Upper-limb accelerometry in children with cerebral palsy: a scoping review

Purpose Among children with cerebral palsy (CP), gains on standardized upper-limb (UL) capacity measures do not necessarily translate into greater spontaneous arm use in everyday environments. This scoping review mapped how UL accelerometry – predominantly bilateral wrist-worn accelerometry – is used in children and adolescents with CP to examine (i) construct validity against standardized UL measures and (ii) intervention monitoring in treatment settings and everyday environments.Methods Following Joanna Briggs Institute methodology and PRISMA-ScR guidance, PubMed, Embase, and CINAHL were searched for English-language studies published between January 2000 and May 16, 2026. Peer-reviewed studies were included if they involved children and adolescents (0–18 years) with CP that used UL-worn accelerometers to report quantitative metrics in standardized or free-living settings.Results Nineteen studies met the inclusion criteria. In construct-validity studies, asymmetry-focused metrics showed the most consistent associations with standardized bimanual measures and manual ability classifications, with the strongest relationships observed in standardized sessions and weaker, context-dependent associations in free-living monitoring. In intervention studies, accelerometry was consistently sensitive to within-session dose and immediate behavioral change during intervention sessions. However, translation to spontaneous everyday arm use was heterogeneous, with several studies showing improved clinical scores without parallel gains in free-living accelerometry metrics.Conclusion UL accelerometry provides objective performance information to complement assessment and neurorehabilitation monitoring in children with CP, especially when asymmetry-oriented metrics are matched to the clinical question and recording context. Its clearest current role is as a complement to standardized assessment and as a tool to quantify in-session engagement, whereas evidence for sustained transfer to everyday arm use remains inconsistent. Future studies should consistently report device model and placement, wear protocol, valid wear time, preprocessing methods, and metric definitions.