RESUMEN
PURPOSE: The purpose of this study is to estimate the accuracy of diagnosing a radio-lucent soft tissue foreign body with using ultrasound according to the different sizes and depth of the foreign bodies in an experimented model. METHODS: In the forefeet of swine, we made 45 incisions (4 in each forefeet) of 2 mm, 5 mm and 10 mm sized toothpicks at depths of 1 cm, 2 cm and 3 cm each. 15 incisions without foreign bodies were cannulated, simulating damage to the soft tissue made by a non captured foreign body. The twelve emergency physicians kept records not only for the presence or absence of a foreign body, but also for the depth of it. We then calculated the sensitivity, specificity, PPV (positive predictive value) and NPV (negative predictive value) according to the sizes of the toothpicks and the depths. RESULTS: The sensitivity, specificity, PPV and NPV of all the depths and sizes were under 80%. The results of the staff and residents were not significantly different. The result of a 3 cm depth was especially poor. CONCLUSION: In our model, ultrasound performed by emergency physicians was neither sensitive nor specific for diagnosing a soft tissue foreign body of various sizes and depths.