarkers. Still, the logic within the field of to mental health conditions. They go even further when DP works in the way that the continual flux of ‘real life claiming they could indeed predict mental health status, data’ gives us access to ‘real life’, so to say to a deeper e.g. relapses into depression, before individuals themselves truth or a truth behind it all, which is an authenticity and professionals would be able to know about it (Dagum, claim. This seems to be equivalent to being closer to 2019). They conceptualise big data on behaviours as a people. Critical data studies rather suggest that data traces reflection of lived experiences. Actively and passively gen- and the individual behind retroact with each other and erated data from digital devices of people is framed as differ according to who is zooming in, how and with resulting in scientifically measured, analysable proxies for which epistemology and ontology (Grommé, 2016; Loi, mental health and thus hold objective (and thus ultimate) 2019). truth. I would like to add that it is not just about the This assumption of having data being produced continu- amount of data (which is usually important in big data) ously and so close to the subject of interest (and thus, tied to and the collection by a digital device, but data that has the logic of accuracy) upholds the illusion of knowing a been collected close enough to the subject and so continu- deeper truth about this very subject. It claims to have ously that it seems as if it would be a direct window to access to all the relevant information about the data peek into real life. subject needed for certain assessments. This seemingly Looking into the depth of the type of data and in line exhaustive knowledge, continuously streamed from real with other authors I propose that for DP, truth is thought life and conceptualised as objective is the basis of the to be found in peoples’ behaviours: what time they get assumption that we will be better in deciding which treat- up, when and how long they sleep, when and in what ment options are ideal, even more so, we will be able to way they take part in social media activities, how often classify disease categories anew. In contrast, a critical the text/call, how much they move, etc. From a social social scientist perspective remarks that ‘data doubles’ science perspective, it is the digital traces of people that lead their own lives and sometimes do not have that may represent digitisable traces of pieces of everyday life much similarity with the person from whom they are practices. However, it is known that the social is digitalised derived and therefore defy the illusion of being a window only in a rudimentary form (Birk and Samuel, 2020). Also, to real life (Lyon, 2003: 22; Mager and Mayer, 2019). many aspects of our everyday lives and routines are not recognised by digital devices, as our complete experiences Consequences and future aspects with those practices are also not digitisable (Mau, 2017; Quinn, 2021). There are different ways to conceptualise PM will have different impacts on the field of medicine as the collected digital features. The data about a person has scientific discipline and healthcare depending on what is the been described as ‘data doubles’, ‘data traces’, ‘data silhou- centre of analysis: the field’s ontology and epistemology or ettes’ and ‘digital fingerprints’ (Lyon, 2003: 22; the patient. The logic behind data analysis is driven by Mindstrong Health, 2021; Quinn, 2021: 2). Loi (2019: mathematical and statistical aspects leading to a shift to cal- 158) criticises this sort of framing that assumes being a culate outcomes for the individual from population data. In one-to-one reflection of a person and suggests, digital the examples of breast cancer diagnosis, the sample of 8 Big Data & Society individuals with breast cancer is compared to knowledge 2019: 571). Big data is framed as a missing piece that about ‘genetic characteristics of a population sub-group’ might help to understand diseases in their ‘expression in (Low et al., 2018; Prainsack, 2017: 4). This logic of PM terms of the lived experience of individuals’ (Burnett, from the 1990s and 2000s meant, clinically collected data 2015; Huckvale et al., 2019; Jain et al., 2015: 463). The of certain subpopulations served to find targeted treatment continuous, in situ and live monitoring of a patient’s activ- for the individual (Prainsack, 2017). This logic stemming ities are stylised as the missing piece in understanding from epidemiology also pervades other medical disciplines. mental health. Additionally, PM has a ‘systems medicine’ approach when ‘Empowerment of the patient’ is a frequently found trope it comes to the object under investigation. The focus here in DP, hoping that people take agency over their health has been redirected from molecular or cell biology to which seems just another form of ‘responsibilisation’ systems and models of disease pathways. This development found within PM in general (Erikainen and Chan, 2019; presents an ontological and epistemological change in bio- Prainsack, 2017). However, it is still questionable who medicine (Erikainen and Chan, 2019). When centring the will profit from digitalised psychiatry and who will rather patient, there seems to be a twofold shift. One shift is see the downsides. While it has been shown that access from the individual to the population when it comes to to health information, e.g. through wearable use is empow- the logic of prediction, where population data is used for. ering for some people, usually those with more At the same time, there is a shift from population medicine resources, social minorities might find it rather disempow- to individualisation when it comes to how responsibility is ering (Birk and Samuel, 2020; Prainsack, 2017). DP is dis- framed. This means that rather than looking at population cussed to disregard social inequalities and conceptualise it disease risk, the discipline focuses on individual risk or rather as individual mental health conditions (Birk and individual prediction. This entails a shift from the popula- Samuel, 2020). Real-life monitoring might bring diagnostic tion to the individual regarding the responsibility for and therapeutic opportunities, however, it opens also the health. Consequently, the individual can more easily be possibility to surveillance (Dagum and Montag, 2019; made responsible, and population systemic aspects are not Lyon, 2003). Banner (2019: 7–8), e.g. describes how sur- being taken into account. Erikainen and Chan (2019: 314) veillance through DP might be used ‘to regulate, define describe this with ‘responsibilization’ of the individual and control’ certain populations such as POC or disabled (rather than medical professionals) which comes from people and ‘used to enforce neoliberal regimes of austerity’, being seen as individual with autonomy first and not as bringing more risk to historically discriminated populations part of a collective with a solidaristic arrangement than to others. (Prainsack, 2017). Additionally, healthcare shifts from Big data in general and related data in health, in particu- being reactive to being proactive, which can be seen by pre- lar, is considered to be the new gold. Health data is com- dictions that are thought to be possible based on genomic modified and profit is made from peoples’ health data data (Erikainen and Chan, 2019; Ferryman and Pitcan, (Banner, 2019; Cosgrove et al., 2020; Prainsack, 2017). 2018). Breast cancer is an example of that when Low Digital platforms are programmed with the company’s et al. (2018: 503) speak of ‘genomic profiling by clinical purpose of collecting as much data as possible and it is a sequencing’ as the next step to identify cancer risks in fairly frequent practice that data is sold and analysed for healthy individuals. This is already a reality for people completely distinct purposes compared to the primary with a strong family history of cancer who are screened reason for data collection. Cosgrove et al. (2020) lay out for BRAC1/2 gene mutations to decide about more frequent that 92% of mental health apps are known to share data check-ups, preventive measures, or specific forms of with third parties such as Facebook and Google without treatment. users having the choice to weigh in on that practice. This Similarly, for DP, experts of the field claim how DP is practice is called data repurposing and has been critiqued all about ‘person-centered care’ which also here results in from ethical and social scientist perspectives (Kitchin, more responsibility being given to each individual 2014). (Huckvale et al., 2019: 88). However, there are additional To recapitulate and expand: What we see in both PM and far-ranging consequences when analysing the processes DP is the assumption that knowing the truth is enhanced within DP. I will now focus on those related to data through the data-driven techniques of the field. For DP, types. Regarding the field’s ontology and epistemology the basis is continuous, in situ and live monitoring of every- DP even more so than PM is hoped to finally provide day life practices. For PM based on genetics and genomics, long longed for (digital) biomarkers for mental health con- it is genetic data marked as holding the ultimate truth about ditions. The RDoC funded in 2009 by the NIMH is one life and being extra precise. The digital form of data is an prominent example for the drive for a change of the essential aspect for the expressed logic because only then fieldaimingatmovingpsychiatrybeyondsofar ‘descrip- can necessary practices of collection, transfer, analysis tive diagnosis’ for mental health conditions with a new and interpretation be possible. Digitalisation, datafication, classification system based on ‘basic science’ (Rüppel, big data and their analysis provide additional information Baumgartner 9 to both fields to be able to claim being closer to their own When data is the primary focus, the rationale is usually construction of knowledge. With this newly generated that more data, and sometimes more precise data as in the knowledge, experts in the field claim that they can offer case of PM, is better. This is a standard rationale within more fitting treatment options and finally, better health. big data, even though it has been criticised extensively Ultimately, data is an essential component for knowledge also by the stakeholders involved (boyd and Crawford, production within the field through which the field seeks 2012; Kitchin, 2014). This means medicine and computer to differentiate itself from other, less digital-data-prone science and statistics now share this same mantra being fields. Being data-driven leads to further consequences, more data is better. To satisfy this need for data within such as the commodification of individual’s health data PM and DP, health institutions have to move further into and how through ‘personalisation’ responsibility shifts to the direction of digitalisation. This might bring new stake- the individual. holders, experts of data-intensive tools, into the picture who offer related products and solutions for problems. The value system of those new stakeholders will find a place in med- icine and health. One result will be a further commodifica- Discussion tion of disease and health and an increased dependence on While both PM and DP have a focus on digitisable data, I those entities that deal with data collection, analysis and have pointed out the differences and similarities of PM interpretation, which are often private companies (Lupton, and DP related to several aspects of data use. First, I have 2018b; Prainsack, 2017; Ruckenstein and Schüll, 2017; laid out the type of data and data producers: genetic data Saukko, 2018). deriving from lab facilities for PM versus big data from a Healthcare for patients may change extensively. variety of digital devices with sensors used in DP. Then I Individualisation seems to lead to responsibilisation. Care have shown the different systematic of data collection: might change from reactive to proactive and predictive. active and rare data production for PM versus How will it affect people to know that they are continuously moment-by-moment, in situ and live data collection for being monitored? To which recursive effects will this lead? DP. The focus on digital data also informs the dominant Digital sociologists and STS scholars discuss that it is understanding within the fields of PM and DP, and results unclear how much patients or so-called data subjects will in genetic determinism and advancing the genetic imagin- ultimately profit (Lupton, 2018a). Those people with ary for PM based on genetics and data fundamentalism in more resources might indeed be empowered by tailored DP. Both logics share the aspect of trust in digitisable treatment and access to their own health data and more data, which is thought to hold the ultimate truth leading information. However, critical data scientists point out to better health. With this logic, these two data-intensive that social minorities might not be the ones profiting but approaches uplift themselves over other less experience the downside: being monitored can just as digital-data-prone medical disciplines. Proof for this easily result in surveillance. People at the social margins aspect is the discursive distancing of experts of DP. They might be depersonalised and disempowered (Prainsack, frame digital traces as ‘objective’‘biomarkers’ for mental 2017). This opens an entirely new array of questions regard- illnesses against the ‘subjective’ standard of self-reports ing surveillance and what Foucault (2008: 1) called ‘biopo- of mental health symptoms which can also be due to the litics’, the establishment of state control over functions and unsuccessful search for genetic and molecular biomarkers processes of life. These aspects are especially salient for DP within psychiatry (Birk and Samuel, 2020; Brietzke et al., and will be even more salient once PM and DP are 2019: 223; Huckvale et al., 2019: 1). In medicine, as in combined. other scientific disciplines, objectivity ranks higher than Being data-driven approaches, many advantages, and subjectivity (Reiss and Sprenger, 2020). Information disadvantages of datafication and matters pertaining to big seems to gain in resilience and worth once it has been col- data and AI come into play. As critical scholars in social lected by a digital device (Lupton, 2015). Even more so if it and human science have shown, only certain knowledge has been collected close to the patient as is the case in DP. is quantifiable and thus digitisable. Many aspects of life ‘Real-life’ data is conceptualised as being ‘objective’ versus cannot be translated into numbers, such as experiential, self-reports being framed as ‘subjective’. Consequences intuitive, tactile and emotional knowledge (Quinn, 2021). might be a shift from reactive medicine to proactive and If data-driven methods gain more importance in the field preventive medicine resulting in individualisation and of health and illness, it is supposed to have huge conse- responsibilisation of the patient through both PM and DP. quences on basic epistemologies and ontologies around Although DP may offer more possibilities of exchange health and illness. Some of the non-quantifiable and non- and support for patients, the constant flow of ‘real-life’ digitasable aspects will not find their way in, e.g. decision data might open the door to more surveillance through con- support tools. Not only will the outcome be different if nected devices, which is the other side of 24 h monitoring they do not inform the decision, but they will also be that, e.g. Mindstrong proposes for its users (Metz, 2018). hidden and in the end, valued less (Lupton, 2015; 10 Big Data & Society Mennicken and Espeland, 2019; Quinn, 2021). This is even stemming from it. Since possibilities for data collection more critical as both fields are rich in descriptions how data have risen, PM is seeking to include all types of data to con- and knowledge gathered in PM and DP will change classi- struct a patient’s ‘unique thumbprint’ (Prainsack, 2017: 3). fication of diseases, identifying influences data-based From the side of DP efforts exist to introduce clinical data to approaches are expected to have on basic epistemologies the phenotype analysis, so-called ‘enriched data for DP’ and ontologies in medicine and healthcare. Therefore, crit- (Liang et al., 2019: 290). Both fields work closely together ical scholars should have a close eye on these paramount with data and computer science and share a similar big data changes. Prainsack (2017: 188) explains how the data logic. Thus, combining all data available, no matter if valuing logic in PM follows a ‘tacit hierarchy of utility, genetic and clinical data or ‘real-life data’, would be the with digital and computerable data on top, and unstructured, next logical step. Some experts already depict a horror sce- narrative, and qualitative evidence at the bottom’. A logic nario of genetic determinism in psychological diseases if that was introduced by the technologies used and not by biological and real-life data would be collected for mental professional experts. DP in particular shows an health conditions (Comfort, 2018). all-too-simplified view on mental health through its Additionally, the COVID-19 pandemic has shown us search for digital biomarkers based on behaviour pheno- how viral genetic information can be used to influence types. Birk and Samuel (2020) rightfully critique the every aspect of our life, including surveillance and quaran- usage of digital data as a proxy for social life for many tine restrictions. Also, it is thought to change our mental reasons. First, important knowledge will not be taken into healthcare, e.g. through the increased access to telehealth account and may ultimately be lost. Many non-quantifiable (Melcher et al., 2020). These transformations call for aspects have for a long time been essential for how deci- social science to engage in the field. There are many ques- sions have been made within medicine. Social science tions still to be asked: epistemological and ontological shows medicine had always had a more ‘human’ aspect to redirections in the field of medicine and healthcare, trans- it than science would claim. Practical knowledge, intuition formations in categories around health and illnesses with and nuances in the physician–patient relationship have new technologies, changes in doctor–patient relationships always been very important not just to how decisions and how institutions and people handle health-related around health and illness have been made, but also to data. Institutionally, it will be interesting to analyse power what constitutes the professions in health care (Kim et al., shifts amongst stakeholders and the changes of values and 2018; Spinnewijn et al., 2020). Second, if decisions are logics that accompany them. Last but not the least, addres- taken based on historical data, existing categories might sing questions around health equality and which conse- be reified and naturalised (Mau, 2017; Mennicken and quences these developments entail for population health Espeland, 2019; Quinn, 2021). This entails also assump- and individual patients is crucial in these changing times. tions around normality and bias in algorithms which is harder to dismantle because of black box phenomena Acknowledgements (Birk and Samuel, 2020). The author would like to thank three anonymous reviewers for Finally, all the processes described before can also be their very insightful comments and Tamara Schwertel, Regina conceptualised as different aspects of biomedicalisation, Ammicht-Quinn and Ursula Offenberger for their valuable feed- the transformance of biomedicine by technoscientific inter- back. Furthermore, she would like to thank Mrunmayee Sathye ventions such as computers and genetics, with the conse- and Lukas Haeberle for their help on the manuscript. quences that the biological gains in importance. In the analysis before I found the following processes of biomedi- Declaration of conflicting interests calisation: a focus on health, surveillance and risk which we The author declared no potential conflicts of interest with respect find in PM (genomic profiling) but above all in DP, then the to the research, authorship and/or publication of this article. ‘technoscientisation of biomedicine’ which can be found in both logics of PM and DP, and a change in biomedical Funding knowledge represented by the aim for new disease classifi- The author received no additional financial support for the cations through data in both PM and DP (Clarke et al., research, authorship and/or publication of this article. 2003: 166). Also, the responsibilisation of the individual in both PM and DP and commodifying health data in DP ORCID iD fits under this umbrella term. Renate Baumgartner https://orcid.org/0000-0002-3401-1870 Conclusion and outlook Notes PM and DP have the potential and are already changing 1. For example, the mandatory CLIA certificate in the United medicine and healthcare. The basis for these developments States for laboratories performing tests of human specimen’s is also the digital data used and the logics of the field diagnosis, prevention or treatment of disease or health Baumgartner 11 problems. Genetics tests are classified as moderate or high the-hidden-biases-in-big-data. (last accessed 6 December complexity tests. 2021). 2. Erikainen and Chan (2019) also describe how this argumenta- Dagum P (2019) Digital brain biomarkers of human cognition and tion leads to big money flows to PM. The construction of truth mood. In: Baumeister H and Montag C (eds) Digital with the production of ‘hope’ within the field, in the end, is able Phenotyping and Mobile Sensing: New Developments in to accumulate different forms of capital. Psychoinformatics. Cham: Springer Nature Switzerland AG, 3. Some practices in DP involve both DP and subjective assess- pp.93–108. ments. Smartphones can be used for ecological momentary Dagum P and Montag C (2019) Ethical considerations of digital phe- assessments which repeatedly sample behaviours, thoughts notyping from the perspective of a healthcare practitioner. In: and feelings in real time (Garcia-Ceja et al., 2018). Baumeister H and Montag C (eds) Digital Phenotyping and 4. The aspect of being able to predict mental health issues and the Mobile Sensing: New Developments in Psychoinformatics. consequences thereof is an important topic that asks to be dis- Cham: Springer Nature Switzerland AG, pp.13–29. cussed, but which exceeds the focus of this paper. Erikainen S and Chan S (2019) Contested futures: Envisioning 5. Invoking being able to monitor real life might also be due to the “personalized,”“stratified,” and “precision” medicine. New frustration in psychiatry of not having sufficient diagnosis and Genetics and Society 38(3): 308–330. treatment tools for people with mental health conditions Ferryman K and Pitcan M (2018) What is precision medicineData (Mindstrong Health, 2021). & Society. Foucault M (2008) The Birth of Biopolitics. Lectures at the Collège de France, 1978–79. Hampshire and New York: Palgrave Macmillan. References Garcia-Ceja E, Riegler M, Nordgreen T, et al. (2018) Mental Banner O (2019) Technopsyence and Afro-Surrealism’s cripis- health monitoring with multimodal sensing and machine learn- temologies. Catalyst: Feminism, Theory, Technoscience 5(1): ing: A survey. Pervasive and Mobile Computing 51: 1–26. 1–29. Grommé F (2016) Data mining ‘problem youth’:Looking closer Barnett I, Torous J, Staples P, et al. (2018) Relapse prediction in But Not seeing better. In: van der Ploeg I and Pridmore J schizophrenia through digital phenotyping: A pilot study. (eds) Digitizing Identities. New York; London: Routledge Neuropsychopharmacology 43(8): 1660–1666. Taylor & Francis, pp.163–183. DOI: 10.4324/9781315756400. Birk RH and Samuel G (2020) Can digital data diagnose mental Hood L, Balling R and Auffray C (2012) Revolutionizing medi- health problems? A sociological exploration of ‘digital pheno- cine in the 21st century through systems approaches. typing’. Sociology of Health and Illness 42(8): 1873–1887. Biotechnology Journal 7(8): 992–1001. boyd D and Crawford K (2012) Critical questions for big data. Hopman R (2020) Opening up forensic DNA phenotyping: The Information, Communication & Society 15(5): 662–649. logics of accuracy, commonality and valuing. New Genetics Brietzke E, Hawken ER, Idzikowski M, et al. (2019) Integrating and Society 39(4): 424–440. digital phenotyping in clinical characterization of individuals Huckvale K, Venkatesh S and Christensen H (2019) Toward with mood disorders. Neuroscience and Biobehavioral clinical digital phenotyping: A timely opportunity to Reviews (104): 223–230. DOI: 10.1016/j.neubiorev.2019.07. consider purpose, quality, and safety. npj Digital Medicine 009. 2(1): 1–11. Burnett F (2015) Four questions for NIMH Director Thomas Jain SH, Powers BW, Hawkins JB, et al. (2015) The digital phe- R. Insel. Available at: https://ct.counseling.org/2015/03/four- notype. Nature Biotechnology 33(5): 462–463. questions-for-nimh-director-thomas-r-insel/ (last accessed 6 Keller EF (2002) The Century of the Gene, 3rd ed. Cambridge and December 2021). London: Harvard University Press. Cedars-Sinai Blog (2019) New Recommendations for BRCA Kim K, Heinze K, Xu J, et al. (2018) Theories of Testing: Should You Be Screened? Available at: https:// health care decision making at the end of life: A www.cedars-sinai.org/newsroom/new-federal-guidelines-for- meta-ethnography. Western Journal of Nursing Research brca-testing-should-you-be-screened/. (last accessed 6 40(12): 1861–1884. December 2021). Kitchin R (2014) Big data, new epistemologies and paradigm Clarke AE, Shim JK, Mamo L, et al. (2003) Biomedicalization: shifts. Big Data and Society 1(1): 1–12. Technoscientific transformations of healht, illness, and U.S. Kumari S, Chouhan U and Suryawanshi SK (2017) Machine biomedicine. American Sociological Review 68(2): 161–194. learning approaches to study HIV / AIDS. Bioscience Comfort N (2018) Genetic determinism. Nature 561: 461–463. Biotechnology Research Communications 10(1): 34–43. Cooper R and Paneth N (2020) Will precision medicine lead to a Liang Y, Zheng X and Zeng DD (2019) A survey on big data- healthier population? Issues in Science and Technology driven digital phenotyping of mental health. Information XXXVI(2): 64–71. Available at: https://issues.org/precision- Fusion 52: 290–307. medicine/. Loi M (2019) The digital phenotype: A philosophical and ethical Cosgrove L, Karter JM, McGinley M, et al. (2020) Digital pheno- exploration. Philosophy & Technology 32: 155–171. typing and digital psychotropic drugs: Mental health surveil- Low SK, Zembutsu H and Nakamura Y (2018) Breast cancer: The lance tools that threaten human rights. Health and Human translation of big genomic data to cancer precision medicine. Rights 22(2): 33–40. Cancer Science 109(3): 497–506. Crawford K (2013) The hidden biases in big data. Havard Lupton D (2015) Digital Sociology. London and New York: Business Review April. Available at: https://hbr.org/2013/04/ Routledge. 12 Big Data & Society Lupton D (2016) The Quantified Self, 1st ed. Cambridge: Polity Prainsack B (2017) Personalized Medicine. Empowered Patients Press. in the 21st Century? New York: New York University Press. Lupton D (2018a) Digital Health – Critical and Cross-Disciplinary Quinn RA (2021) Artificial intelligence and the role of ethics. Perspectives. London and New York: Routledge. Statistical Journal of the IAOS 37(1): 75–77. Lupton D (2018b) How do data come to matter? Living and Ranjan Y, Rashid Z, Stewart C, et al. (2019) Radar-base: Open becoming with personal data. Big Data and Society 5(2): 1–11. source mobile health platform for collecting, monitoring, and Lyon D (2003) Surveillance as Social Sorting. London and analyzing data using sensors, wearables, and mobile devices. New York: Routledge. JMIR mHealth and UHealth 7(8): 1–12. Mennicken A and Espeland WN. (2019) What’s new with Reiss J and Sprenger J (2020) Scientific objectivity. Stanford numbers? Sociological approaches to the study of quantifica- Encyclopedia of Philosophy. Winter 20. Metaphysics tion. Annual Review of Sociology 45: 223–245. Research Lab, Stanford University. Available at: https://plato. McLean S, Ressel K, Koenen KC, et al. (2020) The AURORA stanford.edu/archives/win2020/entries/scientific-objectivity/. study: A longitudinal, multimodal library of brain biology Ruckenstein M and Schüll ND (2017) The datafication of health. and function after traumatic stress exposure. Molecular Annual Review of Anthropology 46: 261–278. Psychiatry 25(2): 283–296. Rüppel J (2019) “Now is a time for optimism”: The politics of Mager A and Mayer K (2019) Body data—data body: Tracing personalized medicine in mental health ambiguous trajectories of data bodies between empower- research. Science Technology and Human Values 44(4): 581–611. ment and social control in the context of health. Saeb S, Lattie EG, Schueller SM, et al. (2016) The relationship Momentum Quarterly – Zeitschrift für sozialen Fortschritt between mobile location sensor data and depression 8(2): 95–109. symptom severity. PeerJ 4: e2537. Mau S (2017) Das Metrische Wir. Über Die Quantifizierung Des Saukko P (2018) Digital health – a new medical cosmology? The Sozialen. Berlin: Suhrkamp Verlag. case of 23andMe online genetic testing platform. Sociology of Melcher J, Hays R and Torous J (2020) Digital phenotyping for Health and Illness 40(8): 1312–1326. mental health of college students: A clinical review. Spinnewijn L, Aarts J, Verschuur S, et al. (2020) Knowing what Evidence-based Mental Health 23(4): 161–166. the patient wants: A hospital ethnography studying physician Metaxiotis KS and Samouilidis JE (2000) Expert systems in med- culture in shared decision making in the Netherlands. BMJ icine: Academic illusion or real power? Information Open 10: e032921. Management and Computer Security 8(2–3): 75–79. Topol EJ (2019) High-performance medicine: The convergence Metz R (2018) The smartphone app that can tell you’re depressed of human and artificial intelligence. Nature Medicine 25: 44–56. before you know it yourself. Available at: https://www. Torous J, Kiang MV, Lorme J, et al. (2016) New tools for new technologyreview.com/s/612266/the-smartphone-app-that- research in psychiatry: A scalable and customizable platform can-tell-youre-depressed-before-you-know-it-yourself/. (last to empower data driven smartphone research. JMIR Mental accessed 6 December 2021). Health 3(2): 16. Milne R (2020) The rare and the common: Scale and the genetic U.S. Government (2021) Electronic Code of Federal Regulations: imaginary in Alzheimer’s disease drug development. New Part 493 – Laboratory Requirements. Available at: https:// Genetics and Society 39(1): 101–126. www.ecfr.gov/current/title-42/chapter-IV/subchapter-G/part- Mindstrong Health (2021) Using science to help us seek the truth. 493 (last accessed 6 December 2021). Available at: https://mindstrong.com/science/ (last accessed 6 van Dijck J (2014) Datafication, dataism and dataveillance: Big December 2021). data between scientific paradigm and ideology. Surveillance Mol A (2008) The Logic of Care. London: Routledge. and Society 12(2): 197–208. National Human Genome Research Institute (n.d.) The CLIA Weiner K, Martin P, Richards M, et al. (2017) Have we seen the Framework. Available at: https://www.genome.gov/Pages/ geneticisation of society? Expectations and evidence. PolicyEthics/GeneticTesting/The_CLIA_Framework.pdf (last Sociology of Health and Illness 39(7): 989–1004. accessed 6 December 2021). Weiss KM (2017) Is precision medicine possible? Issues in Peters T (2012) Playing God? Genetic Determinsm and Human Science and Technology 34(1). Available at: http://issues.org/ Freedom, 2nd ed. London and New York: Routledge. 34-1/is-precision-medicine-possible/. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Big Data & Society SAGE

Precision medicine and digital phenotyping: Digital medicine's way from more data to better health:

Baumgartner, Renate
Big Data & Society , Volume 8 (2): 1 – Dec 22, 2021
Download PDF
12 pages

Loading next page...
 
/lp/sage/precision-medicine-and-digital-phenotyping-digital-medicine-s-way-from-MU0LIHTOKR

References (78)

Publisher
SAGE
Copyright
Copyright © 2022 by SAGE Publications Ltd, unless otherwise noted. Manuscript content on this site is licensed under Creative Commons Licenses.
ISSN
2053-9517
eISSN
2053-9517
DOI
10.1177/20539517211066452
Publisher site
See Article on Publisher Site

Abstract

Precision medicine and digital phenotyping are two prominent data-based approaches within digital medicine. While pre- cision medicine historically used primarily genetic data to find targeted treatment options, digital phenotyping relies on the usage of big data deriving from digital devices such as smartphones, wearables and other connected devices. This paper first focusses on the aspect of data type to explore differences and similarities between precision medicine and digital phenotyping. It outlines different ways of data collection and production and the consequences thereof. Second, it shows how these sorts of data influence dominant beliefs in the field: The field of precision medicine relying on the dominant understanding of ‘genetic determinism’ imported from genetics, digital phenotyping building on the logic of ‘data fundamentalism’. In the end, the analysis shows how digital data informs potentials as well as challenges of precision medicine and digital phenotyping: a better health care for (some) individuals connected with individualisation and respon- sibilisation for all, with a prognosed shift from reactive to preventive medicine. Additionally, data-based approaches might facilitate epistemological and ontological redirections for the whole field of medicine that will also affect knowledge pro- duction and a reassessment of the value of different types of knowledge (quantifiable vs. non-quantifiable) with all its con- sequences. Institutionally, it might lead to shifts in distribution of power to experts in big data related technologies, i.e. private companies. Keywords Personalised medicine, genetic determinism, data fundamentalism, data-driven medicine, big data, digital health This article is a part of special theme on Digital Phenotyping. To see a full list of all articles in this special theme, please click here: https://journals.sagepub.com/page/bds/collections/digitalphenotyping only solve major healthcare challenges, but could shift the Introduction whole medical field from responsive medicine to predictive Precision medicine (PM) and digital phenotyping (DP) are medicine (Hood et al., 2012). DP seeks to revolutionise buzzwords in medicine. The promise they contain is already mental health by collecting continuous ‘real life data’ expressed in their naming. PM aims to offer a precise and from digital devices of everyday use such as smartphones targeted treatment for an individual or a group of people and wearables. With this additional data, knowledge gaps based on clinical data. DP promises to categorise people in mental health diagnosis could be closed without burden- into clinically relevant categories for the treatment of ing patients with further tasks (Huckvale et al., 2019). PM mental diseases, with the help of digital data. Both promise revolutionary shifts in medicine and healthcare. Both approaches are data-driven, have been developed University of Tübingen, Center for Gender and Diversity Research only in recent decades and hold out the prospect of chan- (ZGD), Tübingen, Germany ging the field of medicine on the basis of digitisable data. Corresponding author: They are both said to transform medicine as a scientific Renate Baumgartner, University of Tübingen, Center for Gender and field and as a healthcare discipline. PM claims to offer indi- Diversity Research (ZGD), Brunnenstr. 30, D-72074 Tübingen, Germany. vidualised treatment for specific diseases which could not Email: renate.baumgartner@uni-tuebingen.de Creative Commons CC BY: This article is distributed under the terms of the Creative Commons Attribution 4.0 License (https:// creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access page (https://us.sagepub.com/en-us/nam/open-access-at-sage). 2 Big Data & Society and DP, being data-based, require specific statistical and might look like an odd choice, since DP is practically a computational expertise, which results in new stakeholders part of PM, while PM is recently described as an approach entering the field of medicine. to collect as much varied data as possible to compose a Data today is said to be the new gold. Many publications patient’s health map. Historically, however, big parts of deal with the datafication of our lives as well as with the dis- PM predominantly deal with genetic and genomic data, tinct influence of data not only being collected and aggre- while DP is till date firmly grounded in mental health gated but also analysed by algorithmic means (Lupton, issues. Also, different stakeholders seem to concentrate 2018b; Ruckenstein and Schüll, 2017; van Dijck, 2014). on different types of data or different disease types Technical breakthroughs around data saving, transfer and (Mindstrong, 2021; Prainsack, 2017). Contrasting PM analysis within the last 20 years have made it possible based on genetics with DP dealing with mental health that approaches like PM and DP have a technical basis to conditions, we see the similarities as well as the differ- even be thought of and developed. Datafication and an ences within the working logic in a field that tends to introduction of algorithmic tools such as artificial intelli- mental illnesses versus a field that predominantly gence (AI) have taken place in different fields including deals with illnesses experts ground in the realm of the politics, economy, and law, but significantly also in health body. For this endeavour, I will focus on the following and the field of medicine (Topol, 2019). examples: breast cancer prediction, diagnosis, and treat- Digitalisation in medicine already started in the 1970s, ment as an example of PM based on genetics, and the then by the name of expert systems which were used to digital platform Mindstrong which claims to treat mental get diagnosis and treatment recommendations (Metaxiotis health issues such as depression, bipolar disorder and and Samouilidis, 2000). Over the last 15 years datafication, PTSD basedonDP. or ‘the conversion of qualitative aspects of life into quanti- fied data’, has picked up (Ruckenstein and Schüll, 2017: Precision medicine 261). The process began with more and more patient infor- mation finding its way from paper records to electronic PM is a medical approach that is best known for aiming to patient files, including the involvement of new techniques offer a specifically targeted treatment regime to a person or with digital output, e.g. in radiology. However, when a certain group of people, ideally leading to a ‘personalised’ people recently started using smartphones and wearables, treatment for each person, which in the end could declare engaging in digital communities, thereby producing also common disease labels obsolete (Ferryman and Pitcan, health-related data that can be collected, shared and ana- 2018; Prainsack, 2017). Other terms used for this approach lysed, datafication picked up momentum (Ruckenstein in medicine are ‘personalised medicine’ or ‘stratified med- and Schüll, 2017). Thus, both technical breakthroughs icine’. The terms have appeared and been used in different and sociocultural developments such as wearable usage historical, geographical and disciplinary contexts. I will use and sharing health-related experiences on social media, the term ‘precision medicine’ because it is the more recent e.g. as part of the quantified self-movement, made it possi- term and for the sake of simplicity (Erikainen and Chan, ble for data-driven methods to also win ground in medicine 2019; Ferryman and Pitcan, 2018). Since 1990, PM has and healthcare (Lupton, 2018b; Ruckenstein and Schüll, also been a buzzword and the closely related field of 2017). genomic science has been able to acquire considerable All these processes have resulted in PM and DP not just funding. An example is the human genome project (HGP) remaining a vision but becoming reality and constituting or the ‘all of us’ project in the United States (Cooper and two of the most spoken about data-driven approaches in Paneth, 2020; Ferryman and Pitcan, 2018), both with enor- medicine till date. Still, the two approaches have had his- mous expenditures. The ‘all of us’ project will reach total torically distinct starting points. They use different types costs of about 1 billion US$ once finalised (Cooper and of data which is based on different ways of data produc- Paneth, 2020). PM represents an interdisciplinary endeav- tion. This paper will explore what PM and DP has in our combining biology, medicine, informatics, computer common and what separates them based on the type of science, mathematics and statistics (Erikainen and Chan, data they use. In the end, it will show how this data-driven 2019). Computer science, mathematics and statistics func- aspect of PM and DP results in similar logics within both tion to store and analyse data. The collaboration between approaches, whether in dealing with mental health condi- human molecular genetics with modern computer science tions or diseases based in the realm of the body. First, both since the 1950s has been a thriving force to establish PM approaches will be analysed according to the sort of data (Cooper and Paneth, 2020; Ferryman and Pitcan, 2018). they use. In the second part, I will outline the logic The most prominent desiderata of PM for the field of med- behind the main arguments in the field, also based on icine are disease understanding and prediction with a strong these types of data. In the end, I will conclude with the focus on cancer, aiming to find a suitable treatment with challenges PM and DP might pose due to their role as low incidence of adverse drug reactions (Cooper and digital data-based approaches. Comparing PM and DP Paneth, 2020; Prainsack, 2017). Examples are targeted Baumgartner 3 breast cancer treatment or ‘personalised’ antiretroviral treat- (Barnett et al., 2018; Dagum, 2019; Jain et al., 2015). There ment for HIV-positive individuals based on treatment opti- is a lot of ongoing research trying to tackle different needs, misation tools (Kumari et al., 2017; Low et al., 2018). Data e.g. the RADAR-CNS project from European Union aiming about genetic predisposition, lifestyle information, clinical to develop an open-access platform around mobile health data, etc., all of which has to be ‘structured, digital, quanti- data, and initiatives to provide mental health treatment by fied, and computable data’ is combined creating a ‘unique smartphones to low-income populations. However, much thumbprint’ of a person to inform their diagnosis and treat- of the research is still in an experimental stage (Melcher ment, also called ‘personal health maps’ (Prainsack, 2017: et al., 2020; Ranjan et al., 2019). DP is currently used 3–4). The best-known examples till date, however, mostly in the realm of mental illnesses. However, thinking involve genetic and genomic data. of DP as part of a PM that gathers all sorts of data to compose a personal health map, it could be used for all types of diseases (Huckvale et al., 2019; Prainsack, 2017). Digital phenotyping As far as PM aims at incorporating all sorts of data Psychiatry got more interested in PM since struggling for a for diagnosis and personalised treatments, DP seems to be long time under the lack of objective markers for mental the part of a personalisation which Rüppel describes health conditions that science finds are a key basis for being ‘increasingly rearticulated as ‘Big Data’ project’ correct diagnosis and optimal treatment. Under the name (Prainsack, 2017; Rüppel, 2019: 593). DP and PM are of ‘personalized psychiatry’, and other similar labels scien- both data-driven approaches aimed at understanding tists aimed to identify genomic or molecular biomarkers of diseases and offering targeted treatments, only the data mental health conditions to be able to offer targeted treat- they use can vary from biological, genetic and genomic ments to individuals Rüppel (2019: 593) describes being data to sensor-collected data from digital devices such as ‘increasingly rearticulated as ‘Big Data’ project’ smartphones and wearables (DP). (Prainsack, 2017). DP is an advancement of this search for biomarkers that includes data collected by digital Different sorts of data – different devices with sensors used in everyday life in its assessment consequences? to identify ‘digital biomarkers’ (Dagum and Montag, 2019: 14). DP goals are finding ‘(digital) diagnostic markers’ for This section will focus on an analysis of the distinct sorts mental health conditions by correlating a collection of of data used in PM and DP, based on the cases of breast sensor data and self-reported data and (thereby) monitoring cancer treatment for genetics-based PM as an example of and predicting mental health statuses based on these bio- disease grounded in the biological and the platform markers, and behavioural phenotypes (Birk and Samuel, Mindstrong for DP as an example of mental health. In the 2020: 1874; Dagum and Montag, 2019). The idea of DP following, I will point out how the type of data influences or the digital phenotype was raised by two groups of scien- the way the dominant beliefs within the fields of PM and tists, Jain and colleagues and Torous and colleagues at the DP work. same time, both referring to the usage of large amounts of digital data to find patterns in human behaviours and Type of data and data collection traits, linking those to ‘disease phenotypes’ (Birk and Samuel, 2020: 1876). The assumption is that mental Ferryman and Pitcan (2018: 3) define PM based on their health conditions show themselves in digital traces of a ethnographic research as ‘the effort to collect, integrate, person and ‘behaviour-expressed symptoms’ can thus be and analyze multiple sources of genetic and non-genetic identified as ‘behavioural phenotypes’ (Dagum and data, harnessing methods of big data analysis and Montag, 2019: 14; Garcia-Ceja et al., 2018). DP aims to machine learning, in order to develop insights about gain data from ‘naturalistic settings in-situ, leveraging the health and disease that are tailored to the individual.’ The actual real-world’ and was only possible once digital data involved is genetic data and a variety of other data devices such as smartphones and wearables were ubiqui- including clinical data and lifestyle data (Prainsack, 2017). tously available (Birk and Samuel, 2020: 1876). One method of data accumulation within PM is to start Examples of DP are the assessment of the lacking efficacy with data collection just for the means of having a large of lithium for individuals with ALS in slowing down data pool. The HGP would be an example of such an disease progression, assessed through the analysis of approach. Human DNA data began to be collected with online disease communities (Jain et al., 2015). Other exam- the promise that it would inform solutions to health ples are the analysis of insomnia-related tweets, and moni- issues. The project started in 1990 with the goal of toring and prediction of (clinically) relevant outcomes in mapping the entire sequence of human DNA. The main people with mood disorders such as major depression, rationale behind it was to find new insights into human bipolar disorder, and schizophrenia based on sensory mea- health and disease, which helped to secure a huge amount surements but also on peoples’ participation in social media of funding for this project (Ferryman and Pitcan, 2018). 4 Big Data & Society The other approach is to conduct genome and sample ana- data can be used to predict depressive symptom severity lysis within pathology for specific targets. The examples on and might therefore serve as a biomarker for depression. In which I will focus involve the identification of specific summary, DP is based on the digital measurement, collec- genetic characteristics in the genome of women with a tion, analysis and interpretation of enormously varied activ- family history of breast cancer or the pathological sample ities to enhance understanding and treatment of mental of breast cancer patients, to decide upon targeted forms of health conditions. Related data is collected continuously, therapy. People with inherited BRCA1 and BRCA2 gene live and in situ within ‘real-life settings’ (Birk and Samuel, mutations have an increased risk of developing breast 2020). ovarian cancers and should follow therefore more rigid pre- ventive measures. Oncogenic human epidermal growth Data producers factor receptor (HER2) positive breast cancer patients profit from treatment with trastuzumab and lapatinib that Within the genetic part of PM, the processes of data produc- specifically target HER2, the receptor regulating ‘cell tion take place in highly regulated and controlled environ- growth, proliferation and differentiation’ (Low et al., ments. Following the example of pharmaceutical labs, 2018: 502). Additionally, genomic research in breast those facilities need to prove that they follow specific pro- cancer is still ongoing to identify other targets. In general, tocols even to be granted permission to produce the data. this genetic data is analysed, i.e. ‘produced’, at specific The adherence to the protocols has to be assured through times by specific experts in health institutes or external labs. different measures such as quality control, quality assurance Conversely, DP works with digital data collected and audits (National Human Genome Research Institute, through the usage of smartphones, computers, tablets n.d.; U.S. Government, 2021). The facilities are frequently and wearables, such as ‘FitBit’ or the ‘Smartwatch’,and audited regarding their adherence to these protocols. other devices with digital sensors. A variety of hardware Additionally, special devices used by trained expert person- and software sensors come into play collecting data such nel, e.g. lab technicians, are required to obtain the data. The as inertial sensors measuring walking speed, physiologi- data can only be produced by people with specialised train- cal sensors measuring heart rate and ambient sensors mea- ing (lab technicians) who have access to certain facilities suring temperature (all hardware sensors); but also (labs). However, it is not just laborious but also expensive software sensors that measure internet activity, social to produce this data. The necessary facilities and lab equip- media presence, typing speed, etc. (Birk and Samuel, ment are costly and so it is to follow protocols. Large teams 2020; Garcia-Ceja et al., 2018). DP can be divided into of trained personnel are needed and have to be paid to behavioural phenotyping and digital biomarkers (Dagum develop and adhere to protocols and assure their quality, and Montag, 2019). Behaviour phenotyping uses digital and are able to hold the approval status for the whole facility information such as location, physical activity, mood, (U.S. Government, 2021). As a result of the speech patterns, typing speed and call activity, but also resource-intensive and high-priced nature of data produc- social media usage and search terms to search for tion, somebody has to bear the costs and the expenditures ‘behaviour-expressed symptoms’ (Dagum and Montag, are sometimes regulated and tried to be kept at a 2019: 14; Garcia-Ceja et al., 2018). It uses, e.g. passive minimum (Low et al., 2018). In the example of breast sensing data of GPS location and call logs as a proxy cancer, testing of the genetic properties of a patient and for behaviourally phenotyping loneliness (Birk and their sample is done once (Cedars-Sinai Blog, 2019). To Samuel, 2020). Digital biomarkers aim at measuring summarise, the production of genetic data is initiated ‘trait and state changes in neuropathology that can be indi- actively, conducted at specific time points, and is a labor- cative of disease risk, disease onset, disease progression or ious and expensive process. recovery’ (Dagum and Montag, 2019: 14). Mindstrong, For DP, the collected data is produced and shared by e.g. claims to have identified ‘a set of digital biomarkers users of digital devices, whose live and local data can be from human-smartphone interactions that correlate collected continuously. The ways of collection are enor- highly with select cognitive measures, mood state, and mously varied, depending on the digital device and brain connectivity’ that can be related to depression, sensors used (Birk and Samuel, 2020; Garcia-Ceja et al., anxiety, negative, positive affect and other mental condi- 2018). Often this data is collected as a by-product of prac- tions (Mindstrong Health, 2021). Assessments tices, such as typing speed, click behaviour and not collect- Mindstrong uses, e.g. in the ongoing AURORA study ing personal data, as in the example of Mindstrong. on PTSD are ‘continuous-time accelerometry data, key- Compared to the actively initiated aspect in PM, this data stroke characteristics, time and duration of phone calls, is provided actively (when self-reported) as well as pas- time and character length of text messages, text words/ sively (e.g. when cell phone usage is monitored) by the symbols used, time and number of emails, smartphone users. Prainsack (2017: 21) calls the users of the devices screen time, and intermittent GPS data’ (McLean et al., ‘prosumers’, a combination of producers of data and consu- 2020: 5). Saeb et al. (2016) claim mobile phone location mers of information. Most of this data falls under the term Baumgartner 5 big data, which has been described with properties as could be explained with the help of genetics. This ‘volume, velocity, variety, exhaustive in scope, resolution, mindset is called ‘genetic determinism’ (Peters, 2012). relational and flexible’ (Kitchin, 2014: 1; Rüppel, 2019). Cooper and Paneth (2020) point out that in genetically This means that it entails huge amounts of data with a based PM, genetic determinism prevails. Weiss (2017) high granularity. The aspect of continuous real-life data col- also calls this logic ‘mendelian fundamentalism’, which is lection is named as one big advancement of DP compared an even more drastic description. Both are based on the to other diagnostic tools for mental health conditions assumption that living beings are their DNA, or differently (Dagum, 2019). Big data is simultaneously described as put ‘Genes R Us’, and it is thus possible to find solutions precise, because it is produced close to a person and in against illnesses knowing about the genetics behind real life, and as messy or unclean because it might have (Peters, 2012: 10). Also, if all information and logic of been collected for other purposes and in an uncontrolled the DNA are revealed, it would be possible to understand environment with sensors that might not be equipped to how the body (and the mind) work (Peters, 2012). collect data in research quality (Huckvale et al., 2019; Similarly Weiner et al. (2017: 989) speak of a ‘genetic Kitchin, 2014; Torous et al., 2016). There is a whole indus- imaginary’ being invoked where the biological is seen as try dealing (in the double sense) with health-related data key for understanding diseases and a molecular understand- called the data industry (Cosgrove et al., 2020; Prainsack, ing of diseases is supposed to lead to a new type of medi- 2017). These companies offer services around data cine. Also this logic has its roots in genetics’‘molecular saving, collecting, cleaning, trading and analysis (Kitchin, vision of life’ (Weiner et al., 2017: 999). The imaginary 2014). The producer of the data (or ‘prosumer’) is every ‘is being continuously remade and rearticulated’ and gets individual using digital technologies like smartphones and actualised as new hopes are being constructed based on wearables, or other connected devices with sensors new biotechnologies such as gene editing, and the develop- (Garcia-Ceja et al., 2018; Prainsack, 2017). Additionally, ment of biological drugs (Milne, 2020: 103). Ultimately, any person who interacts on an online platform such as the imaginary works for the persistence of the cultural Facebook and Twitter or is simply conducting an online power of genetics. However, in focussing on technological search provides data for this corpus of data. inventions, it is obscuring the social determinants of health The focus on the different types of data used so far in PM and illness (Milne, 2020). based on genetic data and DP reveals the following differ- Erikainen and Chan (2019: 320) showed how the choice ences between PM and DP: (1) the type of data being lab- of rebranding ‘personalised medicine’ that happened in the generated data versus sensor-derived data from ubiquitous U.S. policy context fell on ‘precision medicine’ also available digital devices; (2) data being collected rarely because the adjective ‘precision’ has an ‘ethically neutral versus moment by moment, live and in situ; and (3) the pro- or even positive’ connotation. I suggest that another essen- ducers being professional labs versus users of digital tial argument within the field is the very aspect of ‘preci- devices. This next section will now show how these charac- sion’, which as a characteristic not only seems to speak teristics of data influence the logic within each of the two for the possibility of choosing a targeted treatment but sur- approaches. rounds genetic data like an aura. It is the conception that genetic data is precise: first because of the effort and the ways how it is measured, second because it is conceptua- Logic within the field lised as being the basis of all living beings and third Drawing on Annemarie Mol’s ‘logic of care’, I will use the having the image of being all-encompassing and everything term ‘logic’ for a persisting rational in the field or a style we have to know to base our decisions on; the last two argu- that seems to be appropriate (Mol, 2008: 1). One could ments being firmly rooted in genetic determinism and rear- say the logic within the field is also a certain dominant ticulating the genetic imaginary (Keller, 2002; Peters, 2012; rational existing in the field which is not questioned. This Weiner et al., 2017). It seems like these aspects give the logic within the field seems to have different nuances whole field more credibility, and as if the truth is to be based on the type of data used in the field. Historically, found in the details, and the closer we look, the closer we genetic or genomic data was used for PM, which was first get to understanding it (Erikainen and Chan, 2019; Keller, expanded to clinical data in general and resulted nowadays 2002; Weiss, 2017). The hope to find more molecular in the usage of a whole variety of data that is helpful when markers for breast cancer through genomic profiling is composing a personal health map (Prainsack, 2017). As I one example of this logic (Low et al., 2018). This argument have pointed out, the most prominent examples of PM till fits very well into the scientific worlds of medical science date involve genetic or genomic data. With the establish- and biology, which operate under positivist assumptions. ment of genetics as a key discipline within biology, a It is the mechanistic assumption that understanding a certain dominant mindset was also established, driven by process detailed enough will reveal the truth behind it. the assumption that every question regarding the functions Based on this assumption, analysing the problem more pre- of living organisms including human health and disease cisely will lead to better suitable treatment options simply 6 Big Data & Society because the premises were more precise and thus closer to Huckvale et al., 2019; Jain et al., 2015: 463). The Research the truth. Hopman (2020: 425) in her study on forensic Domain Criteria Initiative (RDoC) in the U.S. aiming at DNA identifies the logic of accuracy within “the search establishing a new classification system based on ‘basic of the uniqueness of the individual”. In a search for the science’ for mental health conditions is an institutional genetic uniqueness of people, this logic results in the accu- step in this direction (Rüppel, 2019: 571). The continuous, mulation of data. It functions also as a ‘logic of expansion’ in situ and live monitoring of individuals’ behavioural that constantly searches for new genetic ‘“territories” to activities collected as big data are stylised as the missing map’ (Hopman, 2020: 428–429). In this constant search piece in understanding mental health because they can be for more data, it permanently has to attract more money. collected continuously and close to real life (Huckvale The logic resembles very much the search for the ‘unique et al., 2019; Jain et al., 2015). At the same time, current thumbprint’ to compile personal health maps Prainsack diagnostic practices for mental illnesses relying on self- (2017: 3) speaks about within PM. Also, PM is constantly reporting of symptoms are framed as not conclusive aiming to attract more money, which will seem to be well enough for diagnosis (Huckvale et al., 2019). One quest invested as ‘precision’ implies PM will make medicine of DP is the search for digital biomarkers of a field fru- ‘more effective, and thus also cheaper’ (Prainsack, 2017: strated by a long unsuccessful search for biological 79). markers (Birk and Samuel, 2020; Brietzke et al., 2019). A Several points of criticism can be raised to counter this variety of data is used, aimed at understanding mental logic. First, experts in genomic medicine reject that ‘the health diseases better. This is framed as having the potential more we learn about the genome, the more distant it not only to guide new ways of measurement and treatment seems to be from a role as a causative agent in most wide- but also to change the classification of diseases and thera- spread diseases’ and instead acknowledge the role of peutic measures (Burnett, 2015; Huckvale et al., 2019; genomic medicine as ‘a way to do science, not medicine’ Jain et al., 2015). This is interesting when looking at differ- (Cooper and Paneth, 2020: 67). Medicine based on ent conceptualisations of big data. Within some fields like genetic knowledge might be helpful for diseases firmly data science, big data is still regarded as unclean and grounded in genetics, however, not for other diseases. messy and there exists a whole industry that focuses on Thus, for multifactorial diseases, other measures such as practices of ‘cleaning’ and preparing big data for analysis. public health seem to be more effective, even if they may Stakeholders who are pro-DP mark common practices of not have the connotation of ‘precision’ and ultimate truth disease identification like interviews between patients and (Cooper and Paneth, 2020; Weiss, 2017). Second, critical doctors as subjective or at least not sufficient and at the social science perspectives would challenge the idea that same time see digital device collected data as the new a solution can only be found based on mechanistic princi- Holy Grail. Thus, it seems as if lived experiences (or ples. As we have seen in the genetic imaginary argument, digital traces of everyday practices), once they are collected they privilege technical solutions over taking into account through digital devices which are able to provide them con- systemic implications and social determinants of health. tinuously, in situ and live, gain in worth also because it was Those principles may be important for reasoning in the a digital device which gathered them and because they are field. However, decisions in the field of medicine seem to collected close to the individual in ‘real-life settings’ (Mau, be far more complex, entail much more information and 2017; Quinn, 2021). Self-tracking from digital devices is rely on manifold practices as studies, e.g. in medical anthro- framed as providing trustworthy data in contrast to the indi- pology, show (Kim et al., 2018; Spinnewijn et al., 2020). vidual body’s perception which is marked as untrustworthy For DP, the logic behind the argument is distinct but or at least not reliable enough to be the sole ground on arrives at a similar conclusion. Generally for DP, the fol- which diagnosis should be based upon (Lupton, 2015). lowing goals are expressed: early disease detection and – I propose that this way of conceptualising data and gues- surveillance, identifying and incentivising healthy behav- sing it would change a whole field can be termed as ‘data iour, developing new, more targeted interventions and treat- fundamentalism’, a term coined by Crawford (2013). Data ment strategies (Jain et al., 2015). Additionally, fundamentalism is defined by ‘the notion that correlation stakeholders claim that DP will be ‘providing a more com- always indicates causation, and that massive data sets and prehensive and nuanced view of the experience of illness, predictive analytics always reflect objective truth’ [because] an individual’s interaction with digital technolo- (Crawford, 2013). Additionally, digital traces are concep- gies affects the full spectrum of human disease from diag- tualised as (digital) ‘biomarkers’ to use a common nosis, to treatment, to chronic disease management’ (Jain concept known in science. Accordingly, Mindstrong’s et al., 2015: 462). Several experts within DP expand their homepage indicates ‘[T]o identify the digital phenotyping hopes towards a description of how this additional knowl- features that could be clinically useful, Mindstrong used edge gained through big data analysis will also lead to powerful machine learning methods to show that specific changes in classification and diagnosis and treatment of dis- digital features correlate with cognitive function, clinical eases ‘in ways that matter most to patients’ (Burnett, 2015; symptoms, and measures of brain activity in a range of Baumgartner 7 clinical studies (Mindstrong Health, 2021).’ Thus, the case information can indeed be labelled as ‘(digital) extended of DP seems to be a combination of data fundamentalism phenotype’ because it may actually show health-related and biologisation of digital traces. It correlates digital fea- conditions of a person and additionally bears traces of tures to states of health and illness (‘digital biomarkers’) how the user and other people retroact with it. His which have big data as a prerequisite and thus, comes to framing strengthens the aspect that it is also ‘a collective the conclusion that those ‘reflect objective truth’. creation, involved in social feedback loops’. Lupton However, so far no digital biomarker merits the scientific claims that the ‘human-data assemblages’ connect body definition of a biomarker, i.e. ‘a characteristic that is objec- and data, but that people make their data, as well as data, tively measured and evaluated as an indicator of normal makes people (Lupton, 2016, 2018b: 1). Still, these ‘data biologic processes, pathogenic processes, or pharmacologic bodies’ are ‘lively’ in that they are ‘unstable and generative’ responses to a therapeutic intervention’ (Dagum, 2019: 14). and ‘lead a life of their own’ (Lupton, 2018b: 2; Mager and Different scientists working on DP come to the conclusion Mayer, 2019: 95, 98–99). These data doubles ‘are not inno- that digital biomarkers are objective, and ‘reliable informa- cent or innocuous virtual fictions […]. They have ethics, tion on each individual’s behaviour’ (Brietzke et al., 2019: politics’ and are thus, ‘no longer “doubles” […] but they 223; Huckvale et al., 2019; Jain et al., 2015). Insel, e.g. are intrinsically interwoven with bodily practices and bio- compares the possibility of monitoring brain function by politics’ (Lyon, 2003: 27; Mager and Mayer, 2019: 99). DP to a ‘continuous glucose monitor in the world of dia- Companies like Mindstrong might have less interest in betes’ (Metz, 2018). These statements show how medical tracing back data to one particular individual but more in experts in the field stylise DP to provide real and reliable aggregating data about a collective to gather more data information about peoples’ behaviour that can be correlated for digital biomarkers. Still, the logic within the field of to mental health conditions. They go even further when DP works in the way that the continual flux of ‘real life claiming they could indeed predict mental health status, data’ gives us access to ‘real life’, so to say to a deeper e.g. relapses into depression, before individuals themselves truth or a truth behind it all, which is an authenticity and professionals would be able to know about it (Dagum, claim. This seems to be equivalent to being closer to 2019). They conceptualise big data on behaviours as a people. Critical data studies rather suggest that data traces reflection of lived experiences. Actively and passively gen- and the individual behind retroact with each other and erated data from digital devices of people is framed as differ according to who is zooming in, how and with resulting in scientifically measured, analysable proxies for which epistemology and ontology (Grommé, 2016; Loi, mental health and thus hold objective (and thus ultimate) 2019). truth. I would like to add that it is not just about the This assumption of having data being produced continu- amount of data (which is usually important in big data) ously and so close to the subject of interest (and thus, tied to and the collection by a digital device, but data that has the logic of accuracy) upholds the illusion of knowing a been collected close enough to the subject and so continu- deeper truth about this very subject. It claims to have ously that it seems as if it would be a direct window to access to all the relevant information about the data peek into real life. subject needed for certain assessments. This seemingly Looking into the depth of the type of data and in line exhaustive knowledge, continuously streamed from real with other authors I propose that for DP, truth is thought life and conceptualised as objective is the basis of the to be found in peoples’ behaviours: what time they get assumption that we will be better in deciding which treat- up, when and how long they sleep, when and in what ment options are ideal, even more so, we will be able to way they take part in social media activities, how often classify disease categories anew. In contrast, a critical the text/call, how much they move, etc. From a social social scientist perspective remarks that ‘data doubles’ science perspective, it is the digital traces of people that lead their own lives and sometimes do not have that may represent digitisable traces of pieces of everyday life much similarity with the person from whom they are practices. However, it is known that the social is digitalised derived and therefore defy the illusion of being a window only in a rudimentary form (Birk and Samuel, 2020). Also, to real life (Lyon, 2003: 22; Mager and Mayer, 2019). many aspects of our everyday lives and routines are not recognised by digital devices, as our complete experiences Consequences and future aspects with those practices are also not digitisable (Mau, 2017; Quinn, 2021). There are different ways to conceptualise PM will have different impacts on the field of medicine as the collected digital features. The data about a person has scientific discipline and healthcare depending on what is the been described as ‘data doubles’, ‘data traces’, ‘data silhou- centre of analysis: the field’s ontology and epistemology or ettes’ and ‘digital fingerprints’ (Lyon, 2003: 22; the patient. The logic behind data analysis is driven by Mindstrong Health, 2021; Quinn, 2021: 2). Loi (2019: mathematical and statistical aspects leading to a shift to cal- 158) criticises this sort of framing that assumes being a culate outcomes for the individual from population data. In one-to-one reflection of a person and suggests, digital the examples of breast cancer diagnosis, the sample of 8 Big Data & Society individuals with breast cancer is compared to knowledge 2019: 571). Big data is framed as a missing piece that about ‘genetic characteristics of a population sub-group’ might help to understand diseases in their ‘expression in (Low et al., 2018; Prainsack, 2017: 4). This logic of PM terms of the lived experience of individuals’ (Burnett, from the 1990s and 2000s meant, clinically collected data 2015; Huckvale et al., 2019; Jain et al., 2015: 463). The of certain subpopulations served to find targeted treatment continuous, in situ and live monitoring of a patient’s activ- for the individual (Prainsack, 2017). This logic stemming ities are stylised as the missing piece in understanding from epidemiology also pervades other medical disciplines. mental health. Additionally, PM has a ‘systems medicine’ approach when ‘Empowerment of the patient’ is a frequently found trope it comes to the object under investigation. The focus here in DP, hoping that people take agency over their health has been redirected from molecular or cell biology to which seems just another form of ‘responsibilisation’ systems and models of disease pathways. This development found within PM in general (Erikainen and Chan, 2019; presents an ontological and epistemological change in bio- Prainsack, 2017). However, it is still questionable who medicine (Erikainen and Chan, 2019). When centring the will profit from digitalised psychiatry and who will rather patient, there seems to be a twofold shift. One shift is see the downsides. While it has been shown that access from the individual to the population when it comes to to health information, e.g. through wearable use is empow- the logic of prediction, where population data is used for. ering for some people, usually those with more At the same time, there is a shift from population medicine resources, social minorities might find it rather disempow- to individualisation when it comes to how responsibility is ering (Birk and Samuel, 2020; Prainsack, 2017). DP is dis- framed. This means that rather than looking at population cussed to disregard social inequalities and conceptualise it disease risk, the discipline focuses on individual risk or rather as individual mental health conditions (Birk and individual prediction. This entails a shift from the popula- Samuel, 2020). Real-life monitoring might bring diagnostic tion to the individual regarding the responsibility for and therapeutic opportunities, however, it opens also the health. Consequently, the individual can more easily be possibility to surveillance (Dagum and Montag, 2019; made responsible, and population systemic aspects are not Lyon, 2003). Banner (2019: 7–8), e.g. describes how sur- being taken into account. Erikainen and Chan (2019: 314) veillance through DP might be used ‘to regulate, define describe this with ‘responsibilization’ of the individual and control’ certain populations such as POC or disabled (rather than medical professionals) which comes from people and ‘used to enforce neoliberal regimes of austerity’, being seen as individual with autonomy first and not as bringing more risk to historically discriminated populations part of a collective with a solidaristic arrangement than to others. (Prainsack, 2017). Additionally, healthcare shifts from Big data in general and related data in health, in particu- being reactive to being proactive, which can be seen by pre- lar, is considered to be the new gold. Health data is com- dictions that are thought to be possible based on genomic modified and profit is made from peoples’ health data data (Erikainen and Chan, 2019; Ferryman and Pitcan, (Banner, 2019; Cosgrove et al., 2020; Prainsack, 2017). 2018). Breast cancer is an example of that when Low Digital platforms are programmed with the company’s et al. (2018: 503) speak of ‘genomic profiling by clinical purpose of collecting as much data as possible and it is a sequencing’ as the next step to identify cancer risks in fairly frequent practice that data is sold and analysed for healthy individuals. This is already a reality for people completely distinct purposes compared to the primary with a strong family history of cancer who are screened reason for data collection. Cosgrove et al. (2020) lay out for BRAC1/2 gene mutations to decide about more frequent that 92% of mental health apps are known to share data check-ups, preventive measures, or specific forms of with third parties such as Facebook and Google without treatment. users having the choice to weigh in on that practice. This Similarly, for DP, experts of the field claim how DP is practice is called data repurposing and has been critiqued all about ‘person-centered care’ which also here results in from ethical and social scientist perspectives (Kitchin, more responsibility being given to each individual 2014). (Huckvale et al., 2019: 88). However, there are additional To recapitulate and expand: What we see in both PM and far-ranging consequences when analysing the processes DP is the assumption that knowing the truth is enhanced within DP. I will now focus on those related to data through the data-driven techniques of the field. For DP, types. Regarding the field’s ontology and epistemology the basis is continuous, in situ and live monitoring of every- DP even more so than PM is hoped to finally provide day life practices. For PM based on genetics and genomics, long longed for (digital) biomarkers for mental health con- it is genetic data marked as holding the ultimate truth about ditions. The RDoC funded in 2009 by the NIMH is one life and being extra precise. The digital form of data is an prominent example for the drive for a change of the essential aspect for the expressed logic because only then fieldaimingatmovingpsychiatrybeyondsofar ‘descrip- can necessary practices of collection, transfer, analysis tive diagnosis’ for mental health conditions with a new and interpretation be possible. Digitalisation, datafication, classification system based on ‘basic science’ (Rüppel, big data and their analysis provide additional information Baumgartner 9 to both fields to be able to claim being closer to their own When data is the primary focus, the rationale is usually construction of knowledge. With this newly generated that more data, and sometimes more precise data as in the knowledge, experts in the field claim that they can offer case of PM, is better. This is a standard rationale within more fitting treatment options and finally, better health. big data, even though it has been criticised extensively Ultimately, data is an essential component for knowledge also by the stakeholders involved (boyd and Crawford, production within the field through which the field seeks 2012; Kitchin, 2014). This means medicine and computer to differentiate itself from other, less digital-data-prone science and statistics now share this same mantra being fields. Being data-driven leads to further consequences, more data is better. To satisfy this need for data within such as the commodification of individual’s health data PM and DP, health institutions have to move further into and how through ‘personalisation’ responsibility shifts to the direction of digitalisation. This might bring new stake- the individual. holders, experts of data-intensive tools, into the picture who offer related products and solutions for problems. The value system of those new stakeholders will find a place in med- icine and health. One result will be a further commodifica- Discussion tion of disease and health and an increased dependence on While both PM and DP have a focus on digitisable data, I those entities that deal with data collection, analysis and have pointed out the differences and similarities of PM interpretation, which are often private companies (Lupton, and DP related to several aspects of data use. First, I have 2018b; Prainsack, 2017; Ruckenstein and Schüll, 2017; laid out the type of data and data producers: genetic data Saukko, 2018). deriving from lab facilities for PM versus big data from a Healthcare for patients may change extensively. variety of digital devices with sensors used in DP. Then I Individualisation seems to lead to responsibilisation. Care have shown the different systematic of data collection: might change from reactive to proactive and predictive. active and rare data production for PM versus How will it affect people to know that they are continuously moment-by-moment, in situ and live data collection for being monitored? To which recursive effects will this lead? DP. The focus on digital data also informs the dominant Digital sociologists and STS scholars discuss that it is understanding within the fields of PM and DP, and results unclear how much patients or so-called data subjects will in genetic determinism and advancing the genetic imagin- ultimately profit (Lupton, 2018a). Those people with ary for PM based on genetics and data fundamentalism in more resources might indeed be empowered by tailored DP. Both logics share the aspect of trust in digitisable treatment and access to their own health data and more data, which is thought to hold the ultimate truth leading information. However, critical data scientists point out to better health. With this logic, these two data-intensive that social minorities might not be the ones profiting but approaches uplift themselves over other less experience the downside: being monitored can just as digital-data-prone medical disciplines. Proof for this easily result in surveillance. People at the social margins aspect is the discursive distancing of experts of DP. They might be depersonalised and disempowered (Prainsack, frame digital traces as ‘objective’‘biomarkers’ for mental 2017). This opens an entirely new array of questions regard- illnesses against the ‘subjective’ standard of self-reports ing surveillance and what Foucault (2008: 1) called ‘biopo- of mental health symptoms which can also be due to the litics’, the establishment of state control over functions and unsuccessful search for genetic and molecular biomarkers processes of life. These aspects are especially salient for DP within psychiatry (Birk and Samuel, 2020; Brietzke et al., and will be even more salient once PM and DP are 2019: 223; Huckvale et al., 2019: 1). In medicine, as in combined. other scientific disciplines, objectivity ranks higher than Being data-driven approaches, many advantages, and subjectivity (Reiss and Sprenger, 2020). Information disadvantages of datafication and matters pertaining to big seems to gain in resilience and worth once it has been col- data and AI come into play. As critical scholars in social lected by a digital device (Lupton, 2015). Even more so if it and human science have shown, only certain knowledge has been collected close to the patient as is the case in DP. is quantifiable and thus digitisable. Many aspects of life ‘Real-life’ data is conceptualised as being ‘objective’ versus cannot be translated into numbers, such as experiential, self-reports being framed as ‘subjective’. Consequences intuitive, tactile and emotional knowledge (Quinn, 2021). might be a shift from reactive medicine to proactive and If data-driven methods gain more importance in the field preventive medicine resulting in individualisation and of health and illness, it is supposed to have huge conse- responsibilisation of the patient through both PM and DP. quences on basic epistemologies and ontologies around Although DP may offer more possibilities of exchange health and illness. Some of the non-quantifiable and non- and support for patients, the constant flow of ‘real-life’ digitasable aspects will not find their way in, e.g. decision data might open the door to more surveillance through con- support tools. Not only will the outcome be different if nected devices, which is the other side of 24 h monitoring they do not inform the decision, but they will also be that, e.g. Mindstrong proposes for its users (Metz, 2018). hidden and in the end, valued less (Lupton, 2015; 10 Big Data & Society Mennicken and Espeland, 2019; Quinn, 2021). This is even stemming from it. Since possibilities for data collection more critical as both fields are rich in descriptions how data have risen, PM is seeking to include all types of data to con- and knowledge gathered in PM and DP will change classi- struct a patient’s ‘unique thumbprint’ (Prainsack, 2017: 3). fication of diseases, identifying influences data-based From the side of DP efforts exist to introduce clinical data to approaches are expected to have on basic epistemologies the phenotype analysis, so-called ‘enriched data for DP’ and ontologies in medicine and healthcare. Therefore, crit- (Liang et al., 2019: 290). Both fields work closely together ical scholars should have a close eye on these paramount with data and computer science and share a similar big data changes. Prainsack (2017: 188) explains how the data logic. Thus, combining all data available, no matter if valuing logic in PM follows a ‘tacit hierarchy of utility, genetic and clinical data or ‘real-life data’, would be the with digital and computerable data on top, and unstructured, next logical step. Some experts already depict a horror sce- narrative, and qualitative evidence at the bottom’. A logic nario of genetic determinism in psychological diseases if that was introduced by the technologies used and not by biological and real-life data would be collected for mental professional experts. DP in particular shows an health conditions (Comfort, 2018). all-too-simplified view on mental health through its Additionally, the COVID-19 pandemic has shown us search for digital biomarkers based on behaviour pheno- how viral genetic information can be used to influence types. Birk and Samuel (2020) rightfully critique the every aspect of our life, including surveillance and quaran- usage of digital data as a proxy for social life for many tine restrictions. Also, it is thought to change our mental reasons. First, important knowledge will not be taken into healthcare, e.g. through the increased access to telehealth account and may ultimately be lost. Many non-quantifiable (Melcher et al., 2020). These transformations call for aspects have for a long time been essential for how deci- social science to engage in the field. There are many ques- sions have been made within medicine. Social science tions still to be asked: epistemological and ontological shows medicine had always had a more ‘human’ aspect to redirections in the field of medicine and healthcare, trans- it than science would claim. Practical knowledge, intuition formations in categories around health and illnesses with and nuances in the physician–patient relationship have new technologies, changes in doctor–patient relationships always been very important not just to how decisions and how institutions and people handle health-related around health and illness have been made, but also to data. Institutionally, it will be interesting to analyse power what constitutes the professions in health care (Kim et al., shifts amongst stakeholders and the changes of values and 2018; Spinnewijn et al., 2020). Second, if decisions are logics that accompany them. Last but not the least, addres- taken based on historical data, existing categories might sing questions around health equality and which conse- be reified and naturalised (Mau, 2017; Mennicken and quences these developments entail for population health Espeland, 2019; Quinn, 2021). This entails also assump- and individual patients is crucial in these changing times. tions around normality and bias in algorithms which is harder to dismantle because of black box phenomena Acknowledgements (Birk and Samuel, 2020). The author would like to thank three anonymous reviewers for Finally, all the processes described before can also be their very insightful comments and Tamara Schwertel, Regina conceptualised as different aspects of biomedicalisation, Ammicht-Quinn and Ursula Offenberger for their valuable feed- the transformance of biomedicine by technoscientific inter- back. Furthermore, she would like to thank Mrunmayee Sathye ventions such as computers and genetics, with the conse- and Lukas Haeberle for their help on the manuscript. quences that the biological gains in importance. In the analysis before I found the following processes of biomedi- Declaration of conflicting interests calisation: a focus on health, surveillance and risk which we The author declared no potential conflicts of interest with respect find in PM (genomic profiling) but above all in DP, then the to the research, authorship and/or publication of this article. ‘technoscientisation of biomedicine’ which can be found in both logics of PM and DP, and a change in biomedical Funding knowledge represented by the aim for new disease classifi- The author received no additional financial support for the cations through data in both PM and DP (Clarke et al., research, authorship and/or publication of this article. 2003: 166). Also, the responsibilisation of the individual in both PM and DP and commodifying health data in DP ORCID iD fits under this umbrella term. Renate Baumgartner https://orcid.org/0000-0002-3401-1870 Conclusion and outlook Notes PM and DP have the potential and are already changing 1. For example, the mandatory CLIA certificate in the United medicine and healthcare. The basis for these developments States for laboratories performing tests of human specimen’s is also the digital data used and the logics of the field diagnosis, prevention or treatment of disease or health Baumgartner 11 problems. Genetics tests are classified as moderate or high the-hidden-biases-in-big-data. (last accessed 6 December complexity tests. 2021). 2. Erikainen and Chan (2019) also describe how this argumenta- Dagum P (2019) Digital brain biomarkers of human cognition and tion leads to big money flows to PM. The construction of truth mood. In: Baumeister H and Montag C (eds) Digital with the production of ‘hope’ within the field, in the end, is able Phenotyping and Mobile Sensing: New Developments in to accumulate different forms of capital. Psychoinformatics. Cham: Springer Nature Switzerland AG, 3. Some practices in DP involve both DP and subjective assess- pp.93–108. ments. Smartphones can be used for ecological momentary Dagum P and Montag C (2019) Ethical considerations of digital phe- assessments which repeatedly sample behaviours, thoughts notyping from the perspective of a healthcare practitioner. In: and feelings in real time (Garcia-Ceja et al., 2018). Baumeister H and Montag C (eds) Digital Phenotyping and 4. The aspect of being able to predict mental health issues and the Mobile Sensing: New Developments in Psychoinformatics. consequences thereof is an important topic that asks to be dis- Cham: Springer Nature Switzerland AG, pp.13–29. cussed, but which exceeds the focus of this paper. Erikainen S and Chan S (2019) Contested futures: Envisioning 5. Invoking being able to monitor real life might also be due to the “personalized,”“stratified,” and “precision” medicine. New frustration in psychiatry of not having sufficient diagnosis and Genetics and Society 38(3): 308–330. treatment tools for people with mental health conditions Ferryman K and Pitcan M (2018) What is precision medicineData (Mindstrong Health, 2021). & Society. Foucault M (2008) The Birth of Biopolitics. Lectures at the Collège de France, 1978–79. Hampshire and New York: Palgrave Macmillan. References Garcia-Ceja E, Riegler M, Nordgreen T, et al. (2018) Mental Banner O (2019) Technopsyence and Afro-Surrealism’s cripis- health monitoring with multimodal sensing and machine learn- temologies. Catalyst: Feminism, Theory, Technoscience 5(1): ing: A survey. Pervasive and Mobile Computing 51: 1–26. 1–29. Grommé F (2016) Data mining ‘problem youth’:Looking closer Barnett I, Torous J, Staples P, et al. (2018) Relapse prediction in But Not seeing better. In: van der Ploeg I and Pridmore J schizophrenia through digital phenotyping: A pilot study. (eds) Digitizing Identities. New York; London: Routledge Neuropsychopharmacology 43(8): 1660–1666. Taylor & Francis, pp.163–183. DOI: 10.4324/9781315756400. Birk RH and Samuel G (2020) Can digital data diagnose mental Hood L, Balling R and Auffray C (2012) Revolutionizing medi- health problems? A sociological exploration of ‘digital pheno- cine in the 21st century through systems approaches. typing’. Sociology of Health and Illness 42(8): 1873–1887. Biotechnology Journal 7(8): 992–1001. boyd D and Crawford K (2012) Critical questions for big data. Hopman R (2020) Opening up forensic DNA phenotyping: The Information, Communication & Society 15(5): 662–649. logics of accuracy, commonality and valuing. New Genetics Brietzke E, Hawken ER, Idzikowski M, et al. (2019) Integrating and Society 39(4): 424–440. digital phenotyping in clinical characterization of individuals Huckvale K, Venkatesh S and Christensen H (2019) Toward with mood disorders. Neuroscience and Biobehavioral clinical digital phenotyping: A timely opportunity to Reviews (104): 223–230. DOI: 10.1016/j.neubiorev.2019.07. consider purpose, quality, and safety. npj Digital Medicine 009. 2(1): 1–11. Burnett F (2015) Four questions for NIMH Director Thomas Jain SH, Powers BW, Hawkins JB, et al. (2015) The digital phe- R. Insel. Available at: https://ct.counseling.org/2015/03/four- notype. Nature Biotechnology 33(5): 462–463. questions-for-nimh-director-thomas-r-insel/ (last accessed 6 Keller EF (2002) The Century of the Gene, 3rd ed. Cambridge and December 2021). London: Harvard University Press. Cedars-Sinai Blog (2019) New Recommendations for BRCA Kim K, Heinze K, Xu J, et al. (2018) Theories of Testing: Should You Be Screened? Available at: https:// health care decision making at the end of life: A www.cedars-sinai.org/newsroom/new-federal-guidelines-for- meta-ethnography. Western Journal of Nursing Research brca-testing-should-you-be-screened/. (last accessed 6 40(12): 1861–1884. December 2021). Kitchin R (2014) Big data, new epistemologies and paradigm Clarke AE, Shim JK, Mamo L, et al. (2003) Biomedicalization: shifts. Big Data and Society 1(1): 1–12. Technoscientific transformations of healht, illness, and U.S. Kumari S, Chouhan U and Suryawanshi SK (2017) Machine biomedicine. American Sociological Review 68(2): 161–194. learning approaches to study HIV / AIDS. Bioscience Comfort N (2018) Genetic determinism. Nature 561: 461–463. Biotechnology Research Communications 10(1): 34–43. Cooper R and Paneth N (2020) Will precision medicine lead to a Liang Y, Zheng X and Zeng DD (2019) A survey on big data- healthier population? Issues in Science and Technology driven digital phenotyping of mental health. Information XXXVI(2): 64–71. Available at: https://issues.org/precision- Fusion 52: 290–307. medicine/. Loi M (2019) The digital phenotype: A philosophical and ethical Cosgrove L, Karter JM, McGinley M, et al. (2020) Digital pheno- exploration. Philosophy & Technology 32: 155–171. typing and digital psychotropic drugs: Mental health surveil- Low SK, Zembutsu H and Nakamura Y (2018) Breast cancer: The lance tools that threaten human rights. Health and Human translation of big genomic data to cancer precision medicine. Rights 22(2): 33–40. Cancer Science 109(3): 497–506. Crawford K (2013) The hidden biases in big data. Havard Lupton D (2015) Digital Sociology. London and New York: Business Review April. Available at: https://hbr.org/2013/04/ Routledge. 12 Big Data & Society Lupton D (2016) The Quantified Self, 1st ed. Cambridge: Polity Prainsack B (2017) Personalized Medicine. Empowered Patients Press. in the 21st Century? New York: New York University Press. Lupton D (2018a) Digital Health – Critical and Cross-Disciplinary Quinn RA (2021) Artificial intelligence and the role of ethics. Perspectives. London and New York: Routledge. Statistical Journal of the IAOS 37(1): 75–77. Lupton D (2018b) How do data come to matter? Living and Ranjan Y, Rashid Z, Stewart C, et al. (2019) Radar-base: Open becoming with personal data. Big Data and Society 5(2): 1–11. source mobile health platform for collecting, monitoring, and Lyon D (2003) Surveillance as Social Sorting. London and analyzing data using sensors, wearables, and mobile devices. New York: Routledge. JMIR mHealth and UHealth 7(8): 1–12. Mennicken A and Espeland WN. (2019) What’s new with Reiss J and Sprenger J (2020) Scientific objectivity. Stanford numbers? Sociological approaches to the study of quantifica- Encyclopedia of Philosophy. Winter 20. Metaphysics tion. Annual Review of Sociology 45: 223–245. Research Lab, Stanford University. Available at: https://plato. McLean S, Ressel K, Koenen KC, et al. (2020) The AURORA stanford.edu/archives/win2020/entries/scientific-objectivity/. study: A longitudinal, multimodal library of brain biology Ruckenstein M and Schüll ND (2017) The datafication of health. and function after traumatic stress exposure. Molecular Annual Review of Anthropology 46: 261–278. Psychiatry 25(2): 283–296. Rüppel J (2019) “Now is a time for optimism”: The politics of Mager A and Mayer K (2019) Body data—data body: Tracing personalized medicine in mental health ambiguous trajectories of data bodies between empower- research. Science Technology and Human Values 44(4): 581–611. ment and social control in the context of health. Saeb S, Lattie EG, Schueller SM, et al. (2016) The relationship Momentum Quarterly – Zeitschrift für sozialen Fortschritt between mobile location sensor data and depression 8(2): 95–109. symptom severity. PeerJ 4: e2537. Mau S (2017) Das Metrische Wir. Über Die Quantifizierung Des Saukko P (2018) Digital health – a new medical cosmology? The Sozialen. Berlin: Suhrkamp Verlag. case of 23andMe online genetic testing platform. Sociology of Melcher J, Hays R and Torous J (2020) Digital phenotyping for Health and Illness 40(8): 1312–1326. mental health of college students: A clinical review. Spinnewijn L, Aarts J, Verschuur S, et al. (2020) Knowing what Evidence-based Mental Health 23(4): 161–166. the patient wants: A hospital ethnography studying physician Metaxiotis KS and Samouilidis JE (2000) Expert systems in med- culture in shared decision making in the Netherlands. BMJ icine: Academic illusion or real power? Information Open 10: e032921. Management and Computer Security 8(2–3): 75–79. Topol EJ (2019) High-performance medicine: The convergence Metz R (2018) The smartphone app that can tell you’re depressed of human and artificial intelligence. Nature Medicine 25: 44–56. before you know it yourself. Available at: https://www. Torous J, Kiang MV, Lorme J, et al. (2016) New tools for new technologyreview.com/s/612266/the-smartphone-app-that- research in psychiatry: A scalable and customizable platform can-tell-youre-depressed-before-you-know-it-yourself/. (last to empower data driven smartphone research. JMIR Mental accessed 6 December 2021). Health 3(2): 16. Milne R (2020) The rare and the common: Scale and the genetic U.S. Government (2021) Electronic Code of Federal Regulations: imaginary in Alzheimer’s disease drug development. New Part 493 – Laboratory Requirements. Available at: https:// Genetics and Society 39(1): 101–126. www.ecfr.gov/current/title-42/chapter-IV/subchapter-G/part- Mindstrong Health (2021) Using science to help us seek the truth. 493 (last accessed 6 December 2021). Available at: https://mindstrong.com/science/ (last accessed 6 van Dijck J (2014) Datafication, dataism and dataveillance: Big December 2021). data between scientific paradigm and ideology. Surveillance Mol A (2008) The Logic of Care. London: Routledge. and Society 12(2): 197–208. National Human Genome Research Institute (n.d.) The CLIA Weiner K, Martin P, Richards M, et al. (2017) Have we seen the Framework. Available at: https://www.genome.gov/Pages/ geneticisation of society? Expectations and evidence. PolicyEthics/GeneticTesting/The_CLIA_Framework.pdf (last Sociology of Health and Illness 39(7): 989–1004. accessed 6 December 2021). Weiss KM (2017) Is precision medicine possible? Issues in Peters T (2012) Playing God? Genetic Determinsm and Human Science and Technology 34(1). Available at: http://issues.org/ Freedom, 2nd ed. London and New York: Routledge. 34-1/is-precision-medicine-possible/.

Journal

Big Data & Society – SAGE

Published: Dec 22, 2021

Keywords: Personalised medicine; genetic determinism; data fundamentalism; data-driven medicine; big data; digital health

There are no references for this article.