e24215 Background: We developed a website that allows hematologic cancer patients undergoing stem cell transplant (SCT) to access a searchable, filterable database of peer support narratives—written first-hand accounts of SCT experiences and advice created by SCT survivors. The website, called Mosaic (Mosaico in Spanish), pairs the narratives with resources informed by patients, clinicians, and behavioral scientists including psychoeducation, support for patient-provider communication, and stress management and coping resources. This presentation will describe feedback about Mosaic from interviews with SCT clinicians and discuss pros and cons of delivering peer support in written form. Methods: We conducted semi-structured interviews with 13 SCT clinicians (3 oncologists, 3 nurse practitioners, 3 social workers, 3 registered nurses, 1 physician assistant; 92% women; 53.8% white). After describing Mosaic’s content and features, we asked for feedback on Mosaic’s potential benefits, harms, and recommended changes. Interviews were audio recorded, transcribed, and content analyzed by trained coders. Results: All clinicians noted potential benefits of Mosaic. Some focused on novel informational benefits (e.g., getting “the real patient experience,” giving patients “a sense of less isolation”). Others liked that Mosaic could help patients develop reasonable expectations of SCT, including things that can go wrong. A few clinicians liked features designed to meet patients’ differing needs (e.g., “they can filter out what they [don’t] want”). Overall, there was enthusiasm for Mosaic as a patient resource. Yet, all clinicians also mentioned at least one concern. Several wanted to see Mosaic’s full content before recommending it to patients. Others wanted evidence showing Mosaic’s benefits. A more frequent concern about misinformation suggested the need to edit narratives to remove medical advice and identifiable information. Some clinicians were concerned that patients might inappropriately worry about medical risks and complications described in narratives, although others felt that Mosaic’s disclaimers and other features mitigated this concern. Some recommendations for Mosaic addressed this concern (e.g., clearly disclosing that “it’s not medical advice” and that their SCT “experience may be completely different”). Other recommendations included a need for disclosures to help patients make an informed decision about using Mosaic and the need to monitor for adverse events such as distress. A recommended new feature included grouping narratives by year of SCT due to changing treatments. Conclusions: SCT clinicians felt Mosaic could provide important emotional and informational benefits. Their concerns and recommendations will be used to refine Mosaic to optimize safety and potential benefits in preparation for a clinical efficacy trial.