After the Crash

Despite seat belts and air bags and the other safety devices now in our cars, driving is still the riskiest thing we do every day. To make it safer, researchers are going beyond the behavior of crash dummies and back to the scene of the accident.

By Vincent Lytle|Wednesday, May 01, 1996
It was over in a matter of moments.

Lindsey Gordon, an eight-year-old girl, was riding in the backseat of her mother’s Nissan Maxima when a drunk driver drove her Jeep Grand Cherokee through a red light and into the car’s side, crushing the rear door. Maryland police found Lindsey, pale and screaming, so tightly wedged in the wreckage that it took them half an hour to pry her loose. When a helicopter finally delivered her to Children’s Hospital in Washington, D.C., doctors discovered that her stomach wall was ruptured, her spleen was lacerated, and her collarbone and several leg bones were broken.

Lindsey would eventually recover, but only after seven weeks in the hospital, many operations, and more than $300,000 in medical expenses. Even as she was being treated in the hospital, however, crash reconstructionist Mike Warner was at the scene of the accident, trying to determine why she sustained so many serious injuries. Surprisingly, he found only minor damage to the Jeep’s front end. The Maxima, however, was a different matter.

I was a little shocked to see the side impact as bad as it was, Warner recalls. He took out his tape measure and systematically went through the passenger compartment to get a precise record of the damage. Then he photographed the car and read the police report. Back at his office at Dynamic Science in Annapolis, he plugged his data into a computer program that simulates accidents based on crash profiles of various makes of car. Ultimately Warner found that the girl’s severe injuries resulted chiefly from the fact that the Gordons’ car--like most passenger sedans-- was not designed to withstand side impact from a vehicle like the Jeep, which, since it rides so relatively high off the road, missed the Maxima’s steel frame and instead hit the much weaker door, crushing it inward.

Warner is engaged in the field of trauma-based safety studies, whose modest goal is to help lower the risk of serious injury in traffic accidents. Over the years researchers have vastly reduced the fatality rate for automobile accidents by bringing us the seat belt, the air bag, and countless other safety devices. To make cars safer, however, they now need more precise information than is found in traffic accident reports, and more true-to-life data than they can get from crude crash-dummy tests. Crash dummies don’t really reproduce what happens to a passenger, says John H. Siegel, a surgeon and trauma researcher at the New Jersey Medical School in Newark. Dummies have no physiology. Nor do they act much like humans. Put crash dummies in a car and they just sit there, says epidemiologist Catherine Gotschall at Children’s Hospital. Put children in a car and they squirm, wiggle, and lie down.

To home in on the specific causes of injury and death, trauma- based crash researchers like Warner begin their work as soon as a crash victim arrives in the emergency room. They go to the cars themselves to try to piece together precisely how the victim’s injuries were sustained. Then they sit down with doctors and traffic safety experts, and sometimes biomechanical engineers and rehabilitation specialists, to arrive at a total picture of the accident and its impact on the patient. The easy answers have already been found, says Frances Bents, a research manager at Dynamic Science. The kinds of changes required now are much more complex and require a multidisciplinary approach. That’s the real beauty of trauma center research. All parties sit around the table, and each provides a clue to try to solve the puzzle.

The main impetus for trauma-based crash research has come from the National Highway Traffic Safety Administration (NHTSA). Over two decades, the agency had put together an extensive research database of traffic accidents. When in the late 1970s some safety researchers began to clamor for more detailed information from doctors about the kinds of injuries sustained in accidents, the NHTSA began funding trauma-based crash studies. One of the first was a 1988 study at the Maryland Shock Trauma Center at University Hospital in Baltimore that followed 144 patients with severe injuries from frontal and side collisions. At that time government safety standards were based mainly on tests in which a car crashed head-on into a barrier. Researchers found, however, that most real-life crashes are not straightforward frontal crashes but offset, or corner, crashes. Because only a portion of the front end absorbs the energy of such a collision, the crash can be devastating: the car crushes easily, and the instrument panel and toe pan intrude into the passenger compartment. The good news was that air bags and seat belts were found to do an adequate job of preventing head and chest injuries in corner crashes; the bad news was that they did a poor job of protecting the legs.

These early findings underscored the need for safety researchers to shift their focus from preventing death, which in the past they had concentrated on almost exclusively, to preventing injury. People who would once have died from head and chest injuries were surviving, says Bents, but they were surviving with expensive, debilitating leg injuries. Since then, injury prevention has emerged as a major theme of trauma-based research.

Further studies at the Maryland Shock Trauma Center found that the people most susceptible to foot and ankle injures were women--or more accurately, short people. To learn why, biomechanical engineers at the University of Virginia videotaped volunteers, both tall and short, in the act of braking. Shorter drivers, they saw, lift their foot to step on the brake, whereas taller drivers tend to rest their heel on the floorboard. When the researchers plugged these observations into computer crash simulations, they found that as a crash pushes the floorboard inward, it slams into the foot of shorter drivers, resulting in injury. Taller drivers avoid this fate because they rest their heel on the floorboard and ride it up during the crash. The solution, it turned out, was simple: an inch of padding on the floor of the car below the brake pedal can cut the force on shorter drivers’ ankles in half.

One of the trauma researchers’ ultimate goals, of course, is to spur auto companies to design safer cars. Partly because of the work of the Maryland researchers and others, General Motors has begun to develop dummies with more accurate legs, with more lifelike joints and more sensors. Doctors have also begun documenting ankle injuries more precisely, giving biomedical engineers better data to work with. And Mercedes Benz is pioneering the use of pedals that bend under the stress of a crash, as well as new designs that better protect passengers’ feet and that redirect the force of frontal collisions down the side rail and center of the car’s frame.

Trauma investigators have also been taking a closer look at injuries that occur in spite of--and sometimes because of--safety devices. Work at Children’s Hospital suggests that children between 40 and 60 pounds are at present particularly vulnerable in car crashes. The problem is that although these children are usually allowed by law to wear adult seat belts, the seat belts don’t fit them very well. When a child squirms, the lap belt can ride up across the stomach, making the lower spine vulnerable. Child safety would be promoted by the use of booster seats for children 40 to 60 pounds, says Catherine Gotschall, one of the principal investigators. Her group also found some children to be at risk even when they have adequate car seats because parents often use them incorrectly. As a result, the American Academy of Pediatrics, the NHTSA, and other groups are trying to educate parents about such dangers as putting a rear-facing infant seat in the front seat of a car with air bags (if the air bag inflates, the child can be struck in the back of the head or crushed against the car seat). The nhtsa recently decided to allow auto manufacturers to include a switch so that drivers can disable the passenger-side air bag when a child is sitting up front.

The more precise information available through trauma-based studies, researchers hope, should prove to be useful in the treatment of accident victims. Paradoxically, by cutting down on serious injuries, devices such as air bags have made it more difficult to diagnose those less apparent injuries that do occur. Researchers at the University of Miami School of Medicine have found, for instance, that despite air bags, many injuries occur because riders are either not wearing their seat belts or are sitting too close to the steering wheel. In the world of air bags, a patient who looks good can later turn out to have occult injuries, says surgeon Jeffrey Augenstein. As a result, paramedics, who usually decide whether a crash victim needs urgent attention on little more than a quick look, may underestimate the extent of hidden internal injuries. To help them, Augenstein and his colleagues are trying to identify which accidents typically lead to hidden injuries. For instance, if the driver is wearing an automatic shoulder belt without a lap belt, he is at increased risk of liver injury, Augenstein says.

Trauma-based studies are no substitute for epidemiological studies, in which researchers spot broad trends by gathering statistics on a great many accidents. Since trauma centers see only patients with bad injuries, their findings are not representative of all crashes. Trauma- based studies also don’t replace studies in the laboratory, which are better suited for exploring certain specific issues in the biomechanics of accidents. Field investigations of real accidents give you good injury data, says Lawrence Schneider, head of the biosciences division at the University of Michigan Transportation Research Institute in Ann Arbor. In the controlled environment of the lab, you have less injury data, but you can control impact conditions. You have to work the two together.

Trauma research has been useful, however, in identifying which problems need lab work. Studies of real accidents have shown, for instance, that crash victims who sustain seemingly minor brain injuries can suffer long-lasting learning and behavior problems. In response, scientists are using what they know about how much nerve cells can be stretched to develop computer models of the brain during impact. NHTSA scientists hope the models can eventually be used to produce improved crash dummies, with dummy heads that yield more information about the effects of a crash on the human brain.

Ultimately, some trauma researchers want to automate themselves out of a job. Augenstein, for instance, sees no reason that automobiles can’t take their own trauma histories. Racing cars already carry a device, like the black box of an airplane, that records crash information. Why couldn’t all cars provide information about the speed and direction of the crash and whether seat belts were being used at the time? That way, by the time the rescue crew arrived--after the car had automatically called 911 and relayed its location--they would already have an idea of what injuries to expect. Although it makes perfect sense, Augenstein does not expect any of this to happen soon. Change comes slowly. After all, there are still people around who refuse to buckle their seat belts.
Comment on this article