COSMIN review finds no broad-band ADHD child scale meets all psychometric criteria

A systematic review of 36 studies examining 12 broad-band behavioral rating instruments found that no ADHD-related subscale met all COSMIN requirements for recommended use in children aged 18 years or younger. The Strengths and Difficulties Questionnaire (SDQ) Hyperactivity-Inattention subscale had the most favorable evidence among the properties assessed, while the Child Behavior Checklist (CBCL) Attention Problems subscale received a recommendation against use because high-quality evidence showed insufficient cross-cultural validity. The review focused on subscales explicitly measuring inattention, hyperactivity or impulsivity within broad-band instruments.

Such instruments assess a wider range of emotional, behavioral and social functioning, but their ADHD-related subscales cannot be assumed to perform as well as the overall scale. The authors therefore evaluated each relevant subscale separately, consistent with version 2.0 of the COSMIN guidelines. Following PRISMA 2020 guidance, researchers searched PubMed, Web of Science, Scopus and ProQuest from each database’s inception through December 1, 2025.

The search was supplemented by citation tracking and reference chaining. Two reviewers independently screened records, extracted data, evaluated risk of bias, rated measurement properties and synthesized evidence, resolving discrepancies by consensus. Methodological quality was assessed with the COSMIN Risk of Bias Checklist, measurement results were judged against updated criteria for good measurement properties, and evidence quality was graded using an adapted GRADE approach.

The original PROSPERO protocol, registered as CRD420250655032, evaluated broad-band scales as whole instruments. During the review, the unit of analysis was changed to individual ADHD-related subscales in line with updated COSMIN guidance, with no other substantive protocol change reported. The protocol change was documented to preserve methodological transparency and reproducibility.

For the SDQ Hyperactivity-Inattention subscale, structural validity and internal consistency received sufficient ratings supported by moderate-quality evidence. Cross-cultural validity or measurement invariance, reliability, and convergent and discriminant validity also received sufficient ratings supported by high-quality evidence. Content validity was supported only by low-quality evidence, while measurement error and responsiveness were not evaluated.

The reviewers described this profile as comparatively favorable but cautioned that it did not establish the SDQ’s overall superiority for ADHD assessment. The Child and Adolescent Disruptive Behavior Inventory (CADBI) Inattention and Hyperactivity/Impulsivity subscales showed sufficient results for structural and cross-cultural validity with high-quality evidence and for internal consistency with moderate-quality evidence. The Five to Fifteen (FTF) Attention and Hyperactive/Impulsive subscales had sufficient structural validity and internal consistency results supported by moderate-quality evidence.

Single studies also reported sufficient test–retest reliability and convergent validity for the FTF subscales. The Child and Adolescent Behavior Inventory (CABI) Attention Deficit and Hyperactivity subscale had high-quality evidence for reliability, whereas its structural validity and internal consistency findings were supported by low-quality evidence. For the CBCL Attention Problems subscale, two studies with very good methodological quality found insufficient measurement invariance between US and Australian populations and between Dutch and Israeli populations.

The cross-cultural evidence was graded high quality. Under COSMIN criteria, this insufficient performance supported a recommendation against use. The authors advised caution for users outside North America while noting that the review evaluated the CBCL Attention Problems syndrome scale rather than its separate DSM-oriented ADHD scale.

Evidence for several other instruments was limited or unsatisfactory. ANSER-PQ, CBAI, YI-4 and DSMD were each represented by one study that did not meet the minimum methodological threshold for evidence synthesis, so their results were described separately. The ECI-4 was supported by multiple studies examining internal consistency and convergent and discriminant validity.

Its content-validity evidence was low quality because no relevant development or content-validity studies were identified, while its internal-consistency evidence was downgraded because structural-validity studies were lacking. The ABC and CSI-4 were each represented by two studies conducted by the same research group. Across the 36 included studies, internal consistency was examined in 28, structural validity and convergent validity in 16 each, discriminant validity in 13, cross-cultural validity or measurement invariance in nine, and test–retest reliability in eight.

Content validity, criterion validity, measurement error and responsiveness remained important gaps. The limited quantity of research and heterogeneous methodological quality prevented any instrument from being assessed across all COSMIN properties. Restriction to English-language publications may also have constrained the assessment of cross-cultural validity and measurement invariance.

Other limitations concerned the COSMIN framework’s treatment of some analytical choices, including rotation methods and eigenvalue criteria in exploratory factor analyses, and the need for reviewer judgment when development evidence was incomplete. For several older instruments, unavailable manuals or sparse original reports also required greater reviewer judgment for content-validity ratings.

Subject of Research: COSMIN-based evaluation of ADHD-related subscales in broad-band behavioral rating instruments for children

Article Title: How do the measurement properties of the broad-band scales for attention-deficit/hyperactivity disorder in children perform based on the COSMIN guidelines?

Article References: Chen, Y., Cai, Q., Hu, S., Zhao, Z., Wang, J., & Niu, L. (2026). How do the measurement properties of the broad-band scales for attention-deficit/hyperactivity disorder in children perform based on the COSMIN guidelines?. BMC Pediatrics, 26(1), Article 907. https://doi.org/10.1186/s12887-026-07661-1

Image Credits: AI Generated

DOI: 10.1186/s12887-026-07661-1

Keywords: ADHD, children, behavioral rating scales, COSMIN, measurement properties, psychometrics, SDQ, CBCL, systematic review, cross-cultural validity, measurement, properties