Pushing to Close Gaps in Compiling Vehicular Deaths
By Cindy Skrzycki
Tuesday, March 16, 2004; Page E01
Every year, awful things happen to children in cars that do not show up in a fatality database run by the National Highway Traffic Safety Administration. Kids die in power window accidents. They get locked into cars that heat up like ovens. They are backed over by unsuspecting adults.
Also missing from the Fatality Analysis Reporting System, or FARS, in some instances, are deaths caused when vehicles on the side of the road are hit and burst into flames.
The reason for the information gap is simple: The system was not designed to include accidents that didn't occur in traffic on the nation's highways. Thus, injuries in vehicles in parking lots, on a highway shoulder or in a driveway aren't counted in the FARS. In the regulators' lingo, these are known as non-traffic, non-crash events. To make it into the database, the vehicle must be "in transport" on a public road or highway and the death must occur within 30 days of the accident.
These distinctions matter because the FARS database is the nation's premier census for crash-related deaths. Statistics derived from it -- seat belt usage, what time of day accidents happen, and whether it was at a stop sign or a stop light -- help drive almost everything the safety agency does.
The FARS, created in 1975, depends on states gathering information about a crash, from police reports, medical records, death certificates and other sources. Information then is compiled on the vehicle, the persons involved and the nature of the accident. About 6.3 million crashes were reported to police last year. Among them were 37,000 fatal crashes and 43,000 fatalities that ended up in the federal database.
Now, consumer activists want the FARS expanded.
"I want to start a new organization called Missing in FARS which is open to people who have a deceased family member killed in a motor vehicle whose death was not recorded in FARS or which, in the case of fire, was not called a fire death," said Clarence Ditlow, director of the Center for Auto Safety.
He is particularly concerned about fires in Crown Victoria police cruisers made by Ford Motor Co.
Ditlow said his research shows gaps in FARS in incidents of Crown Victorias parked by the side of the road that have burst into flame after being hit by another vehicle. Sometimes the cause of the accident and, ultimately, death, does not indicate it involved a fire, he said. Or there is no information about the vehicle at all because it was not "in transport." The death might be in the system, he found, but it might show up in the pedestrian category. And it would not appear in a computer search for deaths in Crown Victorias.
The center said that for 2002, it counted eight deaths in seven crashes involving fires and Crown Victoria police cars; the database showed four fire crashes and five deaths. So Ditlow's group objected to the agency's decision in October 2002 to close, without taking action, an investigation of fires in Crown Victorias and other models built on the same platform.
It also filed a petition in January asking the NHTSA to reconsider a rule it had just issued to update how manufacturers must build vehicles to protect occupants from fuel tank fires.
Janette E. Fennell, founder and president of Kids and Cars, a child safety advocacy group, also has been pleading with the agency. She wants the FARS to include the deaths of children caught in power windows, backed over by vehicles, or victims of some other mishap when the car is not "in transport."
Fennell counted 154 deaths of children in various vehicle-related accidents last year that are not in the database, or are coded incorrectly. She wants the NHTSA to direct car manufacturers to use technology -- such as auto reverse on power windows -- to stop the accidents.
The safety groups say omitting these deaths deprives regulators of data they need to make regulatory and recall decisions. They also want the agency to count accident deaths that occur within a year, instead of a month.
Jeffrey W. Runge, an emergency room physician before becoming head of the safety agency, said in an interview that the NHTSA needs to have "some measurement devices" for deaths that don't occur on a public road. But the solution, he said, must be commensurate with the size of the problem. This means that the agency and safety groups are not likely to come to the same conclusion about what should be included in, or excluded from, the death tally.
Regulators acknowledge that a small number of mistakes may be made in the FARS, merely because of its size, human error and coding requirements. But officials said Fennell's figures are probably overstated and that Crown Victorias are not any more dangerous than any other car hit on the side of the road at 70 miles per hour.
Runge said the agency has been researching how significant the gaps are and how they might be closed. It has completed, but not released, a report on the prevalence of the kinds of deaths that happen off of roadways, mostly to children.
It's likely, Runge said, that the power window issue may soon be addressed. "It's ripe for regulation or voluntary action," he said. "I think this problem will resolve itself."
The agency also is looking at how to include in the FARS data from fatalities that occur when parked vehicles are struck, Runge said. They could be included in a "not-in-transport" category or a new one called "other motor vehicle." FARS analysts also will be given more training on how to correctly code fire-related deaths when they record an accident's cause in the database.
Joseph Carra, director of the NHTSA's National Center for Statistics and Analysis, said: "People on the other side have to understand these may not be the changes they are looking for."
© 2004 The Washington Post Company