Artificial intelligence as a predictive tool for mental health status: Insights from a systematic review and meta-analysis
Abstract
This systematic review and meta-analysis evaluates the effectiveness of AI-driven tools, particularly conversational agents (CAs), in alleviating psychological distress and improving mental health outcomes. The focus is on their impact across diverse populations, including clinical, subclinical, and older adults. A comprehensive search was conducted in PubMed, Google Scholar, Elsevier, and Scopus using specific MeSH terms and keywords such as “Artificial Intelligence,” “Machine Learning,” “Natural Language Processing,” “Depression,” and “Anxiety.” The timeframe included studies published between January 2000 and July 2024. Inclusion criteria comprised peer-reviewed original research articles, cohort studies, and case reports focusing on AI tools for mental health. Systematic reviews, secondary sources, and non-English publications were excluded. Random-effects meta-analysis was conducted using standardized mean differences, with effect sizes synthesized in forest plots. Twenty studies were included in the qualitative synthesis and six in the quantitative meta-analysis. The analysis demonstrated that AI-based CAs significantly reduce anxiety (Cohen’s d = 0.62, p < 0.01) and depression (Cohen’s d = 0.74, p < 0.001), with higher effectiveness observed in multimodal CAs compared to text-only systems. However, the long-term impact remains inconsistent due to variability in follow-up durations and methodological heterogeneity. Some studies lacked extended observation periods or…
How strong is the evidence?
Evidence Hub reads this as a meta-analysis, done in people. Its evidence level is 1 of 5 (Highest · pooled evidence). Pooled evidence from several studies sits at the top of the evidence pyramid.
Read the paper
Related studies
- Active Learning during Federated Learning for Object Detection
- Fairness-aware federated learning
- Decentralized Federated Learning
- Building a Personally Identifiable Information Recognizer in a Privacy Preserved Manner Using Automated Annotation and Federated Learning
- Preservation of Patient Level Privacy: Federated Classification and Calibration Models
- Efficient explainable federated learning
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.