Assessment of a human cadaver model for training emergency medicine residents in the ultrasound diagnosis of pneumothorax.

OBJECTIVES: To assess a human cadaver model for training emergency medicine residents in the ultrasound diagnosis of pneumothorax.

METHODS: Single-blinded observational study using a human cadaveric model at an academic medical center. Three lightly embalmed cadavers were used to create three "normal lungs" and three lungs modeling a "pneumothorax." The residents were blinded to the side and number of pneumothoraces, as well as to each other's findings. Each resident performed an ultrasound examination on all six lung models during ventilation of cadavers. They were evaluated on their ability to identify the presence or absence of the sliding-lung sign and seashore sign.

RESULTS: A total of 84 ultrasound examinations (42-"normal lung," 42-"pneumothorax") were performed. A sliding-lung sign was accurately identified in 39 scans, and the seashore sign was accurately identified in 34 scans. The sensitivity and specificity for the sliding-lung sign were 93% (95% CI, 85-100%) and 90% (95% CI, 81-99%), respectively. The sensitivity and specificity for the seashore sign were 80% (95% CI, 68-92%) and 83% (95% CI, 72-94%), respectively.

CONCLUSIONS: Lightly embalmed human cadavers may provide an excellent model for mimicking the sonographic appearance of pneumothorax.