The U.S. Department of Transportation-sponsored Crash Injury Research and Engineering Network (CIREN) program offers a reasonable look at the efficacy of second-generation air bags. This paper examines the data from the William Lehman Injury Research Center (WLIRC). The WLIRC data is a near census of crashes in the Miami-Dade region with occupants that appear to be severely injured.
The percentage of deaths among trauma patients in the WLIRC data as a function of delta-V for first-generation air bags was higher than expected at lower delta-V's. There were nine driver fatalities at delta-V's of less than 20 mph (four involving short stature occupants, four with elderly occupants, and one due to significant intrusion and/or vehicle incompatibility). The data supported NHTSA's conclusion that first-generation air bags were too aggressive for occupants in close proximity to the deploying air bag and too aggressive for older persons. For second-generation air bags in the WLIRC data, there were no driver fatalities below 25 mph.
There are a number of conclusions that emerge from the trauma center data: The early deficiencies at low speed have been greatly reduced. Generally, steering columns and air bags work well together to provide high severity protection, although late deployments can still cause chest injury in moderate severity crashes. These injuries to abdominal organs may be difficult to detect at the crash scene.
The fatality rate for first- and second-generation passenger air bags also showed some interesting patterns. There were four fatalities at delta-V's of less than 20 mph with first-generation air bags (two infants in rear-facing child seats, two unbelted children younger than 3). There was one unexpected fatality at moderate severity to an out-of-position adult. An examination of WLIRC data reveals that among second-generation passenger air bags, there have been no child fatalities, no out-of-position related fatalities, and no elderly fatalities below 20 mph delta-V.