P4.A006: Evaluation of a Novel Remote Vital Sign Monitor Compared to Standard Inpatient Monitoring in Pediatric Patients with Cyanotic Congenital Heart Disease

Background: Children with cyanotic congenital heart disease (cCHD) are discharged with multiple remote patient monitoring (RPM) devices for caregivers to intermittently measure, manually record, and interpret vital signs then notify medical providers of concerns. This error-prone process misses clinically significant vital sign events between measurements. VITLS, an FDA 510k cleared pediatric vital sign monitoring platform, continuously measures pulse oximetry (SpO2), heart rate (HR), respiratory rate, and temperature via a wireless probe on the child's chest, instantly storing data in a secure cloud accessible by medical providers and caregivers. Transitioning to VITLS may improve data collection accuracy, but VITLS has not been evaluated in this patient population.

Methods: This IRB approved (H46463), ClinicalTrials.gov registered (NCT05526768) study had two objectives: primarily, evaluate the non-inferiority of VITLS’s measurement of SpO2 compared to standard inpatient monitoring (CTRL) in patients with cCHD (defined as mean difference in SpO2 < 5 in >50% of patients); secondarily, analyze frequency and accuracy of detection of clinically significant vital sign (VS) events by VITLS vs. CTRL (defined as any of the following sustained for 1 minute: SpO2 < 75%, SpO2 > 90%, HR < 100 beats per minute (BPM), or HR > 180 BPM). Predefined study completion was complete data collection on 25 cCHD patients aged 0-1 year.

Results: Twenty-seven patients enrolled, 1 withdrew prior to data collection due to altered goals of care, and 1 transferred units prior to complete data collection. In 25 patients, recording of SpO2 and HR every 2 seconds with VITLS and CTRL for 6 hours minimum per patient was completed. Three patients were excluded due to VITLS data loss, leaving 22 patients analyzed. For each patient, Bland-Altman plots (SpO2) and time series plots (SpO2 and HR) were evaluated, and VS events were analyzed with confusion matrices with chi square testing. Summary plots illustrate the overall difference in SpO2 measurement (Figure 1), and total minutes measured of VS events (Figure 2). We found strong agreement in SpO2 between VITLS and CTRL in 55% of patients (mean differences in SpO2 level < 5).

Conclusion: VITLS was not inferior to CTRL for SpO2 measurement. VITLS may be a reasonable RPM device for SpO2 monitoring in patients with cCHD, however when measuring clinically significant vital sign events, VITLS measured far more minutes of events of SpO2 < 75%, SpO2 >90% and HR < 100 BPM compared to CTRL, indicating that VITLS may have increased false alarms. Additionally, VITLS failed to measure 110.65 minutes of events with HR >180 BPM indicating potential to miss clinically significant HR changes. Overall, VITLS is a promising, but not yet ideal RPM device for this patient population.