This study will evaluate whether an artificial intelligence chatbot can help patients better understand radiation therapy for localized prostate cancer. Participants will be adults with prostate cancer who are planning definitive radiation therapy at Brigham and Women's Hospital. All participants will receive standard radiation therapy education and clinician-led discussions as part of routine care. In addition, participants will interact with an educational chatbot about prostate cancer radiation therapy. The chatbot is designed to provide plain-language information about treatment logistics, preparation, possible side effects, follow-up, and when to contact the care team. The chatbot will not provide diagnosis, treatment recommendations, individualized medical advice, or replace conversations with clinicians. Participants will complete questionnaires before and after using the chatbot. The study will measure changes in patient understanding, confidence in managing symptoms, usability of the chatbot, and safety-related concerns identified during clinician review of chatbot responses.
Study Type
INTERVENTIONAL
Allocation
NA
Purpose
SUPPORTIVE_CARE
Masking
NONE
Enrollment
50
Participants will interact with an LLM-based educational chatbot focused on prostate cancer radiation therapy. The chatbot provides plain-language information about treatment logistics, preparation, expected follow-up, possible urinary, bowel, sexual, and hormonal side effects, symptom monitoring, and when to contact the care team. The chatbot is used as an adjunct to standard radiation therapy education and clinician-led discussions. The chatbot is based on ChatGPT 5.2 deployed within a secure Mass General Brigham Azure environment and configured with physician-reviewed educational source materials and safety guardrails. It is designed to provide educational information only and does not diagnose, recommend treatment choices, provide individualized medical advice, or alter clinical management. Participant interactions will not be used to train, fine-tune, or update the underlying foundation model.
Change in PROMIS Self-Efficacy for Managing Symptoms Score
Within-participant change in symptom self-efficacy will be measured using the Patient-Reported Outcomes Measurement Information System (PROMIS) Self-Efficacy for Managing Symptoms instrument. Scores are reported as PROMIS T-scores, with a mean of 50 and standard deviation of 10 in the reference population. Higher scores indicate greater confidence in managing symptoms, which is a better outcome. There is no fixed minimum or maximum T-score, although typical PROMIS T-scores generally range from approximately 20 to 80.
Time frame: Baseline before chatbot interaction to immediately after chatbot interaction during the same study session
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.