Small retained foreign bodies: what is the limit of detection using current emergency ultrasound equipment?
KeywordsVisual Analog Scale Detection Rate Foreign Body Interventional Radiology Analog Scale
Previous studies of small foreign bodies (FB) have shown a wide range of accuracies of FB detection using animal models, with high accuracy rates for FB > 10 mm and variable accuracy rates for 4 to 5 mm FB.
To determine the lower limit of sonographic detection of FB using current emergency ultrasound equipment in a soft tissue model.
FBs made of metal, glass, wood, and plastic (3 of each) 1 x 1 x 3 mm in size were placed at a depth of 0.5-2.0 cm in 12 pork feet. 8 feet were punctured without FB placement. Pork feet were submerged during this process to minimize air in tissue. 7 ED sonologists with > 2 years experience were blinded to overall number, type and depth of FB, but not to size. FB sites were scanned by each sonologist using either a hockey stick or traditional linear array transducer in a randomized pre-assigned order. Sonologist confidence in the diagnosis was reported using a visual analog scale for each site. Sensitivity, specificity, positive and negative predictive values (PPV and NPV) with 95% confidence intervals were calculated. To determine if sonologist confidence differed by perceived presence or absence of a foreign body, paired t-test was used.
140 ultrasound scans were performed which reported sensitivity, specificity, PPV and NPV as 50% (95% CI: 39%-61%), 50% (37%-61%), 60% (48%-72%), and 40% (28%-52%) respectively. There was little agreement among the sonologists (only 2 sites with 100% agreement). Sensitivity ranged from 25% to 75%, specificity 37% to 62%, PPV 42% to 75%, and NPV 25% to 57% for each sonologist. Sonologists were more confident reporting a positive result (81% vs 51%, p<0.0001), irrespective of the actual presence of FB. The difference between detection rates of 4 types of FB did not reach statistical significance.
Current emergency ultrasound equipment utilized by ED sonologists is unreliable in detection of 3 mm FB in a human extremity soft tissue model. Future studies may further delineate accuracy rates among different sizes and materials of FB.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.