Generative AI Integration in Cancer Care: Mapping Applications and Patient Perspectives

Article Impact Level: HIGH
Data Quality: STRONG
Summary of  European Journal of Cancer https://doi.org/10.1016/j.ejca.2026.116990
Dr. Bradley D. Menz et al.

Points

Flinders University researchers led an international scoping review identifying 2,441 records to assess how cancer patients, survivors, and caregivers perceive generative artificial intelligence applications.
Synthesis of thirty-two eligible studies revealed that patients generally favored generative artificial intelligence when utilized to simplify complex clinical reports and improve health information readability.
Participant trust in artificial intelligence outputs depended heavily on whether generated responses were personalized, emotionally appropriate, and perceived to be clinically accurate by health consumers.
Accountability perspectives were evaluated in only three studies, whereas zero published studies investigated patient views regarding safety, transparency, algorithmic equity, or human autonomy.
Researchers advocate for co-designing artificial intelligence systems alongside oncology patients to establish robust governance, consent frameworks, monitoring protocols, and patient-centered clinical workflows.

Summary

This scoping review evaluated patient, survivor, and caregiver perspectives regarding the implementation of generative artificial intelligence (AI) in oncology clinical practice and healthcare delivery. Led by Bradley Menz and Ashley Hopkins at Flinders University and published in the European Journal of Cancer, the investigation systematically mapped existing literature to identify application contexts, engagement metrics, and critical governance gaps. Given the rapid integration of large language models into medical communications, the research sought to determine how healthcare consumers perceive generative AI utility, trust, and ethical boundaries in cancer care.

A comprehensive literature search identified 2,441 records, of which only 32 studies directly examined the perspectives of individuals affected by cancer. The synthesized data indicated that patients generally held favorable attitudes toward generative AI when applied to improving the accessibility, readability, and simplification of clinical reports and health information. Participant trust was closely correlated with response personalization, emotional appropriateness, and perceived accuracy. However, existing literature remained concentrated within narrow application contexts, leaving key operational domains, workflow integration, and consent processes largely unexamined.

Critical evaluative gaps regarding responsible AI deployment were identified across the included literature. Patient perspectives on accountability were assessed in only three of the 32 studies, while zero studies evaluated patient views on safety, transparency, algorithmic equity, or the preservation of human autonomy. The authors conclude that responsible AI integration requires participatory co-design strategies involving oncology patients, caregivers, and survivors. Expanding research frameworks to encompass ethical considerations, governance structures, and broader clinical applications is essential to ensure patient-centered implementation in oncology.

Link to the article: https://www.ejcancer.com/article/S0959-8049(26)00771-9/fulltext

References

Menz, B. D., Scarfo, N. L., Cornelisse, E., Shahnam, A., Chin-Yee, B., Rickard, C., Haseloff, M., Lau, A. Y. S., Ugalde, A., Paterson, C., Sorich, M. J., Chan, R. J., Rowland, A., & Hopkins, A. M. (2026). Perspectives of people affected by cancer toward generative AI in oncology: A scoping review of applications, evidence gaps, and future priorities. European Journal of Cancer, 246, 116990. https://doi.org/10.1016/j.ejca.2026.116990