A review of interventions to reduce inter-observer variability in volume delineation in radiation oncology

• Published in: Journal of medical imaging and radiation oncology 2016-06, Vol.60 (3), p.393-406
• Main Authors: Vinod, Shalini K, Min, Myo, Jameson, Michael G, Holloway, Lois C
• Format: Article
• Language: English
• Subjects: contouring variation; Humans; inter-observer variability; Lung Neoplasms - diagnostic imaging; Observer Variation; Positron-Emission Tomography; Radiation Oncology - standards; radiotherapy; Tomography, X-Ray Computed; volume delineation
• ISSN: 1754-9477; 1754-9485
• DOI: 10.1111/1754-9485.12462

Introduction Inter‐observer variability (IOV) in target volume and organ‐at‐risk (OAR) delineation is a source of potential error in radiation therapy treatment. The aims of this study were to identify interventions shown to reduce IOV in volume delineation. Methods Medline and Pubmed databases were queried for relevant articles using various keywords to identify articles which evaluated IOV in target or OAR delineation for multiple (>2) observers. The search was limited to English language articles and to those published from 1 January 2000 to 31 December 2014. Reference lists of identified articles were scrutinised to identify relevant studies. Studies were included if they reported IOV in contouring before and after an intervention including the use of additional or alternative imaging. Results Fifty‐six studies were identified. These were grouped into evaluation of guidelines (n = 9), teaching (n = 9), provision of an autocontour (n = 7) and the impact of imaging (n = 31) on IOV. Guidelines significantly reduced IOV in 7/9 studies. Teaching interventions reduced IOV in 8/9 studies, statistically significant in 4. The provision of an autocontour improved consistency of contouring in 6/7 studies, statistically significant in 5. The effect of additional imaging on IOV was variable. Pre‐operative CT was useful in reducing IOV in contouring breast and liver cancers, PET scans in lung cancer, rectal cancer and lymphoma and MRI scans in OARs in head and neck cancers. Conclusion Inter‐observer variability in volume delineation can be reduced with the use of guidelines, provision of autocontours and teaching. The use of multimodality imaging is useful in certain tumour sites.