Rollover crashes account for more than 33% of all motor vehicle related fatalities and have the highest fatality risk at 1.37% in the U.S. There is increased awareness of the high fatality rate associated with this crash type, but there is very limited pediatric-specific data related to rollover crashes in the United States. Previous studies based on data almost twenty years old have revealed that nearly ten percent of all children involved in motor vehicle crashes are in rollover crashes, with the risk of fatality and injury for children in rollovers being nearly twice that of non-rollover crashes. Recent focus on rollover mitigation has resulted in implementation of countermeasures, making it important to evaluate rollover risk for child occupants with a more current data set.

Thus, to provide a contemporary analysis of rollover crashes involving young people, we queried the National Automotive Sampling System’s Crashworthiness Data System (NASS-CDS) from 1998-2011. Rollover crashes for passenger vehicles of model year 1998 or newer with at least one restrained occupant between 0 and 19 years of age were included. Occupant frequency was examined with number of quarter turns, vehicle type, vehicle specific rollover event, rollover type and direction, airbag deployment and Electronic Stability Control availability. Further, occupant age, restraint type, seating position, occupant role, and proximity to the roll direction were analyzed. Univariate and multivariate logistic regression models of MAIS 2+ and MAIS 3+ injury were built to establish the relationship between the key factors and the injury outcomes.

The study cohort consisted of 1560 occupants weighted to represent 515,470 occupants. Results indicate that children restrained in FFCRS or booster seats were less likely to sustain an MAIS 2+ injury than lap/shoulder restrained occupants in a rollover crash. The abdomen was the most commonly injured body region at the AIS 2+ level while the head was most common at the AIS 3+ level, followed by the thorax and spine (for weighted data). However, for unweighted data, the head was the most commonly injured body region followed by the spine at the AIS 2+ level while the head was most common at the AIS 3+ level, followed by the thorax and upper extremities. The variations between the weighted and unweighted distributions points out some of the challenges with conducting child-specific analyses with NASS-CDS, as some cases have extremely high sample weights. Averages of 2.8-quarter turns were associated with an MAIS 2+ injury. Because there were limited cases with rollover mitigation technologies (ESC and airbags), their protective benefits in rollover crashes could not be ascertained.