Feasibility of wearable technology for remote monitoring in high-risk adults with sickle cell disease: Baseline data from the SCD-carre trial Free

Background: Adults with sickle cell disease (SCD) experience reduced physical capacity due to anemia and cardiopulmonary complications. Compromised physical functioning and low engagement in daily activities among individuals with SCD can indicate and cause decline in mental and physical health. Monitoring exertion and functioning in real time can give us insight into risk factors for disease outcomes and improve care in this population, enabling better prediction and prevention of disease progression. Limited research has explored wearable technology to assess disease-related biobehavioral symptoms in this population. Our group has shown that wearable data can be used to collect these data and predict outcomes in the hospital setting. Here we sought to evaluate whether remote monitoring of activity, sleep, and physiology is feasible in ambulatory real-world settings, with a high-risk SCD population. We hypothesized that we could remotely collect valid wearable data from ≥80% of SCD patients enrolled in a multisite trial.

Methods: We conducted an ancillary study within the SCD-CARRE trial, a 12-month multi-site randomized controlled trial, testing whether monthly RBC exchange transfusion added to usual care reduces SCD-related clinical episodes in high-risk adults. High-risk criteria included: TRV ≥3.0 m/sec, TRV 2.5-2.9 m/sec plus NT-proBNP ≥160 pg/mL, or chronic kidney disease with macroalbuminuria/proteinuria or eGFR <60 mL/min/1.73m². Participants wore Garmin devices continuously for 7-day periods at baseline and follow-up visits at 4, 8, and 12 months. Valid data was defined as ≥3 days with ≥10 hours of device wear per day. Feasibility was operationalized as >80% valid data collection among those with available data. Secondary outcomes included descriptive analysis of activity metrics (step counts, sedentary and active time), sleep parameters (deep, light, REM, awake duration), and heart rate variables (average, minimum, maximum HR), averaged across all valid days available.

Results: Of the 173 participants enrolled in SCD-CARRE, 155 (89.6%) had Garmin watch data available, with 18 participants having no data due to device errors, device not placed, or data still pending at time of analysis. Among those with available data, 10 participants (6.5%) had zero valid wear days, while 145 (93.5%) achieved at least one valid wear day. A total of 132 participants (76.3% of enrolled, 85.2% of those with available data) met our primary feasibility criteria.

Participants without valid data were more likely to be on intensive disease-modifying therapies, including hydroxyurea (87.8% vs 64.4%, p<0.01) and other agents such as glutamine or P-selectin inhibitors (31.7% vs 15.2%, p=0.03). There were no other differences between groups.

Participants with valid wearable data demonstrated excellent device compliance with a mean wear time of 21.5 hours per day (SD 3.7), indicating nearly continuous monitoring. Activity metrics showed mean daily steps of 4,226 (SD 3,337), sedentary time of 18.0 hours (SD 3.9), and active time of 3.5 hours (SD 2.0). Heart rate parameters were within expected ranges with mean values of 80.2 bpm (average), 60.0 bpm (minimum), and 121.9 bpm (maximum). Sleep data were available for 106 participants, showing mean total sleep duration of 6.3 hours (SD 1.9), with light sleep averaging 4.0 hours (SD 1.4) and deep sleep 2.4 hours (SD 1.4).

Conclusions: This is the largest study to date to successfully demonstrated feasible wearable data collection adults with SCD in real-world ambulatory settings. Of 155 participants with any data, 85.2% (132) had ³3 days of valid data, comparable to published compliance rates of 80-90% in similar studies. However, 18 participants generated no data, suggesting device distribution, wear, or technical issues. Patients on disease-modifying therapies were more likely to have missing or insufficient data, suggesting greater compliance challenges in those requiring intensive medical management. Mean activity, heart rate, and sleep metrics were within expected ranges and varied significantly across participants.

Study limitations included discontinued oxygen saturation monitoring due to unreliable data quality, sleep compliance issues, and device malfunctions. Future studies will examine correlations with clinical outcomes and treatment differences. These findings support feasibility of remote monitoring in SCD populations for both research and clinical care.