Portal de Periódicos da CAPES

308 Small Retained Foreign Bodies: What Is the Limit of Detection Using Current Emergency Ultrasonography Equipment?

2012; Elsevier BV; Volume: 60; Issue: 4 Linguagem: Inglês

10.1016/j.annemergmed.2012.06.286

1097-6760

Daniel Jafari, K.J. Cody, Nova L. Panebianco, Frances S. Shofer, Bon Ku, Alexander Au, Anthony J. Dean,

Traumatic Ocular and Foreign Body Injuries

Study Objectives: Small foreign bodies may be of clinical significance if undetected in human extremities, especially if composed of organic material. Previous studies have shown a wide range of accuracies of foreign bodies detection using animal models, with high accuracy rates for foreign bodies > 10 mm and variable accuracy rates for 4 to 5 mm foreign bodies. This study sought to determine the lower limit of sonographic detection of foreign bodies using current emergency ultrasonography equipment in a soft tissue model that approximates the appearance of human extremity soft tissues. Methods: A total of 12 foreign bodies made of metal, glass, wood, and plastic (3 of each) of approximately 1 × 1 × 3 mm in size were placed at a depth of 0.5-2.0 cm in 12 randomly selected sites in 12 pork feet. 8 pork feet received puncture wounds without placement of foreign bodies (sham group). All punctures were made with the models submerged in water to minimize introduction of air into the tissue. 7 emergency physicians with at least 2 years experience in sonology (range 2-16 years), blinded to the order, number, type and depth of foreign bodies were randomly assigned to use either a Sonosite M-Turbo ® (Sonosite, Inc., Bothell, WA) or Zonare z.one ultra® (Zonare Medical Systems, Inc., Mountain View, CA) ultrasonography systems to attempt to identify the foreign bodiess. Each machine was equipped with both a “hockey stick” and a traditional high-resolution linear array transducer (Zonare L14-5w 14 - 5 MHz and L14-5sp 14 - 5 MHz; SonoSite HFL38x 13 - 6 MHz and SLAx 13 - 6 MHz). The location of the puncture wound (but not its depth or direction) and the approximate size of the foreign bodies were known to the sonologists. Every foreign bodies site was scanned by each sonologist using either the hockey stick or traditional linear array transducer in a randomized pre-assigned order. The sonologist's confidence in the diagnosis was reported using a visual analog scale for each site by each sonologist. 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. Results: 140 ultrasonography 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. Of the 20 models, only 2 resulted in 100% agreement between all sonologists, of which 1 scan was correctly identified as having no foreign bodies, and one was mistakenly identified as having a foreign body. Sensitivity ranged from 25% to 75%, specificity 37.5% to 62.5%, PPV 42% to 75%, and NPV 25% to 57.1% for individual sonologist. Sonologists were more confident when they were reporting a positive result (81% versus 51%, p<.0001), irrespective of the actual presence of foreign body. The difference between detection rates of 4 types of foreign bodies did not reach statistical significance. Conclusion: Current emergency ultrasonography equipment utilized by experienced emergency physician sonologists is unable to reliably detect 3 mm foreign bodies in a soft tissue model designed to replicate human extremities. Future studies may further delineate extinction rates among different sizes of foreign bodies, different materials, and different transducer types.