Air Bag Failures: NTHSA Data Collection Appears To Be Flawed

News stories have highlighted the apparent failure of air bags in head-on car crashes of many vehicles.

A TV station in Nashville obtained an internal National Highway Traffic Safety Administration (NTHSA) report via a Freedom of Information request that indicated 576 fatalities might have occurred from crashes where air bags failed to deploy between 2001 and 2006.

This matches information from a 2007 story from the Kansas City Star, reporting similar numbers of air bags failing to deploy.

NTHSA denied a request for an interview by the Nashville reporters, but commented that air bags work most of the time. Neither report denies this fact; the question is, why do they fail sometimes when it appears they should deploy?

How Air Bags Work

Air bags are designed to deploy when a sensor, typically located near the front of the vehicle, is triggered by an impact or deceleration.

The deceleration causes a small mechanical element to move and close an electrical circuit. The microchip sensor then triggers a small solid-fuel explosive, which fires and produces the gas that inflates the air bag. The sensors have become increasingly complex, and may be in multiple locations on the vehicle to determine the type and direction of the crash. Older air bags may have more limited triggering criteria.

During the 1990s, there was much concern over air bags deploying and injuring children or small adults in the front passenger seat. That led to the development of air bags which could sense the size and location of the passenger and the severity of the crash, modifying the deployment to meet the conditions presented by the actual crash.

FARS or NASS/CDS?

An examination of data from the Fatality Analysis Reporting System (FARS) and the National Automotive Sampling System/Crashworthiness Data System (NASS/CDS) reveals the current weakness of the data collection.

The Insurance Institute for Highway Safety (IIHS) reviewed the numbers from the Kansas City Star story and they appear to match those from the internal NHTSA report, in that they both determined 576 crashes out of 1,446 resulted in fatalities to the vehicle occupants.

The IIHS reports there are problems with the data collection and that “missing data may result in inaccurate estimates of nondeployment.” FARS appears to show 18 percent nondeployment, compared to the nine percent from the NASS/CDS numbers.

The larger problem is “the inaccuracies in FARS may stem partly from the lack of uniformity among state police crash report forms and coding practices,” according to the IIHS. States vary in reporting air bag deployment as a separate item on the police crash report forms.

The IIHS study notes that “failures of front air bags to deploy in crashes in which drivers or right-front passengers died and in which the front air bags usually would be expected to deploy appear to be relatively uncommon and far less frequent than suggested by FARS data.”

The concern they raise is over the accuracy of the coding of nondeployments in the FARS data. NHTSA should work to improve the quality of this information, to create an accurate picture of how often air bags fail.

“FARS is useful, but the police who supply the data don’t always note correctly whether an air bag deployed,” Elisa Braver, of the IIHS noted. “This is why sometimes you have to dig deeper to see what actually happened in a given crash, and NASS/CDS is a good source for a deeper dig. The Star reporters simply relied too much on FARS.”

The NASS/CDS sample has much better and more detailed information on crashes than FARS. The information is collected by trained investigators using a uniform set of data-gathering tools, compared to highway patrol and police officers from different jurisdictions, using widely differing forms and dealing with injured drivers, other emergency responders and traffic control.

Unfortunately, neither the NASS/CDS numbers nor the FARS data, show any commonality among the frontal motor vehicle accidents that would explain in general why the air bags failed.

Uncertainty Remains

The questions raised by the various reports examining FARS and NASS/CDS data potentially indicate a safety matter of concern to any driver; that of their air bag not deploying during a front-end car collision.

NHTSA should immediately improve its data collection of this information, allowing meaningful interpretations to be made of whether or not there is a problem, and then permitting an assessment of what factors, if any, are resulting in the nondeployments.