Questionnaire-based computational screening of adult ADHD

Area of Science:

• Psychiatry
• Psychometrics
• Neurodevelopmental Disorders

Background:

• Attention-Deficit/Hyperactivity Disorder (ADHD) often persists into adulthood, yet diagnosis is challenging due to lengthy, multidisciplinary processes.
• Adult ADHD is frequently under-diagnosed, highlighting the need for efficient screening tools.

Purpose of the Study:

• To develop and validate a psychometric screening scale for identifying adult ADHD.
• To provide a tool for use in both clinical and experimental settings for ADHD assessment.

Main Methods:

• A scale was designed based on DSM-5 criteria and administered to 110 ADHD and 110 control individuals.
• Item reduction used multiple regression, followed by factorial analyses and machine learning for predictive power assessment.

Main Results:

• The scale (TRAQ10) demonstrated strong internal consistency (Cronbach's alpha = .9) and confirmed a 2-factor model.
• Machine learning analysis showed high accuracy in classifying ADHD versus control groups, outperforming comorbidity scales.

Conclusions:

• The developed scale shows adequate performance for screening and hypothesis testing in adult ADHD.
• Generalizability may be limited by age and gender biases in the study sample.