Analysis of entrance surface dose in general radiographies using body mass index classification

Dose audit is an important topic in radiology and variations in dose attributable to sex have not been fully investigated. The primary purpose of our study was to determine the entrance surface dose (ESD) for men and women, from which we established an easy estimation equation for ESD, using the body mass index (BMI) as a variable. From April 2000 to March 2002, 15424 patients (8586 men, 6838 women; mean age, 58 years old; age range, 20 to 90 years old) were recruited in this study. Estimated equation of ESD was obtained from the dispersion data of BMI and ESD. Welch's t-test was used as the test of significance. P values evaluated levels of significance at 5%. Analysis for correlation coefficients was used for Peason's correlation coefficient test. And, we also performed a simple regression analysis. The results of the simple regression equation with BMI as variable (y: ESD [mGy], x: BMI [kg/m(2)), and the median values of ESD were as follows: for chest PA, men, 0.22, y = 0.0145 x - 0.0876, women, 0.20, y = 0.0126 x - 0.0704; for abdomen AP, men, 2.12, y = 0.2707 x - 3.7206, women, 1.65, y = 0.2561 x - 3.7989; for pelvis AP, men, 1.87, y = 0.0517 x + 0.8753, women, 1.77, y = 0.0674 x + 0.4775; for lumbar spine AP, men, 3.08, y = 0.3462 x - 4.1416, women, 2.66, y = 0.2525 x - 2.8131; and for lumbar spine lateral, men, 9.31, y = 0.6012 x - 3.7216, women, 7.42, y = 0.3147 x + 0.3384. The results of ESD had a significant difference between sexes (P < 0.05). Chest PA, abdomen AP, lumbar spine AP, and lumbar spine lateral each had a good correlation between ESD and BMI (r: 0.4 to 0.8). We successfully obtained the estimation equation using BMI as variable.