Twenty-nine human cadavers were subjected to frontal impacts using two restraint conditions: torso belt-only (14 tests), and torso belt-and-airbag (15 tests). Five radiologists independently evaluated post-test chest films, and their results were compared with detailed necropsy results. The radiologists' ability to identify rib fractures was influenced by the restraint system, rib number; aspect (right or left), circumferential location of the fracture, and the number of fractures on the rib. The probability of fracture detection (sensitivity) was significantly higher for the torso belt-only group (0.44) than for the belt-and-airbag group (0.24). Knowledge of fracture locations increased the overall (both restraint groups) sensitivity from 0.40 to 0.49. It is concluded that the sensitivity is dependent on the restraint system used. Since rib fractures are the injuries most commonly used to describe AIS 2+ thoracic injury severity, this dependence may bias field data-based comparative studies of restraint system effectiveness. (Author/publisher).
Abstract