Decision trees compare caregiver ratings with clinical ADHD diagnoses
Source check: Source checked. Not independent scientific confirmation.
Finding: In 326 Italian children and adolescents referred for suspected ADHD, a decision tree using caregiver questionnaires agreed with clinician diagnoses in 74% of leave-one-out evaluations. Analysis of fitted-model errors linked autism-related symptom ratings to classification patterns. This is a retrospective comparison with clinical assessments, not an independently validated remote diagnostic service. Shared questionnaire information, a selected mostly male sample and exploratory error analysis limit interpretation.
Technology: Decision tree; random forest and support-vector-machine comparisons
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
- AI contribution
- The tree learns questionnaire-score splits that classify clinician-assessed ADHD status; learned classifications and their errors support analysis of which reported symptom patterns depart from the full in-person evaluation.
- Test and result
- Among 342 referred children, sixteen with extensive missing data were excluded, leaving 326. A decision tree was evaluated against DSM-5-based clinician assessments using leave-one-out validation; accuracy was 74%, sensitivity 92% and specificity 50%. Separate error-pattern analyses used fitted full-sample predictions.
- Limitation / still unknown
- Generalization to other referral populations, independent diagnostic reference assessment and prospective telehealth effectiveness remain unestablished. The clinicians’ diagnostic process incorporated caregiver information also available to the algorithm; exploratory autism/error associations do not establish causal confounding or independently validated subgroup performance. Supplementary data-cleaning details were not reviewed.
Evidence sources and annotation context
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC9875192/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 21 Nov 2026. Review scope: Primary relevant participant/reference, questionnaire, tree/LOO Methods, Results and limitations. Supplementary data-cleaning appendix not inspected. Exact primary identity/date and fresh primary PubMed correction/retraction relationships checked..
No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.