A curtain or side air bag is designed to protect an occupant’s head, torso, and/or pelvis in a side impact, specifically a “near-side” impact to the side of the vehicle adjacent to where the occupant is seated. While NHTSA has never explicitly required installation of curtain or side air bags, the agency encouraged it by listing, since 1996, the makes and models of vehicles that offer them in its Buying a Safer Car brochures and at www.safercar.gov; by expanding, in 1996, its NCAP consumer-information program to include star ratings for side impacts; and by upgrading, in 2007, FMVSS No. 214, “Side impact protection,” adding an oblique 20 mph side impact test with a pole, with phase-in scheduled for MY 2011 to 2015. The agency anticipated that head-protection air bags would generally be installed to meet the new requirement. Several types of side air bags have been offered in the United States, including torso bags that deploy from the seat or door (first sold on 1996 Volvos), head curtains that deploy down from the roof-rail area (first sold as an inflatable tubular structure in MY 1998 BMWs), and combination bags that deploy outward from the seat and then quickly upward to provide torso as well as head protection. However, by 2006, the clear preference was for separate curtain and torso bags, the configuration that covers the largest area and intuitively appears to provide the most protection. In model year 2011, 85 percent of new cars and LTVs were equipped with curtains plus torso bags for drivers and right-front passengers. Meanwhile, some vehicles added rollover sensors that make it possible to also deploy the curtains in rollover crashes as well as side impacts (first sold on 2002 Ford Explorers and Mercury Mountaineers). In addition, recent curtains may cover a larger area and stay inflated longer to protect occupants in crashes with multiple impacts and to help prevent occupants’ ejection from the vehicle in rollovers and other crashes. NHTSA anticipates curtains that deploy in rollovers will generally be used to meet FMVSS No. 226, “Ejection mitigation,” which is scheduled to phase in from MY 2014 to 2017. Installations are catching up with the other types of side air bags; in model year 2011, 45 percent of new cars and LTVs were equipped with curtains that deploy in rollovers. NHTSA issued a preliminary evaluation of side air bags in 2007, based on crash data through CY 2005 — before the widespread availability of separate curtain and torso bags. The analyses showed significant fatality reductions in near-side impacts for air bags offering head protection, but to do that the analyses had to merge curtain-plus-torso and combination bags into a single group with a single effectiveness estimate. NHTSA would now like to estimate the fatality reduction in near-side impacts specifically for curtain plus torso bags, as manufacturers are increasingly relying on this technology. Furthermore, the 2007 report had preliminary analyses, based on limited data, showing that curtain bags may be rather effective even for far-side occupants (e.g., the driver in a right-side impact); those analyses need to be revisited, as it is not intuitively clear why the technology would have a substantial effect in far-side impacts. The 2007 report did not study curtains that deploy in rollover crashes. NHTSA’s Fatality Analysis Reporting System through 2011 now has enough crash data to estimate individually the fatality-reducing effectiveness of each of the major types of curtain or side air bags in near-side impacts — including over 10 times as much data on vehicles with curtains plus torso bags as in the 2007 report. A logistic regression, comprising 73,228 FARS cases, found statistically significant fatality reductions for each type of curtain or side air bags, for drivers and RF passengers. Separate curtain and torso bags show the highest effectiveness in near-side impacts: 31-percent fatality reduction, with relatively narrow confidence bounds from 25 to 37 percent. Supplementary contingency-table analyses of FARS and estimates of national fatality rates per 1,000 occupants involved in near-side impacts, based on FARS and NASS-GES data, confirmed the logistic regression’s results for curtain plus torso bags. Effectiveness is approximately the same in cars and in LTVs, and for drivers and RF passengers. Analyses for far-side impacts did not show corresponding, large benefits for curtain plus torso bags. Although some of the regressions show positive, but relatively small effects, NHTSA does not believe there is, on the whole, enough evidence to quantify a specific fatality reduction in far-side impacts at this point. The analyses of far-side impacts should perhaps be repeated in about 3 or 4 years, when there will be considerably more data available. FARS now also has enough data for initial statistical analyses of curtains that deploy in rollover crashes. Although the data is still limited, it shows these curtains save lives in first-event rollovers. The estimated fatality reduction is a statistically significant 41.3 percent, with confidence bounds ranging from 22.5 to 55.5 percent. These curtains help prevent occupant ejection and also mitigate injuries from contact with components and surfaces inside the vehicle: they are about equally effective in reducing ejection and non-ejection fatalities in first-event rollovers. With the existing data, NHTSA did not see a fatality reduction in subsequent-event rollovers and also could not determine if the recent curtains that deploy in rollovers are more beneficial than earlier curtain designs in near-side or far-side impacts. These analyses, too, should be repeated in about 3 or 4 years. (Author/publisher)
Abstract