55 Radiotherapy Peer Review in Neuro-Oncology: National Survey of Practice, Barriers and AI Integration

Abstract Introduction In neuro-oncology, radiotherapy peer review is central to quality assurance, due to anatomical complexity and proximity to critical structures. However, current UK practice is not well described. Method An online survey was designed by the Tessa Jowell Academy Peer Review Working Group and distributed nationally (October–December 2025). Survey questions covered peer review structure, scope, multidisciplinary involvement, artificial intelligence (AI) use, and perceived impact. Data were analysed using descriptive statistics at both institutional- and practitioner-level. Results Seventy-two clinicians from 38 hospitals responded, representing 29 out of 30 UK neuro-oncology networks and one in the Republic of Ireland. Most respondents were oncologists (60/72, 83%), and 34/72 (47%) had practised >10 years. Structure Most hospitals (30/38, 79%) peer reviewed all cases and held weekly meetings (29/38, 76%); however, 27/72 (38%) reported insufficient protected time for peer review and 2/38 (5%) did not peer review. Scope and practice Most reviewed contours only (28/38, 74%), while 7/38 (18%) reviewed both contours and final plans. Minor changes occurred in > 25% cases in 12 centres; major changes were rarer. MDT involvement Multidisciplinary involvement varied, radiographers participated in 16/38 centres (42%), physicists in 12/38 (32%), and neuroradiologists in 10/38 (26%). Use of AI AI-based auto-contouring was used by 51/72 (71%) respondents with 60% of AI users describing mixed or negative experiences, particularly with complex neuroanatomy. Perception Most respondents (67/72, 93%) viewed peer review positively. Key challenges reported were resource-related, particularly lack of protected time. Suggested improvements included dedicated job-planned time, increased multidisciplinary input, and regional collaboration. Conclusions Peer review is widely embedded in UK neuro-oncology and valued but remains heterogeneous and resource-constrained. As AI integration expands, strengthened workforce support and governance are essential to retain quality.