Randomized controlled trial compares AI-enhanced and clinician-guided writing for cancer survivors' mental health, indicating effective digital support options.
Aims Digital tools offer scalable support for cancer survivors' psychological well‐being, however, direct efficacy comparisons between artificial intelligence (AI)‐Enhanced and clinician‐guided digital interventions are scarce. This study evaluated an AI‐Enhanced expressive writing intervention against clinician‐guided digital support and usual care for cancer survivors. Methods In this randomized controlled trial, 120 cancer survivors (111 completers) were randomized (1:1:1) to AI‐Enhanced, clinician‐guided, or usual care groups for a 4‐week expressive writing intervention. Outcomes, including depression, anxiety, fear of progression (FoP), and resilience, were assessed at baseline, post‐intervention, and 1‐month follow‐up between September 2024 and April 2025. Data were analyzed using repeated‐measures ANOVAs and thematic analysis of semi‐structured interviews. Results Both AI and clinician‐guided interventions significantly reduced depression (AI: d = 0.92; Clinician: d = 0.98) and anxiety (AI: d = 0.59; Clinician: d = 0.66) versus usual care, with comparable efficacy. However, clinician‐guided support demonstrated greater efficacy in reducing FoP ( d = 0.62 vs. AI). Resilience improved over time across all groups, with no significant inter‐group differences. Qualitatively, AI offered reduced stigma and accessibility, while the clinician‐guided approach provided valued empathetic connection. Conclusion AI‐Enhanced expressive writing is a viable, scalable tool for reducing depression and anxiety in cancer survivors. Yet, the greater efficacy of clinician‐guided digital support for complex issues like FoP highlights the value of human‐mediated empathy. This advocates for hybrid models in psychological care that integrate AI's scalability with the relational depth of clinical expertise to enhance survivor well‐being. Trial Registration Chinese Clinical Trial Registry (ChiCTR2400089455)
No takes yet. Share an insight, caveat, or question.
Yang et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: