Learning curve of residents of a radiation oncology service in a training program.
CAROLINA DE LA PINTA,
Spain
PO-1033
Abstract
Learning curve of residents of a radiation oncology service in a training program.
Authors: CAROLINA DE LA PINTA1, Mireia Valero1, Elisa Pinto2, Victor Duque1, Lira Pelari1, Francisco López Mesa1, Elena Centelles1, Maria Elena Hernando2, Blanca Iñigo2, Eva Fernández-Lizarbe1, Elena Canales3, Raquel García-LaTorre3, Manuel Garví4, Vanesa Pino4, Sonsoles Sancho García1
1Ramón y Cajal Hospital, Radiation Oncology, Madrid, Spain; 2Universidad Politécnica de Madrid, Madrid, Spain, Bioengineering and Telemedicine Centre, Madrid, Spain; 3Ramón y Cajal Hospital, Radiology, Madrid, Spain; 4Ramón y Cajal Hospital, Radiation Therapist, Madrid, Spain
Show Affiliations
Hide Affiliations
Purpose or Objective
Resident training depends on the volume of patients in the training center. A pilot contouring training program for residents has been implemented in our center. This program consists of several contouring modules by pathology. Our study shows the first data obtained from the program comparing the delineation of GTV of non-spinal bone metastases on MRI and CT.
Material and Methods
After institutional review boards approved the study, we analyzed anonymized CT and MRI scans obtained from nine non-spinal bone metastases. Five radiation oncology residents at our center used CT and MRI to delineate nine GTVs of non-spinal bone metastases. The locations of the non-spinal bone metastases were pelvis and skull. The cases show different characteristics: blastic and lytic metastases, different primaries (lung, breast, prostate, rectum and urothelial). GTV volumes on CT and MRI were compared. The concordance index according to Landis and Koch was evaluated.
Results
The median GTV volume was 3.01 cc on CT (0.076-79.14 cc), while on MRI it was 45.17 cc (0.1-75.12 cc). The interobserver variance and standard deviation in CT were lower than in MRI (674.3 vs. 834.1, and 25.97 vs. 28.88, respectively). The concordance index was slight (0.2) between CT and MRI in all patients. If we divide the residents into first and second year and compare them with third- and fourth-year residents, the median GTV volume on MRI was 51.23 and 42.69 cc in group 1 and 2 (p=0.018), versus 2.32 cc and 3.78 on CT in group 1 and 2 (p=0.022). The variance and standard deviation among observers in group 2 in CT and MRI were lower than in group 1 (780.80 vs. 619.62, and 27.94 vs. 24.89 respectively). The agreement index was moderate (0.53 and 0.51) between CT and MRI by groups.
Conclusion
This study demonstrates the evolving learning curve of physicians in training. This curve will improve if residents have at their disposal a regulated and contoured training plan with expert supervision. We observed a great difference in the contour between CT and MRI by proposing a specific MRI study module within the training program