The KIDS-KNEES-ACL is a valid PROM for children and adolescents with an ACL injury; item reduction using Rasch analysis and confirmatory factor analysis

PURPOSE: The purpose of this study was to analyse the structural validity of the preliminary paediatric Knee Numeric-Entity Evaluation Score-Anterior Cruciate Ligament (KIDS-KNEES-ACL) Version 1.0 by Rasch analysis and Confirmatory Factor Analysis (CFA), to establish a final item-reduced and structurally valid KIDS-KNEES-ACL Version 2.0.

METHODS: KIDS-KNEES-ACL Version 1.0 is a content-valid patient-reported outcome measure (PROM) for the evaluation of children and adolescents with anterior cruciate ligament (ACL) deficiency. It consists of 60 items across 9 subscales. KIDS-KNEES-ACL Version 1.0 was completed by patients from a national cohort of children treated with ACL reconstruction. For each subscale, structural validity was evaluated by fit to the Rasch model for 141 respondents. Item fit, local dependence (LD), and Differential Item Functioning (DIF) for age and sex were assessed. These properties guided the process of item reduction until an adequate fit to the Rasch model was reached. The structure, as suggested by the Rasch model, was confirmed with CFA. Internal consistency was evaluated by Cronbach's α.

RESULTS: Rasch analysis revealed item misfit of four of 60 items, and LD for several items across all subscales. When adjusted for, omitting 16 items from the item pool, excellent item fit statistics were achieved. Four subscales could be merged into the two new 'Friends and Family' and 'Looseness and Weakness' subscales. CFA confirmed the structure of all subscales with excellent χ 2 fit (p > 0.05) without further modification. Internal consistency was acceptable with α values ranging from 0.71 to 0.88. There was no evidence of DIF with respect to age or sex.

CONCLUSION: The KIDS-KNEES-ACL 2.0, consisting of 44 items across seven subscales, is a new content-valid and structurally valid PROM for children and adolescents with ACL deficiency. It can serve as an evaluative outcome for ACL injury and reconstruction.

LEVEL OF EVIDENCE: N/A.