Effectiveness of artificial intelligence–enhanced cognitive behavioral therapy compared to alternative non-pharmacological therapies for treating depression or anxiety among adults: a systematic review protocol

Abstract

OBJECTIVE: The objective of this systematic review is to evaluate the effectiveness of artificial intelligence-enhanced cognitive behavioral therapy (AI-enhanced CBT) compared to traditional CBT or alternative non-pharmacological therapies in treating depression and anxiety among adults. INTRODUCTION: Mental health disorders, particularly anxiety and depression, affect millions of individuals worldwide. The World Health Organization estimated that in 2021, approximately 359 million people experienced anxiety and 332 million experienced depression, with global prevalence rates of 4.4% and 4.0%, respectively. CBT has long been an effective treatment for these conditions, with structured interventions for symptom reduction. However, AI-enhanced CBT introduces a new dimension by personalizing treatment through adaptive algorithms and automated feedback. While traditional therapies such as psychotherapy, medication, or mindfulness have shown varying degrees of success, the integration of AI with CBT holds potential for scalable, accessible, and individualized care. ELIGIBILITY CRITERIA: This review will include studies that evaluate AI-enhanced CBT for treating adults with depression and/or anxiety. Comparators may include traditional CBT or alternative non-pharmacological interventions. Eligible studies must report clinical outcomes such as symptom reduction, patient engagement, or treatment feasibility. METHODS: The review will follow the JBI methodology for systematic reviews of effectiveness. Comprehensive searches will be conducted in MEDLINE (Ovid), Scopus, Cochrane CENTRAL, CINAHL (Ovid), Embase (Ovid), and Epistemonikos. We will also search relevant gray literature sources, including ProQuest Dissertations and Theses Global (EBSCOhost), OAIster, and Google Scholar. Two reviewers will independently screen studies, appraise quality using the JBI checklist for randomized controlled trials, and extract data using a standardized form. The review will follow PRISMA reporting guidelines, and, if data permit, meta-analysis will be undertaken. REVIEW REGISTRATION: PROSPERO CRD42024580549.

Ähnliche Arbeiten