Digital and AI-enhanced psychosocial interventions in breast cancer patients: a systematic review and meta-analysis
摘要
Digital and AI-supported psychosocial interventions are increasingly implemented to address psychological distress among women with breast cancer. However, their effectiveness and sustainability remain uncertain. This review aimed to systematically evaluate randomised controlled trials (RCTs) assessing the impact of digital and AI-enhanced psychosocial interventions on stress, anxiety, depression, and health-related quality of life (HRQoL).
MethodsFollowing PROSPERO registration and PRISMA 2020 guidelines, comprehensive searches of multiple databases were conducted up to June 2025. Eligible studies included RCTs involving breast cancer populations and reporting validated psychosocial outcomes. Risk of bias was assessed using the Cochrane RoB 2.0 tool, and certainty of evidence was evaluated with GRADE. Random-effects meta-analyses using standardised mean differences (Hedges’ g) were conducted to pool outcomes measured with different instruments.
ResultsThirty-five RCTs (n ≈ 7551 participants) met inclusion criteria. Pooled analyses indicated modest but consistent reductions in stress, anxiety, and depressive symptoms, alongside improvements in HRQoL. Interventions grounded in cognitive behavioural therapy, mindfulness, or psychoeducation produced the most reliable outcomes. AI-assisted and blended formats tended to show stronger effects compared with static or minimally interactive platforms. However, heterogeneity in intervention design, short follow-up periods, reliance on self-reported outcomes, and small single-centre samples limited certainty. Cultural adaptation and representation from low- and middle-income countries were minimal.
ConclusionDigital and AI-supported psychosocial interventions show promise in reducing distress and improving HRQoL among women with breast cancer. Evidence certainty remains moderate, constrained by methodological limitations. Larger, multicentre trials with longer follow-up and diverse populations are needed to establish sustainability, generalisability, and clinical integration.
RelevanceThese findings suggest that digital and AI-enhanced psychosocial interventions may serve as adjuncts to conventional care, with potential to improve access to supportive oncology services. Further high-quality, long-term studies are needed before informing guidelines, policy, or widespread implementation.
Trial registrationPROSPERO (CRD420251074813; registered 16 June 2025).