JMIR Publications

ABSTRACT

Background:

Cancer patients experience complex physical, psychosocial, and behavioral challenges that require comprehensive and continuous support. This need has become increasingly urgent due to the rising global cancer burden and the limited scalability and accessibility of traditional care models. In response, conversational agents (CAs) have emerged as promising digital interventions for enhancing cancer care. However, evidence regarding their feasibility and effectiveness remains limited.

Objective:

This review aimed to evaluate the feasibility and effectiveness of CAs in supporting cancer patient care and to summarize the key characteristics of CA interventions to inform future design and implementation.

Methods:

We systematically searched PubMed, Cochrane Library, Web of Science, and Embase through February 3, 2025. Additional searches of reference lists and clinical trial registries were performed to identify gray literature. Eligible studies included randomized controlled trials (RCTs) and nonrandomized intervention studies (NRIs) that evaluated CA-delivered interventions targeting health outcomes in cancer patients. Two reviewers independently selected studies and extracted data. Study quality then appraised using the Cochrane Risk of Bias 2.0 tool for RCTs and the JBI Critical Appraisal Checklist for NRIs. Extracted data included study characteristics, CA features, and implementation outcomes including feasibility, acceptability and usability. Meta-analyses were conducted to assess the effects of CAs on physical activity, pain, anxiety, depression, psychological distress, and quality of life.

Results:

Seventeen studies involving 1817 cancer patients were included. CA interventions supported various aspects of care, including tumor-specific management, symptom control, radiotherapy support, mental health, physical activity, genetic counseling, and clinical trial information. Overall, CAs demonstrated good feasibility, acceptability, and usability; however, considerable variability in retention rates and long-term user engagement was observed. Meta-analysis showed significant improvements in physical activity (MD=1.44, 95% CI: 0.36 to 2.52, P<.01), pain (MD = -0.91, 95% CI: -1.44 to -0.38, P<.01), anxiety (SMD = -0.19, 95% CI: –0.35 to –0.02, P=.02), and quality of life (SMD = 0.35, 95% CI: 0.03 to 0.67, P=.03). No significant effects were observed for depression (SMD = –0.07, 95% CI: –0.42 to 0.27, P=.68) or psychological distress (SMD = -0.33, 95% CI: –0.66 to 0.01, P=.06).

Conclusions:

CAs demonstrate promising effectiveness in promoting physical activity, reducing pain and anxiety, and improving quality of life among cancer patients. However, evidence regarding their impact on depression and psychological distress remains limited. CAs generally exhibited favorable feasibility, acceptability and usability; nonetheless, notable variability in long-term engagement across studies highlights the need for further optimization. These findings support the integration of CAs into broader cancer care frameworks. Future development should leverage innovative theoretical models and incorporate artificial intelligence (AI)-powered large language models (LLMs), in combination with rule-based mechanisms, to enhance both the effectiveness and sustainability of CA-based interventions. Clinical Trial: PROSPERO CRD42025645982