Dokumentart(en): Sammelwerksbeitrag Forschungsergebnis Online-Publikation
Titel der Veröffentlichung: Towards Optimized Clinical Cut-Off Scores for Mental Disorders in Rehabilitation: A Machine Learning Case Study in Depression

Identifying and treating mental health disorders is a key challenge in rehabilitation care (Baumeister et al., 2017). Mental health self-report questionnaires and rating-scales with clinical sum score cut-offs are often used for the screening of mental disorders. Machine learning (ML) has been shown to achieve high accuracy in various predictive tasks (Dwyer et al., 2018; Yarkoni & Westfall, 2017). However, ML has not been utilized to improve the accuracy of screening instruments. The present study aims 1) to derive optimized traditional sum score cut-offs for commonly used screening instruments in rehabilitation and 2) explore the potential of ML to further improve the accuracy of these screening instruments.