Hand-Use-at-Home Questionnaire: validity and reliability in children with neonatal brachial plexus palsy or unilateral cerebral palsy

Contains fulltext : 196905.pdf (Publisher’s version ) (Open Access)
OBJECTIVE: To investigate construct validity and test-retest reliability of the parent-rated Hand-Use-at-Home questionnaire (HUH) in children with neonatal brachial plexus palsy or unilateral cerebral palsy. DESIGN AND SUBJECTS: For this cross-sectional study, children with neonatal brachial plexus palsy or unilateral cerebral palsy, aged 3-10 years, were eligible. MAIN MEASURES: The HUH, Pediatric Outcome Data Collection Instrument Upper Extremity Scale (neonatal brachial plexus palsy only), and Children's Hand-Use Experience Questionnaire (unilateral cerebral palsy only) were completed. The HUH was completed twice in subgroups of both diagnoses. Lesion-extent (indication of involved nerve rootlets in neonatal brachial plexus palsy as confirmed during clinical observation and/or nerve surgery) and Manual Ability Classification System levels (unilateral cerebral palsy) were obtained from the medical records. Spearman correlation coefficients between the HUH and all clinical variables, agreement, standard error of measurement, smallest detectable change and intra-class correlation were calculated. RESULTS: A total of 260 patients participated (neonatal brachial plexus palsy: 181), of which 56 completed the second HUH (neonatal brachial plexus palsy: 16). Median age was 6.9 years for children with neonatal brachial plexus palsy, 116 had C5-C6 lesions. Median age for children with unilateral cerebral palsy was 6.4 years, 33 had Manual Ability Classification System Level II. The HUH correlated moderately with lesion-extent ( rs =-0.5), Pediatric Outcome Data Collection Instrument Upper Extremity Scale ( rs = 0.6) and Children's Hand-Use Experience Questionnaire ( rs = 0.5) but weakly with Manual Ability Classification System levels ( rs = -0.4). Test-retest reliability was excellent (intra-class correlation2,1 = 0.89, standard error of measurement = 0.599 and smallest detectable change = 1.66 logits) and agreement was good (mean difference HUH1 - HUH2 = 0.06 logits). CO