Psychometrics of Wearable Devices Measuring Physical Activity in Ambulant Children With Gait Abnormalities: A Systematic Review and Meta-analysis.
Huib van Moorsel, Barbara Engels, Jacek Buczny and 5 others
PMID 39822201WHAT IT FOUND
Wearable activity monitors are generally consistent when repeated and agree reasonably well with reference measures in ambulant children with gait problems.
Accuracy varies by setting, activity type and age. This is about measurement, not therapy effect.
Key findings
01Pooled test-retest reliability was high (ICC=0.81; 95% CI, 0.74-0.89), and pooled inter-device reliability was very high (ICC=0.99; 95% CI, 0.98-1.00).
02Pooled criterion validity was high (r=0.70; 95% CI, 0.59-0.82), while pooled construct validity was moderate (r=0.63; 95% CI, 0.36-0.89).
03Moderator analyses showed higher test-retest reliability in laboratory settings (ICC=0.91; 95% CI, 0.81-1.00) than in free-living settings (ICC=0.77; 95% CI, 0.69-0.84), and higher cutoff-based criterion validity in children 13 years or older (r=0.91; 95% CI, 0.79-1.00) than in children younger than 13 years (r=0.61; 95% CI, 0.54-0.68).
STILL TO COME
How it was doneWhat they foundWhat it means for PTs
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
Most validation studies were laboratory based, so the pooled results may not show how devices behave during ordinary daily activity. Measurement error could not be summarized across studies, so clinicians cannot tell from this review how much a child's score must change before it is real. Responsiveness was not assessed, so the review does not show whether wearables can detect clinically meaningful change over time. The evidence is dominated by children with cerebral palsy; 71% of studies focused on CP, and children who use wheelchairs or have non-neuromuscular gait causes were excluded. Construct validity was based on only 3 studies, so that pooled estimate is less stable than the reliability or criterion validity estimates. Methodological quality was mixed: many reliability, validity and measurement-error studies were rated doubtful or inadequate because of missing population details, epoch descriptions or statistical information. The review included children aged 2 to 19 years, but age-specific device accuracy was not fully resolved; cutoff-based validity was lower for children younger than 13 years.
Declared interests
The paper reports support from Foundation Innovation Alliance Regional Attention and Action for Knowledge Circulation (SIA Raak, MKB12.002).
The easy way to misread this
Do not conclude that any wearable activity monitor is equally accurate for every ambulant child in daily life. Most validation studies were laboratory based, free-living reliability was lower than laboratory reliability, measurement error could not be pooled, and one study found poor wrist-device step detection when children used crutches or walkers.