No abstract

No abstract

Oral poster abstracts Methods: We recruited 282 cases of which ultrasonographic examination were done by ten 2nd year residents. EFW by ultrasonography was measured within 48 hours prior to delivery and actual birth weight (BW) was collected. We calculated percentage error that means the difference between ultrasonographic EFW and actual BW divided by actual BW and multiplied by 100. We also obtained the values at upper and lower 25% deviation from median. The range between 25% and 75% quartiles was obtained for each resident. The Pearson correlation was used for statistical analyses. Results: The correlation coefficient between the ultrasonographic EFW and actual BW was 0.907 (P < 0.001). EFW less than 10% of difference was in 74.8% of cases. Ultrasongraphic performance varied according to each resident. The narrowest range between 25% and 75% quartiles was 8.6% and the widest range was 15.7%. Two residents showed median values of higher than 2 percent error and one showed median values of lower than −2 percent error. Conclusions: Although ultrasonographic EFW performed by residents is correlated with the actual BW, the accuracy of EFW is affected by personal skill even in the same grade of residency. We suggest that percentage errors can be used as an evaluation and tuition tool to improve ultrasonographic performance of measuring fetal weight.