Epidemiology Data Collection Methods

7 wearable and sensor innovations pushing health beyond “wellness” tracking this month: 🔘 Sibel Health is developing an AI-enabled wearable that tracks scratching behaviour in people with atopic dermatitis, turning something usually seen as a subjective symptom into a measurable clinical signal that could also support drug development. 🔘 CranioSense is working on a non-invasive approach to measuring intracranial pressure, which today often requires invasive procedures, and if validated could make brain pressure monitoring safer and more continuous in routine clinical care. 🔘 University of Technology Sydney researchers are developing AI-powered sweat sensors that can decode body chemistry in real time, tracking hormones, medication levels and potential early warning signs of disease, potentially offering a non-invasive alternative to some forms of blood testing 🔘 ŌURA rings are being used within Medicare Advantage Plans, with around one-third of eligible members opting in and sharing biometric data, which is already leading to improvements in sleep and light activity and is paving the way for deeper clinical use cases such as hypertension monitoring 🔘 Samsung Electronics is preparing to launch an AI Brain Health tool that uses data from smartphones and wearables, including speech, movement and sleep behaviour, to help detect early signs of dementia while aiming to keep the experience privacy-aware and clinically relevant 🔘 Researchers at the University of Arizona have created a wearable mesh sleeve that monitors gait and subtle movement patterns to identify early signs of frailty in older adults, with the goal of shifting care from reacting after a fall to proactively supporting prevention through continuous remote monitoring 🔘 And China is testing “smart urinals” that analyse urine in real time for markers like glucose and protein, which opens up interesting conversations about passive health screening, consent, and how health data might be gathered in everyday environments. 💬We are steadily moving from episodic health snapshots to passive, continuous and contextual signals across movement, sleep, behaviour and even body chemistry. The technology is getting closer. Now the real work is around validation, governance, reimbursement and making sure the data actually makes a difference in peoples lives 👇 Links to articles in comments #DigitalHealth #Wearables #AI

This reflection paper by the European Medicines Agency provides guidance on the methodological aspects of using real-world data (#RWD) in non-interventional studies (#NIS) to generate real-world evidence (#RWE) for #regulatory purposes. It emphasizes the importance of carefully selecting study designs based on the research question. Studies are classified as either descriptive, focusing on patient characteristics and patterns, or causal, aiming to infer treatment effects. The framework of target #trial_emulation is recommended to improve the internal validity of causal studies by mimicking randomized controlled trials as closely as possible using observational data. A key concern throughout is addressing sources of #bias and #confounding. The paper highlights different types of bias—selection bias, information bias, and time-related bias—and stresses the need to prevent or mitigate these issues during study design rather than trying to correct them post hoc. Special attention is given to confounding, recommending careful identification and handling of confounders through study design choices like active comparators and new user designs, and the use of control exposures or outcomes where appropriate. Effect modification is also discussed in the context of ensuring the generalisability of findings. #Governance and #transparency are critical elements, with the paper advocating adherence to the ENCePP Code of Conduct and EU data protection regulations. It calls for clear study registration, public disclosure of protocols and results, and sharing of analytical codes to enhance reproducibility and trust. #Data_quality is another major focus, with emphasis on evaluating both the reliability (accuracy, completeness, credibility) and relevance (fitness for the research question) of RWD sources. The use of data quality frameworks, transparent validation of data elements, and careful handling of data linkage across sources are recommended to ensure robust evidence generation. Finally, the paper outlines expectations for statistical analyses, stressing pre-specified analysis plans, model transparency, and robust handling of missing data, confounding, and heterogeneity. It advises moving beyond p-values to focus on effect estimates and their clinical relevance, using sensitivity and stratified analyses to assess the robustness of findings. By integrating these principles, the paper aims to improve the quality and regulatory acceptability of RWE derived from NIS. Disclaimer: The opinions shared are solely my own and not express the views or opinions of any of my employers.

3 Healthcare AI papers I'm reviewing today 1. 📚AI in Medicine: Medical multimodal foundation models in clinical diagnosis and treatment: Applications, challenges, and future directions https://lnkd.in/eNV5CDSp Medical Multimodal Foundation Models (MMFMs) combine diverse data types (imaging, text, labs) to improve diagnosis, treatment planning, and precision medicine. Recent advances in large datasets and multimodal architectures (vision-only and vision-language) enable strong generalization across tasks like segmentation, classification, and clinical report generation. Key opportunities lie in holistic integration of multi-organ/multimodal data, but challenges remain in optimizing representations and scaling real-world clinical adoption. 2. 📚 BMJ Dig Health & AI - Optimising large language models for clinical information extraction: a benchmarking study in the context of ulcerative colitis research https://lnkd.in/erXhhX9k This study compared open-source and closed-source LLMs for extracting the Mayo Endoscopic Subscore from colonoscopy reports. It found that QLoRA fine-tuning improves open-source performance significantly, but GPT-4o with prompt engineering still outperforms them by 5–11% and is more cost-effective. Overall, GPT-4o is the most efficient option today, while QLoRA-optimized open-source models are viable fallbacks, though both leave room for improvement in instruction following. 3. 📚 JAMIA Open - Generative artificial intelligence for automated data extraction from unstructured medical text https://lnkd.in/ekfy8-VX A GenAI pipeline using an open-source LLM was developed to extract structured data from right heart catheterization notes with built-in guardrails and a retry mechanism. It achieved high performance (99% precision, 85% recall, 91.5% F1, 90% accuracy), with missed values as the main error and hallucinations extremely rare (<0.01%). The study shows LLM pipelines can reliably mine unstructured clinical data, improving research efficiency and clinical applications.

Your Wardrobe Goes Online From step counters to fart monitors, wearables are changing the epistemology of medicine itself We began with step counters. Then heart rhythms. Then sleep cycles. Then blood oxygen. Then glucose. Now? Farts. Scientists at the University of Maryland are piloting what they jokingly call a "Fitbit for farts"—a tiny hydrogen sensor worn discreetly on the body that continuously measures flatulence. It sounds like late-night comedy. It is, in fact, serious #gastroenterology. And it signals something much larger. This device sits at the crossroads of three powerful trends: extreme miniaturization, continuous monitoring, and edge computing. The same supply chains that gave us smart rings, smart watches, and wireless earbuds now enable a battery-powered sensor small enough to measure something we've never systematically measured before: baseline digestive patterns. Forty percent of American adults report regular digestive disruption. Fiber-rich diets, which reduce colon cancer risk, are often abandoned because of bloating and gas. Yet in 2026, we still don't know how often the average person passes gas in a day. Early data offers a hint at the range: one participant logged 175 emissions. For decades, digestive health relied on self-reporting and invasive lab work. Now we are entering an era of passive, ambient #health telemetry. The #AppleWatch moved the cardiology ward to your wrist. Continuous glucose monitors brought the endocrinology lab to your arm. Each time, the shift was the same: from episodic snapshot to living dashboard. When you move from occasional measurement to continuous signal detection, you don't just gather more data—you change the epistemology of medicine itself. Patterns that were previously invisible become legible. Causation, not just correlation, becomes possible. This is just the beginning. Imagine clothing that tracks inflammatory markers. Glasses that monitor neurological drift. Belts—we make a few of those at Randa Apparel & Accessories—that detect posture, waist measurement, and metabolic change. Fabrics embedded with biosensors that surface early-stage disease before symptoms arrive. Your wardrobe becomes diagnostic infrastructure. Real questions follow: #data ownership, #privacy, psychological burden, the quiet anxiety of living with a dashboard of yourself. Continuous monitoring can empower patients, or produce a nation of worried well, over-interpreting every signal. These are not small concerns. But the direction is unmistakable. #Healthcare is migrating from hospitals to homes to bodies. From appointments to algorithms. From episodic to continuous. Technology has always moved closer, first to our pockets with smartphones, then our wrists. Now it is woven into the textiles we wear and clipped discreetly where biology actually happens. Fitbit (now part of Google) and ŌURA are not the destination. They are the prologue. Walt Whitman sang the body electric. We're adding sensors.

𝗗𝗼 𝘆𝗼𝘂 𝗳𝗲𝗲𝗹 𝗪𝗲𝗮𝗿𝗮𝗯𝗹𝗲𝘀 𝗮𝗿𝗲 𝗹𝗶𝘃𝗶𝗻𝗴 𝘂𝗽 𝘁𝗼 𝘁𝗵𝗲𝗶𝗿 𝗲𝘅𝗽𝗲𝗰𝘁𝗮𝘁𝗶𝗼𝗻𝘀? Because this breakthrough just raised the bar significantly. Researchers have developed a wearable device that monitors glucose levels through sweat – and it doesn't stop there. This disposable patch integrates real-time glucose monitoring with automated transdermal drug delivery for diabetes management. 𝗛𝗲𝗿𝗲'𝘀 𝘄𝗵𝘆 𝘁𝗵𝗶𝘀 𝗺𝗮𝘁𝘁𝗲𝗿𝘀: ✅ 𝗡𝗼𝗻-𝗶𝗻𝘃𝗮𝘀𝗶𝘃𝗲 𝗺𝗼𝗻𝗶𝘁𝗼𝗿𝗶𝗻𝗴 – No more painful finger pricks ✅ 𝗖𝗼𝗻𝘁𝗶𝗻𝘂𝗼𝘂𝘀 𝘁𝗿𝗮𝗰𝗸𝗶𝗻𝗴 – Real-time glucose data from sweat analysis ✅ 𝗦𝗺𝗮𝗿𝘁 𝗱𝗿𝘂𝗴 𝗱𝗲𝗹𝗶𝘃𝗲𝗿𝘆 – Automated treatment response when glucose levels spike ✅ 𝗪𝗲𝗮𝗿𝗮𝗯𝗹𝗲 & 𝗱𝗶𝘀𝗽𝗼𝘀𝗮𝗯𝗹𝗲 – Practical for everyday use This isn't science fiction. It's soft bioelectronics on human skin, creating a closed-loop system that monitors AND treats diabetes simultaneously. 𝗧𝗵𝗲 𝗯𝗶𝗴𝗴𝗲𝗿 𝗽𝗶𝗰𝘁𝘂𝗿𝗲? This technology represents the convergence of AI, wearables, and connected care – three pillars transforming healthcare delivery. We're moving from reactive treatment to proactive, personalized health management. For healthcare organizations exploring digital transformation, innovations like this answer the question: wearables aren't just living up to expectations – they're exceeding them by becoming active treatment devices, not just passive monitors. 𝗧𝗵𝗲 𝗿𝗲𝗮𝗹 𝗾𝘂𝗲𝘀𝘁𝗶𝗼𝗻: 𝗛𝗼𝘄 𝗾𝘂𝗶𝗰𝗸𝗹𝘆 𝗰𝗮𝗻 𝘄𝗲 𝘀𝗰𝗮𝗹𝗲 𝘁𝗵𝗶𝘀 𝗳𝗿𝗼𝗺 𝗿𝗲𝘀𝗲𝗮𝗿𝗰𝗵 𝘁𝗼 𝗿𝗲𝗮𝗹-𝘄𝗼𝗿𝗹𝗱 𝗰𝗮𝗿𝗲? What's your take? Are non-invasive biosensors the next frontier in chronic disease management? 📖 Research: Science Advances (DOI: 10.1126/sciadv.1601314) For more Wearables News and Expertise follow João Bocas #DigitalHealth #AIinHealthcare #Wearables #DiabetesCare #HealthTech #ConnectedCare #Innovation #HealthcareTransformation

One of the biggest opportunities in health is moving from episodic measurement to more continuous, real-world understanding. That’s why this new PLOS Digital Health paper is so meaningful. Led by Michael Chee and his team at the National University of Singapore, the study found that nocturnal PPG signals collected from Oura Ring could estimate vascular age with performance comparable to a clinical-grade fingertip sensor in a cohort of 160 healthy adults. Vascular aging is a key marker for cardiovascular risk, but the traditional ways to measure it are often expensive and hard to scale. Research like this helps show how consumer wearables may expand access to longitudinal, real-world health insights in a way that is more practical and more accessible over time. What also stands out to me is that these signals were captured during sleep, when physiological data can be gathered passively and consistently. At Oura, we’ve long known that sleep is one of the clearest windows into overall health, and research like this proves it. It also points to a more proactive model of health—one built not just on isolated clinical snapshots, but on patterns measured over time. Learn more on the Pulse blog: https://lnkd.in/gmi8U5AK  

This new paper by Sergio Consoli et al explores how generative AI can transform unstructured outbreak data into structured, searchable knowledge. The team developed an epidemiological knowledge graph (eKG) using WHO Disease Outbreak News (DONs), applying an ensemble of large language models to extract details such as disease name, country, date, and number of cases or deaths. The researchers used open-source models including Mistral, Zephyr, and Meta-Llama to extract information from over 3,000 outbreak reports. They structured this data into a FAIR-compliant knowledge graph, linking it with biomedical and geographic ontologies. The resulting resource—comprising nearly 3,000 outbreak events—is now publicly accessible via SPARQL endpoints and visualization tools. This matters because many official outbreak reports remain locked in prose, making them difficult to analyze at scale. With eKG, public health professionals can conduct detailed, structured queries across decades of global outbreak data. This significantly improves our ability to track, analyze, and respond to emerging health threats. The big takeaway: AI can unlock the full value of legacy outbreak data by transforming it into structured, interoperable formats that support real-time analysis and response. This approach opens new possibilities for integrating informal sources like news and social media into formal disease surveillance systems, advancing global preparedness and early warning capabilities. https://lnkd.in/ePc54yvQ #GlobalHealth #PathogenSurveillance #HealthInnovation #PublicHealth

Data quality lies at the core of effective decision-making in health systems, directly impacting the reliability of monitoring, evaluation, and policy formulation. This Comparative Analysis of Data Quality Assessment Tools, developed under the MEASURE Evaluation project, offers a detailed examination of tools designed to enhance data quality in global health programs. With a focus on HIV, tuberculosis, and malaria, the guide provides an indispensable resource for health practitioners, program managers, and policymakers aiming to address the persistent challenges of data inaccuracies. Through a structured comparison of tools such as the Data Quality Audit (DQA), Routine Data Quality Assessment (RDQA), and Data Quality Review (DQR), this document highlights their unique objectives, methodologies, and practical applications. By offering insights into tool selection and implementation, it equips stakeholders to choose the most appropriate approaches for their specific programmatic needs. Furthermore, it emphasizes the importance of integrating these tools within broader health information systems to ensure sustainability and scalability. This resource empowers its audience to move beyond traditional methods of data collection and validation. By leveraging these tools, health programs can achieve greater transparency, accountability, and efficiency, ultimately driving improved outcomes for global health initiatives.

📊 Most medical data lives in unstructured clinical notes, and too often, it not used. Every day, physicians write progress notes, imaging reports, and pathology summaries filled with signals: treatment responses, adverse events, even early signs of success or failure. But buried in free text, this information is unusable for traditional analysis. Clinical trials remain the gold standard, but they’re expensive, slow, and often unfeasible. So how do we unlock the evidence hidden in messy medical records? That’s what TRIALSCOPE set out to do. 🔹 By combining biomedical language models, probabilistic approaches, inferences, TRIALSCOPE automatically structured EMR data from over 1 million cancer patients. 🔹 It reproduced results of published lung cancer trials, generalized to pancreatic cancer, and simulated studies. 🔹 Compared with manual curation, it achieved >20× faster processing and 10× lower cost. -> Clinical text isn’t noise. With the right tools, it’s raw data waiting to become real world evidence.

Wearable technology in healthcare isn’t inherently an asset or a liability. It’s a force multiplier—and the value it creates depends entirely on whether the system around it is designed to absorb, interpret, and act on continuous data responsibly. From a clinical standpoint, the upside is clear. Wearables have enabled earlier detection of conditions like atrial fibrillation and sleep-disordered breathing, created longitudinal visibility that traditional encounters never allowed, and increased patient engagement by making health more tangible in daily life. But the system-level limitations are becoming increasingly apparent: false positives, non-actionable data, workflow burden, clinician fatigue, and heightened patient anxiety—often without a proportional improvement in outcomes. This isn’t fundamentally a technology problem. It’s a system design, governance, and accountability problem. When data ownership is unclear, action thresholds are undefined, and clinicians are expected to manage continuous signal streams without infrastructure or incentives, wearables risk shifting healthcare from prevention to noise management. The real question isn’t whether patients should use wearables. It’s whether our healthcare systems are ready to integrate continuous diagnostic data—across conditions—without overmedicalizing everyday life or overwhelming the workforce meant to interpret it. #Cardiology #WearableTech