Assessing mobility, upper extremity function, and pain interference in pediatric physical therapy: Psychometric properties of the PROMIS® item banks

<h2>Abstract</h2><h3>Objectives</h3> To examine psychometric properties of three Dutch-Flemish pediatric and proxy PROMIS® item banks in pediatric physical therapy (PPT). <h3>Methods</h3> Children (8-18 years) attending PPT in the past year, and their parents, completed Dutch-Flemish PROMIS® Mobility, Upper Extremity, Pain Interference item banks (v2.0). Structural validity, internal consistency, reliability, test-retest reliability, known-group validity, comparability and child-parent agreement were assessed. <h3>Results</h3> 244 children and 361 parents participated. A third of the children (37.1%) had a chronic medical diagnosis. The physiotherapeutic goals of the children in our sample were related to musculoskeletal (57.3%), neurological (28.8%), gastrointestinal, endocrine, and/or hematological (0.8%), cardiovascular/respiratory (7.8%), mental health (2.5%), and behavioral (13.3%) conditions. Structural validity, internal consistency, test-retest reliability, comparability and child-parent agreement were sufficient. Children in PPT scored lower on all three PROMIS instruments compared to the general pediatric population (p<0.05). PROMIS Mobility and Upper Extremity scores were reliable for clinical scores. PROMIS Pain Interference scores were reliable for population average and clinical scores. Computerized adaptive testing provided most information per item. <h3>Conclusion</h3> PROMIS v2.0 pediatric/proxy Mobility, Upper Extremity, and Pain Interference instruments displayed sufficient validity in PPT and sufficient reliability for clinical scores. They are appropriate for use in PPT. <h3>Impact statement</h3> This study provides the first comprehensive validation of the scores on the PROMIS v2.0 pediatric Mobility, Upper Extremity, and Pain Interference item banks for use in pediatric physical therapy, demonstrating that these instruments are suitable for clinical practice. By showing that computerized adaptive testing offers most information per item and that parent-proxy scores are reliable when self-report is not possible, this research advances the science and practical application of PROMs in heterogeneous pediatric populations and highlights key areas for future improvement.