Feasibility of smart ring-based remote monitoring in hospitalized patients: evaluation of signal transmission latency and data acquisition continuity.
Hyunsun Lim, Jung Mi Han, Min Hyun Park, Joo Youn Kim, Haeyong Pak, Seung Jin Oh
Abstract
Open AccessBackground: General wards rely on intermittent vital-sign checks, creating blind spots for deterioration detection under staffing constraints. While wearable sensors promise continuous ward-scale monitoring, real-world feasibility remains uncertain because prior studies emphasized signal accuracy rather than the timeliness and low-latency continuous monitoring system. This study addresses that gap by quantifying transmission latency and signal acquisition rate (SAR) of a ring-based system in routine surgical-ward care. Methods: We conducted a single-center retrospective analysis of 128 adult postoperative inpatients monitored with a finger-worn smart ring (CART-1 plus) from 23 December 2024 to 20 July 2025. Signals streamed via BLE to a dedicated receiver and hospital Wi-Fi. Primary endpoints were (i) transmission latency (device timestamp→server receipt) and (ii) SAR. Cached burst uploads at 0-1-s intervals were excluded from latency analyses. Prespecified acceptability thresholds were ≤10 min for BP latency, ≤5 min for SpO2 latency, and ≥80% for SAR. Associations with patient/environmental factors (mobility status, room type, time of day, network environment) were modeled. Results: Mean transmission latency was 7.23 min (BP) and 4.35 min (SpO2), both within predefined thresholds (10 and 5 min, respectively; p < 0.001). The mean SAR was 90.17% for BP and 90.62% for SpO2, significantly exceeding the 80% threshold (p < 0.001). Transmission latency was significantly lower in patients requiring no or full assistance with mobility, in those residing in single or double rooms (vs. quad), and during daytime monitoring (p < 0.001). For SpO2, latency was affected by mobility and change in network environment but not by room type or monitoring time. SAR was not significantly influenced by clinical variables but was higher after change in network environment (p < 0.05). Conclusion: In real-world surgical wards, a finger-worn ring achieved clinically acceptable transmission latency and high continuity, supporting feasibility for a low-latency continuous monitoring system. Findings identify modifiable environmental and patient-level factors that influence latency and can guide scale-up and integration with early-warning workflows.