Trend AnalysisMedicine & Health
Wearable Sensors and Remote Patient Monitoring: Digital Health Meets Chronic Disease
Chronic diseases (diabetes, heart failure, COPD, CKD) account for 74% of global deaths and consume ~86% of US healthcare spending (and a substantial majority of healthcare spending in high-income coun...
By Sean K.S. Shin
This blog summarizes research trends based on published paper abstracts. Specific numbers or findings may contain inaccuracies. For scholarly rigor, always consult the original papers cited in each post.
The Question
Chronic diseases (diabetes, heart failure, COPD, CKD) account for 74% of global deaths and consume ~86% of US healthcare spending (and a substantial majority of healthcare spending in high-income countries more broadly). Traditional management relies on periodic clinic visits โ snapshots of health status separated by weeks or months. Wearable sensors and remote patient monitoring (RPM) promise continuous, real-time physiological data streams that could detect deterioration before symptoms appear and enable proactive intervention. But does continuous monitoring actually improve clinical outcomes, or does it merely generate data that overwhelms clinicians and creates false alarms?
Landscape
Ghadi et al. (2025) reviewed the integration of wearable technology with AI for remote patient care. Their analysis identifies key barriers to adoptionโdata privacy constraints, EHR integration challenges, digital literacy gaps, and data overloadโand proposes AI-driven solutions including federated learning (for privacy-preserving model training across institutions) and deep learning frameworks (for intelligent alerting that filters clinically actionable signals from raw sensor streams). The core argument is that moving from "data transmission" to "intelligent alerting" is necessary to address alert fatigue and make RPM clinically sustainable.
Deng et al. (2025) reviewed digital health specifically for chronic kidney disease (CKD), where eGFR decline is gradual and often asymptomatic until advanced stages. They documented how wearable biosensors (continuous blood pressure monitors, hydration sensors) combined with mobile health apps and AI decision support can detect CKD progression earlier than quarterly lab visits, potentially slowing decline through timely medication adjustment.
Stradford et al. (2024) describe the WEAR study protocolโa prospective observational study combining wearable activity trackers with electronic patient-reported outcomes (ePROs) in rheumatoid arthritis patients initiating upadacitinib or adalimumab. The study is designed to test whether objective activity data from wearables adds information beyond self-reported disease activity scores; results are forthcoming as data collection completes.
Daich Varela et al. (2024) extended RPM to ophthalmology, reviewing home-based monitoring devices for retinal diseases. For conditions like diabetic retinopathy and age-related macular degeneration, where visual function changes can be detected before irreversible damage occurs, home monitoring enables earlier treatment initiation.
Key Claims & Evidence
<
| Claim | Evidence | Verdict |
|---|
| AI-enabled wearables reduce alert fatigue vs. raw data transmission | Edge AI processing generates clinically actionable alerts only (Ghadi et al. 2025) | Supported in design; clinical outcome evidence still maturing |
| Wearable monitoring detects CKD progression earlier than clinic visits | Continuous BP and hydration data capture subclinical changes (Deng et al. 2025) | Promising; large RCTs needed to confirm |
| Wearable activity data complements patient-reported outcomes | WEAR study designed to test this hypothesis (Stradford et al. 2024); results pending | โ ๏ธ Hypothesis โ study protocol paper; results not yet published |
| Home monitoring enables earlier retinal disease treatment | Self-monitoring devices detect visual changes before scheduled visits (Daich Varela et al. 2024) | Supported; patient adherence to home monitoring varies |
Open Questions
Clinical outcome evidence: Most RPM studies measure process outcomes (hospitalisation rates, time to detection). Do they improve hard outcomes (mortality, disability-free survival)?
Health equity: Wearable RPM requires smartphones, internet access, and digital literacy. How should health systems ensure equitable access?
Data governance: Continuous health monitoring generates vast personal datasets. Who owns this data, and how should it be protected against commercial exploitation?
Reimbursement models: Healthcare payers are slowly adopting RPM reimbursement codes, but coverage varies widely. Can standardised cost-effectiveness evidence accelerate adoption?Referenced Papers
- [1] Ghadi, Y. et al. (2025). Integration of wearable technology and AI in digital health for remote patient care. J. Cloud Computing. DOI: 10.1186/s13677-025-00759-4
- [2] Daich Varela, M. et al. (2024). Digital health and wearable devices for retinal disease monitoring. Graefe's Archive. DOI: 10.1007/s00417-024-06634-3
- [3] Stradford, L. et al. (2024). WEAR study: wearable activity tracker in rheumatoid arthritis. Contemporary Clinical Trials Communications. DOI: 10.1016/j.conctc.2024.101272
- [4] Deng, T. et al. (2025). Digital health integration in chronic kidney disease. Clinica Chimica Acta. DOI: 10.1016/j.cca.2025.120749
๋ฉด์ฑ
์กฐํญ: ์ด ๊ฒ์๋ฌผ์ ์ ๋ณด ์ ๊ณต์ ์ํ ์ฐ๊ตฌ ๋ํฅ ๊ฐ์์ด๋ค. ์ธ์ฉ๋ ๊ตฌ์ฒด์ ์ธ ์ฐ๊ตฌ ๊ฒฐ๊ณผ, ํต๊ณ ๋ฐ ์ฃผ์ฅ์ ํ์ ์ฐ๊ตฌ์์ ์ธ์ฉํ๊ธฐ ์ ์ ์๋ณธ ๋
ผ๋ฌธ์ ํตํด ๊ฒ์ฆํด์ผ ํ๋ค.
์จ์ด๋ฌ๋ธ ์ผ์์ ์๊ฒฉ ํ์ ๋ชจ๋ํฐ๋ง: ๋์งํธ ํฌ์ค์ ๋ง์ฑ ์งํ์ ๋ง๋จ
๋ถ์ผ: ์ํ | ๋ฐฉ๋ฒ๋ก : ์์-๊ธฐ์ ์
์ ์: Sean K.S. Shin | ๋ ์ง: 2026-03-17
์ฐ๊ตฌ ์ง๋ฌธ
๋ง์ฑ ์งํ(๋น๋จ๋ณ, ์ฌ๋ถ์ , COPD, CKD)์ ์ ์ธ๊ณ ์ฌ๋ง์ 74%๋ฅผ ์ฐจ์งํ๋ฉฐ, ๋ฏธ๊ตญ ์๋ฃ๋น ์ง์ถ์ ์ฝ 86%๋ฅผ ์๋นํ๋ค(๊ทธ๋ฆฌ๊ณ ๋ ๋๊ฒ๋ ๊ณ ์๋ ๊ตญ๊ฐ ์๋ฃ๋น ์ง์ถ์ ์๋นํ ๋น์ค์ ์ฐจ์งํ๋ค). ์ ํต์ ์ธ ๊ด๋ฆฌ ๋ฐฉ์์ ์ฃผ๊ธฐ์ ์ธ ์ธ๋ ๋ฐฉ๋ฌธ์ ์์กดํ๋ฉฐ, ์ด๋ ๋ช ์ฃผ ํน์ ๋ช ๋ฌ ๊ฐ๊ฒฉ์ผ๋ก ๊ฑด๊ฐ ์ํ๋ฅผ ์ค๋
์ท์ฒ๋ผ ํฌ์ฐฉํ๋ ๋ฐฉ์์ด๋ค. ์จ์ด๋ฌ๋ธ ์ผ์์ ์๊ฒฉ ํ์ ๋ชจ๋ํฐ๋ง(RPM)์ ์ฆ์์ด ๋ํ๋๊ธฐ ์ ์ ์
ํ๋ฅผ ๊ฐ์งํ๊ณ ์ ์ ์ ๊ฐ์
์ ๊ฐ๋ฅํ๊ฒ ํ๋ ์ฐ์์ ยท์ค์๊ฐ ์๋ฆฌ ๋ฐ์ดํฐ ์คํธ๋ฆผ์ ์ ๊ณตํ ๊ฒ์ ์ฝ์ํ๋ค. ๊ทธ๋ฌ๋ ์ฐ์ ๋ชจ๋ํฐ๋ง์ด ์ค์ ๋ก ์์์ ๊ฒฐ๊ณผ๋ฅผ ๊ฐ์ ํ๋๊ฐ, ์๋๋ฉด ๋จ์ํ ์์์๋ฅผ ์๋ํ๊ณ ๊ฑฐ์ง ๊ฒฝ๋ณด๋ฅผ ์ ๋ฐํ๋ ๋ฐ์ดํฐ๋ง์ ์์ฑํ๋๊ฐ?
์ฐ๊ตฌ ํํฉ
Ghadi et al. (2025)์ ์๊ฒฉ ํ์ ์ผ์ด๋ฅผ ์ํ ์จ์ด๋ฌ๋ธ ๊ธฐ์ ๊ณผ AI์ ํตํฉ์ ๊ฒํ ํ์๋ค. ๊ทธ๋ค์ ๋ถ์์ ๋์
์ ์ฃผ์ ์ฅ๋ฒฝโ๋ฐ์ดํฐ ํ๋ผ์ด๋ฒ์ ์ ์ฝ, EHR ํตํฉ์ ์ด๋ ค์, ๋์งํธ ๋ฆฌํฐ๋ฌ์ ๊ฒฉ์ฐจ, ๋ฐ์ดํฐ ๊ณผ๋ถํโ์ ํ์ธํ๊ณ , ์ฐํฉ ํ์ต(federated learning, ๊ธฐ๊ด ๊ฐ ํ๋ผ์ด๋ฒ์ ๋ณด์กด ๋ชจ๋ธ ํ๋ จ์ ์ํ)๊ณผ ๋ฅ๋ฌ๋ ํ๋ ์์ํฌ(์์ ์ผ์ ์คํธ๋ฆผ์์ ์์์ ์ผ๋ก ์คํ ๊ฐ๋ฅํ ์ ํธ๋ฅผ ๊ฑธ๋ฌ๋ด๋ ์ง๋ฅํ ๊ฒฝ๋ณด๋ฅผ ์ํ)๋ฅผ ํฌํจํ AI ๊ธฐ๋ฐ ์๋ฃจ์
์ ์ ์ํ๋ค. ํต์ฌ ์ฃผ์ฅ์ ๊ฒฝ๋ณด ํผ๋ก(alert fatigue)๋ฅผ ํด๊ฒฐํ๊ณ RPM์ ์์์ ์ผ๋ก ์ง์ ๊ฐ๋ฅํ๊ฒ ๋ง๋ค๊ธฐ ์ํด "๋ฐ์ดํฐ ์ ์ก"์์ "์ง๋ฅํ ๊ฒฝ๋ณด"๋ก์ ์ ํ์ด ํ์ํ๋ค๋ ๊ฒ์ด๋ค.
Deng et al. (2025)์ ๋ง์ฑ ์ ์ฅ ์งํ(CKD)์ ํนํ๋ ๋์งํธ ํฌ์ค๋ฅผ ๊ฒํ ํ์๋ค. CKD๋ eGFR ๊ฐ์๊ฐ ์ ์ง์ ์ด๋ฉฐ ์งํ๋ ๋จ๊ณ๊น์ง ์ข
์ข
๋ฌด์ฆ์์ผ๋ก ๋ํ๋๋ค. ๊ทธ๋ค์ ์จ์ด๋ฌ๋ธ ๋ฐ์ด์ค์ผ์(์ฐ์ ํ์ ๋ชจ๋ํฐ, ์๋ถ ์ผ์)๊ฐ ๋ชจ๋ฐ์ผ ํฌ์ค ์ฑ ๋ฐ AI ์์ฌ๊ฒฐ์ ์ง์๊ณผ ๊ฒฐํฉ๋ ๊ฒฝ์ฐ, ๋ถ๊ธฐ๋ณ ๊ฒ์ฌ ๋ฐฉ๋ฌธ๋ณด๋ค ๋ ์ผ์ฐ CKD ์งํ์ ๊ฐ์งํ๊ณ ์ ์์ ์ธ ์ฝ๋ฌผ ์กฐ์ ์ ํตํด ๊ธฐ๋ฅ ์ ํ๋ฅผ ์ ์ฌ์ ์ผ๋ก ๋ฆ์ถ ์ ์์์ ๋ฌธ์ํํ์๋ค.
Stradford et al. (2024)์ WEAR ์ฐ๊ตฌ ํ๋กํ ์ฝ์ ๊ธฐ์ ํ์๋ค. ์ด๋ upadacitinib ๋๋ adalimumab์ ์์ํ๋ ๋ฅ๋งํฐ์ค ๊ด์ ์ผ ํ์๋ฅผ ๋์์ผ๋ก ์จ์ด๋ฌ๋ธ ํ๋ ์ถ์ ๊ธฐ์ ์ ์ ํ์ ๋ณด๊ณ ๊ฒฐ๊ณผ(ePRO)๋ฅผ ๊ฒฐํฉํ ์ ํฅ์ ๊ด์ฐฐ ์ฐ๊ตฌ์ด๋ค. ์ด ์ฐ๊ตฌ๋ ์จ์ด๋ฌ๋ธ์ ๊ฐ๊ด์ ํ๋ ๋ฐ์ดํฐ๊ฐ ์๊ฐ ๋ณด๊ณ ์ง๋ณ ํ์ฑ๋ ์ ์๋ฅผ ๋์ด์๋ ์ ๋ณด๋ฅผ ์ถ๊ฐํ๋์ง ๊ฒ์ฆํ๋๋ก ์ค๊ณ๋์์ผ๋ฉฐ, ๋ฐ์ดํฐ ์์ง์ด ์๋ฃ๋จ์ ๋ฐ๋ผ ๊ฒฐ๊ณผ๊ฐ ๊ณง ๋ฐํ๋ ์์ ์ด๋ค.
Daich Varela et al. (2024)์ RPM์ ์๊ณผ ๋ถ์ผ๋ก ํ์ฅํ์ฌ, ๋ง๋ง ์งํ์ ์ํ ๊ฐ์ ๊ธฐ๋ฐ ๋ชจ๋ํฐ๋ง ๊ธฐ๊ธฐ๋ฅผ ๊ฒํ ํ์๋ค. ๋น๋จ๋ง๋ง๋ณ์ฆ ๋ฐ ๋
ธ์ธ์ฑ ํฉ๋ฐ๋ณ์ฑ๊ณผ ๊ฐ์ด ๋น๊ฐ์ญ์ ์์์ด ๋ฐ์ํ๊ธฐ ์ ์ ์๊ฐ ๊ธฐ๋ฅ ๋ณํ๋ฅผ ๊ฐ์งํ ์ ์๋ ์งํ์์, ๊ฐ์ ๋ชจ๋ํฐ๋ง์ ๋ ์ด๋ฅธ ์น๋ฃ ์์์ ๊ฐ๋ฅํ๊ฒ ํ๋ค.
์ฃผ์ ์ฃผ์ฅ ๋ฐ ๊ทผ๊ฑฐ
<
| ์ฃผ์ฅ | ๊ทผ๊ฑฐ | ํ์ |
|---|
| AI ๊ธฐ๋ฐ ์จ์ด๋ฌ๋ธ์ ์์ ๋ฐ์ดํฐ ์ ์ก ๋๋น ๊ฒฝ๋ณด ํผ๋ก๋ฅผ ๊ฐ์์ํจ๋ค | Edge AI ์ฒ๋ฆฌ๊ฐ ์์์ ์ผ๋ก ์คํ ๊ฐ๋ฅํ ๊ฒฝ๋ณด๋ง์ ์์ฑํ๋ค (Ghadi et al. 2025) | ์ค๊ณ ์ธก๋ฉด์์ ์ง์ง๋จ; ์์์ ๊ฒฐ๊ณผ ๊ทผ๊ฑฐ๋ ์์ง ์ฑ์ ์ค |
| ์จ์ด๋ฌ๋ธ ๋ชจ๋ํฐ๋ง์ ์ธ๋ ๋ฐฉ๋ฌธ๋ณด๋ค CKD ์งํ์ ๋ ์ผ์ฐ ๊ฐ์งํ๋ค | ์ฐ์ ํ์ ๋ฐ ์๋ถ ๋ฐ์ดํฐ๊ฐ ๋ฌด์ฆ์ ๋ณํ๋ฅผ ํฌ์ฐฉํ๋ค (Deng et al. 2025) | ์ ๋งํจ; ํ์ธ์ ์ํด ๋๊ท๋ชจ RCT ํ์ |
| ์จ์ด๋ฌ๋ธ ํ๋ ๋ฐ์ดํฐ๋ ํ์ ๋ณด๊ณ ๊ฒฐ๊ณผ๋ฅผ ๋ณด์ํ๋ค | WEAR ์ฐ๊ตฌ๊ฐ ์ด ๊ฐ์ค์ ๊ฒ์ฆํ๋๋ก ์ค๊ณ๋จ (Stradford et al. 2024); ๊ฒฐ๊ณผ ๋๊ธฐ ์ค | โ ๏ธ ๊ฐ์ค โ ์ฐ๊ตฌ ํ๋กํ ์ฝ ๋
ผ๋ฌธ; ๊ฒฐ๊ณผ ๋ฏธ๋ฐํ |
| ๊ฐ์ ๋ชจ๋ํฐ๋ง์ ๋ ์ด๋ฅธ ๋ง๋ง ์งํ ์น๋ฃ๋ฅผ ๊ฐ๋ฅํ๊ฒ ํ๋ค | ์๊ฐ ๋ชจ๋ํฐ๋ง ๊ธฐ๊ธฐ๊ฐ ์์ฝ ๋ฐฉ๋ฌธ ์ ์ ์๊ฐ ๋ณํ๋ฅผ ๊ฐ์งํ๋ค (Daich Varela et al. 2024) | ์ง์ง๋จ; ๊ฐ์ ๋ชจ๋ํฐ๋ง์ ๋ํ ํ์ ์์๋๋ ๋ค์ํจ |
๋ฏธํด๊ฒฐ ์ง๋ฌธ
์์ ๊ฒฐ๊ณผ ๊ทผ๊ฑฐ: ๋๋ถ๋ถ์ RPM ์ฐ๊ตฌ๋ ๊ณผ์ ๊ฒฐ๊ณผ(์
์์จ, ๊ฐ์ง ์์ ์๊ฐ)๋ฅผ ์ธก์ ํ๋ค. ์ด๋ฌํ ์ฐ๊ตฌ๊ฐ ์ค์ง์ ์ธ ๊ฒฐ๊ณผ(์ฌ๋ง๋ฅ , ์ฅ์ ์๋ ์์กด)๋ฅผ ๊ฐ์ ํ๋๊ฐ?
์๋ฃ ํํ์ฑ: ์จ์ด๋ฌ๋ธ RPM์ ์ค๋งํธํฐ, ์ธํฐ๋ท ์ ์, ๋์งํธ ๋ฆฌํฐ๋ฌ์๋ฅผ ํ์๋ก ํ๋ค. ์๋ฃ ์์คํ
์ ๊ณตํํ ์ ๊ทผ์ ์ด๋ป๊ฒ ๋ณด์ฅํด์ผ ํ๋๊ฐ?
๋ฐ์ดํฐ ๊ฑฐ๋ฒ๋์ค: ์ง์์ ์ธ ๊ฑด๊ฐ ๋ชจ๋ํฐ๋ง์ ๋ฐฉ๋ํ ๊ฐ์ธ ๋ฐ์ดํฐ์
์ ์์ฑํ๋ค. ์ด ๋ฐ์ดํฐ์ ์์ ๊ถ์ ๋๊ตฌ์๊ฒ ์์ผ๋ฉฐ, ์์
์ ์ฐฉ์ทจ๋ก๋ถํฐ ์ด๋ป๊ฒ ๋ณดํธํด์ผ ํ๋๊ฐ?
์ํ ๋ชจ๋ธ: ์๋ฃ ์ง๋ถ์๋ค์ RPM ์ํ ์ฝ๋๋ฅผ ์์ํ ๋์
ํ๊ณ ์์ผ๋, ๋ณด์ฅ ๋ฒ์๋ ๋งค์ฐ ๋ค์ํ๋ค. ํ์คํ๋ ๋น์ฉ ํจ๊ณผ์ฑ ๊ทผ๊ฑฐ๊ฐ ๋์
์ ๊ฐ์ํํ ์ ์๋๊ฐ?References (4)
Ghadi, Y. Y., Shah, S. F. A., Waheed, W., Mazhar, T., Ahmad, W., Saeed, M. M., et al. (2025). Integration of wearable technology and artificial intelligence in digital health for remote patient care. Journal of Cloud Computing, 14(1).
Daich Varela, M., Sanders Villa, A., Pontikos, N., Crossland, M. D., & Michaelides, M. (2024). Digital health and wearable devices for retinal disease monitoring. Graefe's Archive for Clinical and Experimental Ophthalmology.
Stradford, L., Curtis, J. R., Zueger, P., Xie, F., Curtis, D., Gavigan, K., et al. (2024). Wearable activity tracker study exploring rheumatoid arthritis patientsโ disease activity using patient-reported outcome measures, clinical measures, and biometric sensor data (the wear study). Contemporary Clinical Trials Communications, 38, 101272.
Deng, T., Xue, Y., & Methakanjanasak, N. (2026). Digital health integration in chronic kidney disease. Clinica Chimica Acta, 582, 120749.