Grace Lee1, Anthony Fyles2, B C John Cho2, Alexandra M Easson3, Louis L Fenkell4, Nicole Harnett5, Lee Manchul2, Phillip K Tran2, Wei Wang2, Tim Craig2, Gregory J Czarnota6, Robert E Dinniwell7. 1. Radiation Medicine Program, Princess Margaret Hospital, Toronto, Ontario, Canada. 2. Radiation Medicine Program, Princess Margaret Hospital, Toronto, Ontario, Canada; Department of Radiation Oncology, University of Toronto, Toronto, Ontario, Canada. 3. Department of Surgical Oncology, Princess Margaret Hospital, Toronto, Ontario, Canada. 4. Radiation Medicine Program, Princess Margaret Hospital, Toronto, Ontario, Canada; Department of Radiation Oncology, University of Toronto, Toronto, Ontario, Canada; Department of Radiation Oncology, Stronach Regional Cancer Centre, Newmarket, Ontario, Canada. 5. Radiation Medicine Program, Princess Margaret Hospital, Toronto, Ontario, Canada; Department of Radiation Oncology, University of Toronto, Toronto, Ontario, Canada; Cancer Care Ontario, Toronto, Ontario, Canada. 6. Radiation Medicine Program, Princess Margaret Hospital, Toronto, Ontario, Canada; Department of Radiation Oncology and Imaging Research, Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada; Department of Medical Biophysics, University of Toronto, Toronto, Ontario, Canada. 7. Radiation Medicine Program, Princess Margaret Hospital, Toronto, Ontario, Canada; Department of Radiation Oncology, University of Toronto, Toronto, Ontario, Canada. Electronic address: rob.dinniwell@rmp.uhn.on.ca.
Abstract
PURPOSE: Breast cancer is managed by a multidisciplinary team with a goal for the timely provision of high quality care. Given radiation oncologist (RO) time constraints, an opportunity arises for task delegation of breast seroma target delineation to an advanced practice clinical specialist radiation therapist (CSRT) with clinical and technical expertise to facilitate treatment planning. To explore this further, we quantitatively evaluated the variability in post-surgical seroma delineation between the CSRT and ROs. METHODS: Specialized site specific training was provided to the CSRT, who, with 7 ROs, independently contoured the seroma and graded its clarity, using the cavity visualization score (CVS), for 20 patients with clinical stage Tis-2N0 breast tumors. The conformity indices were analyzed for all possible pairs of delineations. The estimated "true" seroma contour was derived from the RO contours using the simultaneous truth and performance level estimation algorithm. Generalized kappa coefficient and center of mass metrics were used to examine the performance level of the CSRT in seroma delineations. RESULTS: The CVS of the CSRT correlated well with the mean RO-group CVS, (Spearman ρ = 0.87, P < .05). The mean seroma conformity index for the RO group was 0.61 and 0.65 for the CSRT; a strong correlation was observed between the RO and CSRT conformity indices (Spearman ρ = 0.95, P < .05). Almost perfect agreement levels were observed between the CSRT contours and the STAPLE RO consensus contours, with an overall kappa statistic of 0.81 (P < .0001). The average center of mass shift between the CSRT and RO consensus contour was 1.69 ± 1.13 mm. CONCLUSIONS: Following specialized education and training, the CSRT delineated seroma targets clinically comparable with those of the radiation oncologists in women with early breast tumors suitable for accelerated partial breast or whole breast radiotherapy following lumpectomy. This study provides support for potential task delegation of breast seroma delineation to the CSRT in our current multidisciplinary environment. Further study is needed to assess the impact of this role expansion on radiotherapy system efficiency.
PURPOSE:Breast cancer is managed by a multidisciplinary team with a goal for the timely provision of high quality care. Given radiation oncologist (RO) time constraints, an opportunity arises for task delegation of breast seroma target delineation to an advanced practice clinical specialist radiation therapist (CSRT) with clinical and technical expertise to facilitate treatment planning. To explore this further, we quantitatively evaluated the variability in post-surgical seroma delineation between the CSRT and ROs. METHODS: Specialized site specific training was provided to the CSRT, who, with 7 ROs, independently contoured the seroma and graded its clarity, using the cavity visualization score (CVS), for 20 patients with clinical stage Tis-2N0 breast tumors. The conformity indices were analyzed for all possible pairs of delineations. The estimated "true" seroma contour was derived from the RO contours using the simultaneous truth and performance level estimation algorithm. Generalized kappa coefficient and center of mass metrics were used to examine the performance level of the CSRT in seroma delineations. RESULTS: The CVS of the CSRT correlated well with the mean RO-group CVS, (Spearman ρ = 0.87, P < .05). The mean seroma conformity index for the RO group was 0.61 and 0.65 for the CSRT; a strong correlation was observed between the RO and CSRT conformity indices (Spearman ρ = 0.95, P < .05). Almost perfect agreement levels were observed between the CSRT contours and the STAPLE RO consensus contours, with an overall kappa statistic of 0.81 (P < .0001). The average center of mass shift between the CSRT and RO consensus contour was 1.69 ± 1.13 mm. CONCLUSIONS: Following specialized education and training, the CSRT delineated seroma targets clinically comparable with those of the radiation oncologists in women with early breast tumors suitable for accelerated partial breast or whole breast radiotherapy following lumpectomy. This study provides support for potential task delegation of breast seroma delineation to the CSRT in our current multidisciplinary environment. Further study is needed to assess the impact of this role expansion on radiotherapy system efficiency.