Psychometric examination of the patient-reported outcome measurement information system parent proxy pediatric sleep measures from early childhood to adolescence in a nationally representative Spanish sample

• Published in: Sleep (New York, N.Y.) N.Y.), 2024-12, Vol.47 (12)
• Main Authors: Becker, Stephen P, Burns, G Leonard, Montaño, Juan José, Servera, Mateu
• Format: Article
• Language: English
• Subjects: Adolescent; Child; Child, Preschool; Factor Analysis, Statistical; Female; Humans; Male; Parents; Patient Reported Outcome Measures; Proxy; Psychometrics; Reproducibility of Results; Sleep Wake Disorders - diagnosis; Sleep Wake Disorders - epidemiology; Spain; Surveys and Questionnaires
• ISSN: 0161-8105; 1550-9109; 1550-9109
• DOI: 10.1093/sleep/zsae215

To use a nationally representative sample to (1) evaluate the factor structure of the patient-reported outcome measurement information system (PROMIS) parent proxy pediatric sleep scales (Spanish translation), (2) examine the invariance of these scales across sex and across different developmental periods of childhood and adolescence, (3) confirm the information and precision of the scales using item response theory (IRT), and (4) provide age-based normative information. Parents of a nationally representative sample of 5525 Spanish children and adolescents ages 5-16 years (56.1% boys) completed the Spanish translation parent proxy short versions of the sleep disturbance and sleep-related impairment scales. We conducted confirmatory factor analyses, invariance analyses, and graded-response IRT analyses. CFAs conducted separately on males and females within three age groups (early childhood: ages 5-8 years; middle childhood: ages 9-12 years; adolescence: ages 13-16 years) indicated all items had a substantial loading with one exception (the sleep continuity item ["my child slept through the night"] had a substantially lower loading and was removed for subsequent analyses). The scores on the two scales demonstrated invariance across sex within each age group. Using IRT analyses, both scales showed a high degree of information and precision from slightly below the trait means to slightly above two standard deviations above the trait means. The strong psychometric properties of the short versions of the parent proxy PROMIS pediatric sleep disturbance and sleep-related impairment scales, coupled with age-based norms, suggest these scales are likely to be useful for research and clinical applications.