# Validity and Reliability of a Smart Band for Monitoring Cardiorespiratory Parameters in Children and Adolescents with Severe Cerebral Palsy

## TL;DR

This study evaluates the Xiaomi Mi Band 6 for monitoring heart rate and oxygen saturation in children with severe cerebral palsy, finding it reliable for heart rate but not for oxygen saturation.

## Contribution

The study provides the first evaluation of the Xiaomi Mi Band 6's reliability and validity for cardiorespiratory monitoring in children with severe cerebral palsy.

## Key findings

- Mi Band 6 showed good reliability for heart rate (ICC = 0.83) but high measurement error.
- SpO2 measurements had low reliability (ICC = 0.55) and exceeded acceptable clinical error margins.
- Factors like involuntary movements and low body weight affected sensor accuracy.

## Abstract

Cerebral palsy (CP) is a disorder frequently associated with respiratory and cardiac comorbidities, making the monitoring of heart rate (HR) and oxygen saturation (SpO2) essential. This study examined the reliability and validity of Xiaomi Mi Band 6, compared with a clinical pulse oximeter, for measuring HR and SpO2 in 35 children and adolescents with CP classified at GMFCS levels III–V. Mi Band 6 demonstrated good reliability for HR (ICC = 0.83), although the high measurement error (MDC90 = 19.57 bpm) limits its usefulness for small physiological changes. SpO2 results showed low reliability (ICC = 0.55) and substantial variability (MDC90 = 18.85%), exceeding the clinically acceptable error margin of ±2–3%. Validity analyses revealed poor agreement between Mi Band 6 and clinical pulse oximeter for SpO2, and moderate agreement for HR, with large variability in Bland–Altman analyses. Factors such involuntary movements, altered muscle tone, low body weight, and reflective sensors on the wrist may have affected the results. In conclusion, Xiaomi Mi Band 6 demonstrated good reliability and may be cautiously used for general HR monitoring, but it is not suitable for assessing SpO2 in this pediatric population. Further research is needed to identify cost-effective and accurate wearable technologies.