The increased use of digital health technology (DHT) in clinical research is helping to expand our understanding of how people function in their real-world environment. While DHTs bring many potential benefits to clinical research, there are still technical challenges to overcome for DHTs to be used in research across populations, including pediatric participants. A recent report by Rock Health highlights data suggesting that investments in healthcare innovations that benefit children are lagging. One effort to advance digital health benefits for children is led by the Digital Medicine Society (DiMe) and Boston Children’s Hosptial in an effort to create a version of The Playbook for Pediatric Digital Medicine as a guide to help the community address some of these gaps in healthcare.  

Another effort to address these gaps and help pediatric researchers realize the benefits of innovations in DHT was described in a recent publication from the team at Pfizer Innovation Research Lab, Monitoring Activity and Gait in Children (MAGIC) using digital health technologies.  The goal of the study was to -

• validate digital measures for physical activity (PA) and gait in children,
• develop a normative data set for reference, and
• assess the comfort and compliance of wearable DHTs in this population.  

In the May 2024 session of our Digital Health Monthly scientific webinar series, I was joined by a member of the Pfizer team who worked on the MAGIC study, Junrui Di, PhD, Associate Director, Digital Medicine Statistics, to review the results of the study and shed light on the intricacies of this project. 

Gait is an important indicator of our mobility and can provide important information about our health, so much so that gait has been referred to as the 6th vital sign. The importance of gait and physical activity measures in health outcomes has recently been underscored by the qualification of stride velocity 95th centile as a wearable-derived clinical endpoint by the European Medicines Agency (EMA), and moderate-to-vigorous physical activity (MVPA) being recognized by the FDA as a primary endpoint in a phase 3 clinical trial. Therefore, establishing validity and usability of DHTs to collect similar measures is important so that these innovations can also provide meaningful clinical insights and advance research in pediatric health. 

Details and Results from the MAGIC Study 

Gait and physical activity have been studied extensively in children, including children with conditions such as Duchenne Muscular Dystrophy (DMD) and in the context of obesity control, but there remained a lack of standardized measures and further validation of DHTs in developing children. The MAGIC study involved 40 participants aged 3-17 years, divided into three age groups (3-5 years, 6-11 years, and 12-17 years). The study design followed the protocol: 

• Day 1: Collection of data from in-lab walking tasks and a wearability/comfort questionnaire. 

• Day 1-15: At-home monitoring, where participants wore an ActiGraph CentrePoint Insight Watch (CPIW) device on their wrist and lumbar region, maintained a daily activity diary, and had a phone interview on day 15. 

• Post-study: Assessment for adverse events, followed by mailing back the device. 

Dr. Di shared an inside look at the recruitment times between age groups to help inform researchers whose studies include these ages of approximate timelines that they could expect. Recruitment times varied across age groups, with the youngest group (3-5 years) taking the longest (232 days), and the oldest group (12-17 years) the shortest (77 days).  

Compliance is a critical component to the success of clinical trials utilizing DHTs. Therefore, understanding wear compliance, comfort and wearability of DHTs in this pediatric population were one of the main goals of the MAGIC study. Overall, compliance for the lumbar worn device was lower than wrist, especially in the 3-5 years age group at about 40% compliant days, but lumbar wear did increase in the participants >6 years to 80% compliant days. Wrist-worn device compliance was higher than 80% for all age groups. Comfort assessments revealed high acceptability, with 85% of participants agreeing or strongly agreeing that wrist and lumbar DHTs were comfortable to wear.  

In-lab gait validation showed a significant linear correlation with all GAITRite metrics across different walking speeds. Agreement levels ranged from good to excellent at slow and natural speeds, while moderate agreement was observed for fast walking speeds. Additionally, age-related trends indicated a decrease in PA and an increase in sedentary time with age, while most gait characteristics (except cadence) increased with participant age. 

The details of all of the MAGIC study data that was presented can be viewed in the webinar presentation slides. 

The MAGIC study demonstrated that DHTs can reliably and accurately capture PA and gait information in children, and did confirm age effects in these metrics. Given the relatively high wear compliance and positive feedback on the comfort and willingness to wear DHTs for extended periods of time (>7 days), it’s likely that DHTs can help reduce the burden of study participation by collecting data remotely and reducing the number of site visits necessary throughout a study.   

Q&A Session Highlights 

The study's findings prompted a flood of audience questions and created some interesting discussion points, including: 

What cut points would you recommend for the 13 to 17-year age group? 

• The study utilized a Crouter algorithm for adolescents, but Dr. Di mentioned that there could be utility in moving away from threshold-based measures like MVPA when addressing certain research questions. 

How did you manipulate gait speed in young children? 

• The research team applied creative strategies and incorporated games like “the floor is lava” or red light/green light. They also used different language to communicate what they meant by normal and fast walking speeds, such as “a normal walk is the way you usually walk, and fast walk is where you try a little bit harder.” Dr. Di admitted that this type of experiment is not that easy, because kids might jump or run which makes it difficult to get good quality gait data.  

Was a higher variation seen in the younger age groups? 

• Yes. The team did a separate analysis by age group and the youngest cohorts are lower, although not by much.  

(And, as the moderator, I got to sneak in my own question for Dr. Di) In physical rehab research, gait or stride length can change with age. How do you differentiate between a treatment effect versus normal growth changes?  

• Dr. Di saw this as a study design and/or statistical analysis question. Modeling child growth is a popular topic in statistical analysis, and/or researchers can consider doing a crossover design to mitigate age effects. 

There were also several questions related to the device use and the type of data that could be collected, such as:  

Can the range of motion of joints be measured?  

• No, this isn’t likely with the device mounted in the lumbar region. It’s likely that multiple devices mounted on the limb would be necessary to collect this type of data. 

The CentrePoint Insight Watch (CPIW) was used in this study, but it’s not specifically stated that it’s indented for lumbar wear, so why was it used for that placement in this study? How can you validate the wear location in the real-world? 

• The CPIW, although a “watch”, can be put on a lumbar location with appropriate accessories. Essentially, it is a wearable device with tri-axial accelerometer. You can then use the appropriate algorithms (e.g. Scikit-Digital-Health gait package) to derive gait and activity measures.  

• Functional and comfortable accessories are critical to be able to mount the device properly on the lumbar region. 

• In terms of validation, one recommendation is to validate the measures of interest in-clinic, then collect data for a few days in the real-world/at-home environment. Validation of gait measures such as gait speed in the real-world is difficult due to the challenges in recording ground truths for speed, stride length etc. at home or free-living.   

Did devices have to be removed during sports/aquatics? 

• They did not have to be removed during physical activities like sports, but participants were asked to remove the device during aquatic activities (swimming, bathing, etc.). 

Click here to watch the on-demand recording of this installment of our Digital Health Monthly scientific webinar series, Monitoring Activity and Gait in Children using DHTs.