Femur fractures in relatively low speed frontal crashes: the possible role of muscle forces.

In a sample of relatively low speed frontal collisions (mean collision speed change of 40.7 kph) the only major injury suffered by the partly or fully restrained occupant was a femur fracture. However, femur load measurements from standardized barrier crash tests for similar vehicles at a greater speed change (mean of 56.3 kph) showed that in almost all the cases, the occupant's femur would not have fractured because the loads were below fracture threshold. In order to address this discrepancy, the load in the femurs of the occupants in the crash sample were estimated and compared with the femur fracture threshold. Femur load was estimated by inspecting the scene and measuring deformations in each vehicle, defining occupant points of contact and interior surface intrusion, and calculating crash change in velocity and deceleration. From this data, the measured femoral loads from standardized crash test data in a comparable vehicle were scaled to the actual crash by considering crash deceleration, occupant weight, and restraint use. All the occupants (7 males, average age 26.7 years, 13 females, average age 36 years) sustained at least a transverse midshaft fracture of the femur with comminution, which is characteristic of axial compressive impact, causing bending and impaction of the femur. However, the estimated average maximum axial load was 8187 N (S.D. = 4343N), and the average probability for fracture was only 19% (based on the femur fracture risk criteria). In 13 crashes the fracture probability was less than 10%. Two factors were considered to explain the discrepancy. The occupant's femur was out of position (typically the driver's right front leg on the brake) and did not impact the knee bolster, instead hitting stiffer regions of the dashboard. Also, since most victims were drivers with their foot on the brake to avoid the collision, additional compressive force on the femur probably resulted from muscle contraction due to bracing for impact. Adding the estimated muscle load on the femur to the estimated external load increased the femur loads beyond threshold, explaining the fracture in all but one case. Since crash tests using dummies cannot simulate out of position occupants or muscle contraction loading, they may underestimate the total load acting on the femur during actual impacts where the driver is bracing for the crash. These results may have implications for altering knee bolster design to accommodate out of position occupants and the additional load caused by muscle forces during bracing.