Addressing ethical issues in multimodal integration of digital biomarkers: A rapid review (Preprint)

BACKGROUND In recent years, digital biomarkers have become increasingly popular in research and clinical investigations for therapeutic interventions. Traditionally, biomarkers are classified as physiological, pathological, and anatomical agents with specific features or characteristics for biological processes, including monitoring mechanisms and assays. As new generation of biomarkers, wearable technologies, or mobile health (mHealth) are rapidly expanding as outpatient tools for early disease screening and diagnosis. Digital biomarkers are non-imaging (non-invasive) sensors that are used to collect real-time data, perform measurements and/or enter metrics value in portable devices such as smartphones, and immersive technologies for alternatives to conventional biomarkers. When pairing with other advanced technologies including machine learning, and algorithm, multimodal biomarkers are proven efficient for biological monitoring including cognitive and behavioural changes. However, the socio-technological aspects are not well studied, and no specific policy exists for data collection, management and governance. Recent research raised several concerns relating to digital biomarkers including ethical issues, data privacy and security and data governance. This paper presents a short review highlighting the critical issues with digital biomarkers, considering their proliferation in healthcare. OBJECTIVE This research presents an in-depth analysis with critical issues surrounding non-imaging biomarkers. Drawing from empirical data, the literature is scoped systematically to present an in-depth analysis where the issues surrounded non-digital biomarkers are highlighted and examined METHODS Comprehensive electronic searches were conducted in the literature as a first steps, following a five-stages approach for scoping method in social science described by Arksey and O’Malley, (2005). Data from six public databases were analysed searched. Key words: “non imaging biomarker” “safety”, “privacy”, “security” narrow the search to focus relevant issues addressing the problem questions that retrieve and synthesise existing data; in addition to “ethics”, and “social” concerns. Our selection purpose was meant to explore the literature and summarise findings and identifying gaps. RESULTS Emerging App-based devices exploit behaviour patterns to predict or response to the user’s cognitive states or cognitive manipulation. Data collected can potentially be used to trigger mental health interventions. Table 2 highlighted the four types of risks identified (1) ethical concerns, (2) privacy, (3) security/cyber breach, (4) data safeguarding/data governance, and (5) policy/regulation/governance (Table 1). The risk of data hacking through data storage sites increased significantly with open platforms available to users, including those who are inexperienced and vulnerable users. CONCLUSIONS Digital biomarkers enable individuals to remotely personalized and self-assessed their condition for better health management. Between 2014 and 2023, the digital market has seen a boom in digital health devices ranging Portable smartphone App connected wearable devices to neuromodulating devices. However, the socio-technological aspects are not well studied, and no specific policy exists for data collection, management and governance. Recent research raised several concerns relating to digital biomarkers including ethical issues, data privacy and security and data governance. CLINICALTRIAL N/A