Results and Resident Evaluation of the 2007 American College of Radiology In-Training Examination in Radiation Oncology

Purpose: To report the results, demographic data, and resident evaluation of the 2007 ACR radiation oncology in-training (TXIT) examination. Methods: The 2007 TXIT examination consisted of 360 multiple-choice questions covering 13 different subject areas. It included 9 demographic questions and 7 evaluation items. Five hundred seventy-two residents from 85 institutions took the 2007 TXIT examination. Results: The median raw score was 218.3 ± 29 (range, 140-295). The mean item difficulty was 60.7%, with mean item discrimination of 0.19. The reliability coefficient was 0.92. Of the respondents to the demographic questions, 550 (96.2%), 493 (86.2%), and 402 (70.3%) residents had formal physics, biology, and clinical oncology instruction, respectively. Clinical experience was highest in breast, genitourinary, and lung cancers and lowest in soft-tissue sarcomas, pediatric tumors, and orbital neoplasms. Formal instruction in cancer biology, intensity-modulated radiation therapy, and statistics was reported by 416 (72.7%), 383 (67.0%), and 245 (42.8%) residents, respectively. On the basis of residents' responses to the evaluation questions, 31 (5.4%) did not know that the examination book was available to them after the test, 47 (8.2%) did not know that an answer key was available, and 79 (13.8%) residents did not know that rationales were available for future study. Conclusions: The reliability estimate for the ACR TXIT examination was high, typical of carefully constructed in-training examinations. The availability of the examination book, answer key, and rationales was not known by approximately 5% to 15% of residents taking the TXIT examination; therefore, a better method of informing the residents of these educational tools is needed.