Internal medicine and pediatric residents were surveyed online for 42 days. After obtaining demographics and training levels, respondents were asked to rank five different radiation dose quantities, each corresponding to one of the five ACR relative radiation levels (RRLs) expressed using different dose terms. Respondents ranked the choices from least to greatest (ie, from 1 to 5) or indicated if all five were equal. For the final question, the same dose quantity was expressed five different ways.
Fifty-one medicine and 45 pediatric residents responded (a 44% response rate). Mean differences in rankings were as follows: for chest x-rays, 0.109 (95% confidence interval [CI], −0.018 to 0.236); for cross-country flights, 0.462 (95% CI, 0.338 to 0.585); for natural background radiation, −0.672 (95% CI, −0.793 to −0.551); for cancer risk, −0.294 (95% CI, −0.409 to −0.178); and for ACR RRL, 0.239 (95% CI, 0.148 to 0.329). Statistically significant differences were found in the distributions of rankings (P < .001) and percentage of correct rankings across each radiation dose term (P < .001), with the ACR RRL having the highest percentage of correct rankings (61.2%).
Adult and pediatric physicians consistently over- or underestimated radiation dose quantities using different terms to express radiation dose. These results suggest that radiation dose information should be communicated using standard terminology such as the ACR RRL scale to foster consistency and improve the accuracy of physicians’ radiation risk perceptions.