ARLINGTON, Va. --- Seat/head restraint designs in 22 current car models are rated good for rear crash protection, but those in 53 other cars are rated marginal or poor, according to new test results from the Insurance Institute for Highway Safety. The latest evaluations of occupant protection in rear-end collisions by the Insurance Institute for Highway Safety found that the seat/head restraints in more than 60 percent of car models fall short of current state-of-the-art protection from neck injury or whiplash. The ratings of good, acceptable, marginal, or poor are based on geometric measurements of head restraints and simulated crashes that together assess how well people of different sizes would be protected in a typical rear-end collision. Among the winners are seat designs in all Volvos; Audi A4, S4 and A6; Ford Five Hundred/Mercury Montego; Nissan Sentra and Versa; Saab 9-3; and Subaru Impreza and Legacy/Outback. Seat/head restraints in 12 other car models are rated acceptable. These results show some improvement since the Institute began rear-impact tests of seat/head restraints in 2004. Seats in only eight car models earned good ratings in 2004. "Even though we have more good performers, it's disappointing that so many designs are still rated marginal or poor," said Institute President Adrian Lund. "Neck injuries are common in crashes, and it's not difficult or expensive to design more protective seat/head restraints." Rear-end collisions are frequent, and neck injuries are the most common injuries reported in automobile crashes. They account for 2 million insurance claims each year, costing at least $8.5 billion. Such injuries aren't life threatening, but they can be painful and debilitating. "We're simulating what happens when a vehicle rear-ends another one in commuter traffic or at a stop light," Lund said. "People think of head restraints as head rests, but they're not. They're important safety features. You're more likely to need the protection of a good head restraint than the other safety devices in your vehicle because rear-end crashes are so common." When a vehicle is struck in the rear and driven forward, the vehicle seats accelerate occupants' torsos forward. Unsupported, an occupant's head will lag behind the forward movement of the torso. This differential motion causes the neck to bend and stretch. The higher the torso acceleration, the more sudden the motion, the higher the forces on the neck, and the more likely a neck injury is to occur. "The key to reducing whiplash injury risk is to keep the head and torso moving together," Lund explained. "To accomplish this, the geometry of a head restraint has to be adequate, and so do the stiffness characteristics of the vehicle seat. Then the seat and head restraint have to work in concert to support an occupant's neck and head, accelerating them with the torso as the vehicle is driven forward." A head restraint should extend at least as high as the center of gravity of the head of the tallest expected occupant. A restraint also should be positioned close to the back of an occupant's head so it can contact the head and support it early in a rear-end crash. If a restraint isn't positioned behind the head, it cannot support the head, but good restraint geometry by itself isn't sufficient. A seat also needs to be designed so its head restraint doesn't move backward in a rear impact because this would prevent the restraint from catching the head, the Institute pointed out. At the same time, a vehicle seat cannot be too stiff. It has to "give" so an occupant will sink into it, moving the head closer to the restraint. More manufacturers are paying attention to safer seat design. This is reflected in better ratings for some seats compared with ratings the Institute published in 2004. Seat/head restraints in the Audi A4 and S4, Honda Civic, Hyundai Sonata, Kia Optima, and Nissan Sentra improved from poor to good. Seat/head restraints in the Mercedes E class and Subaru Legacy/Outback improved from acceptable to good. Improving to acceptable are seat/head restraints in the BMW 3 series, Ford Focus, Hyundai Elantra, Lexus IS, and Mercedes C class. Many of the models with improved ratings have "active" designs that automatically move a head restraint up and toward the back of the head in a crash. Mercedes models use a spring-loaded system that activates when sensors detect a crash. Other cars have mechanisms in the seatbacks that push the head restraint up when an occupant's torso sinks into the seat during a crash. While many automakers are making improvements, a few are going in the wrong direction. Seat/head restraints in the Chrysler 300, Kia Amanti, and Nissan Altima earned marginal ratings this time compared with acceptable ratings for the earlier designs tested in 2004, the Institute said. The Institute doesn't test seats with head restraints that are rated marginal or poor for geometry. These seats automatically earn the lowest rating of poor because their head restraints cannot be positioned to protect many taller people. In this round of evaluations, the worst ratings based on geometry belong to seats in the Cadillac DTS, Pontiac Grand Prix, Suzuki Forenza and Reno, plus some seats in the BMW 5 series, Buick Lacrosse, and Mitsubishi Galant, the Institute said. "Still it's encouraging that the seat/head restraint combinations in only seven of the models we evaluated didn't make it to the dynamic test because of marginal geometry," Lund said. "When we began evaluating the geometry of head restraints in 1995, most earned the lowest rating of poor. Back then, head restraints in most cars weren't tall enough or close enough to the head to begin to provide adequate protection for people as tall as an average-size man." In response to the Institute's 1995 evaluations, manufacturers began to make changes. They designed head restraints taller and closer so they're more likely to be in position to catch people's heads in rear-end collisions. Further improvements are being driven by the federal government. A new regulation will require head restraints to extend higher and fit closer to the backs of people's heads by the 2009 model year. Automakers also have been spurred to better designs by the Institute's Top Safety Pick award. Winning vehicles have to earn good ratings in all three Institute crash tests --- front, side and rear. "Audi and Subaru redesigned their seats and head restraints specifically to earn Top Safety Pick in 2007," Lund pointed out. "Other automakers are working on this aspect of crashworthiness specifically to earn the award." Recognizing improvements in head restraint geometry and the need to move beyond ratings based solely on geometry, Institute researchers joined with other whiplash injury prevention experts in late 2000 to organize the International Insurance Whiplash Prevention Group (IIWPG). IIWPG conducted extensive research and testing to develop the procedures for the dynamic tests and evaluation criteria used by member research groups, including the Institute, to rate the performance of seat/head restraint combinations in vehicles sold in a number of world markets. Ratings also are being released by IIWPG members in Canada, the United Kingdom, Europe, and Australia. Seat/head restraint ratings are based on a two-step evaluation. In the first step, restraint geometry is rated using measurements of height and distance from the back of the head of a mannequin that represents an average-size man. Seats with good or acceptable geometric ratings are subjected to a dynamic test conducted on a crash simulation sled. The sled test replicates the forces in a stationary vehicle that's rear-ended by another vehicle of the same weight going 20 mph, which accelerates the struck vehicle to 10 mph. The sled is a movable steel platform that runs on fixed rails and can be programmed to re-create the accelerations that occur inside vehicles during real-world crashes. A dummy specially designed to assess rear-end crash protection, BioRID, is used to measure the forces on the neck during the simulated crashes. Researchers also measure how hard the seatback pushes on the dummy's back and how quickly the head restraint supports the head. The Institute's dynamic ratings of good, acceptable, marginal, or poor are derived from two seat design parameters (peak acceleration of the dummy's torso and time from impact initiation to head restraint contact with the dummy's head) plus neck tension and shear forces recorded on BioRID during the test. The sooner a restraint contacts the dummy's head and the lower the acceleration of the torso and the forces on BioRID's neck, the better the dynamic rating. A seat/head restraint's dynamic evaluation is combined with its geometric evaluation to produce an overall rating.