NHTSA Proposes Side Impact Protection for Children

With just six months to go before a Final Rule is due to be published, the National Highway Traffic Safety Administration has released its proposal to add side impact testing to the child seat safety standard for children up to 40 pounds.

To the uninformed, the NPRM reads like the history of an agency moving forward expeditiously toward a test that will lead the way for global child safety standards. And it is true that the proposed sled test adds a new deformable door component. But a more complete dive into the history of side-impact crashes and protection for child shows that this innovation should have been an amendment to an existing side impact test for child safety seats. This proposal is waaaaay late, and the result of not one—but two—Congressional mandates, about a dozen years apart.

Here’s the truth: (And you can measure it by policy, by regulation, by research, by public engagement – it all comes out the same.) Protecting children in crashes has never been a priority for the agency, nor for the automakers. That’s how you get the first side-impact compliance test for child safety seats in 2014.

Nonetheless, some child seat safety experts are applauding the effort.

“I’m very excited that this is finally being implemented,” said Gary Whitman, Vice President, Research and Development at ARCCA Inc, and an expert in occupant crash safety systems who has tested hundreds of child seats and collaborated with NHTSA, the Pennsylvania chapter of the American Academy of Pediatric Child Injury Prevention, National SAFE KIDS Campaign, and the Children’s Hospital of Pennsylvania in child restraint research. “It’s long overdue and it’s a shame that it required an act of Congress.”

At the same time, Whitman and biomechanical expert Salena Zellers Schmidtke agree that the proposal has some serious deficits – it only protects children up to 40 pounds, and makes questionable assumptions about the efficacy of side air curtains for children in booster seats in a side impact crash.

Side Impact Hazard Well Known

The agency, researchers and industry have long known that side impact crashes are a significant safety hazard. NHTSA has periodically assessed the magnitude of side impact crashes as it considered amendments to the side-impact standard, and continue to find that it is the second most frequent crash mode, accounting for slightly under a third of the total.

The dangers to rear-seated children have similarly been established by the research. In 1989, researchers from the University of California at Irvine assessed injury patterns and severity among children 4-9 years old, by seat location and crash type and found that a “marked proportion of serious injuries sustained by restrained children in onsite lateral impacts; 41% of these children had injuries with an MAIS of 2 or greater,” primarily caused by deformation of the door. In 2002, the agency examined FARS data from 1999 and found 1,317 fatalities of children aged 12 and under, with nearly 32 percent in side-impact crashes; 91 of those children were restrained in a child safety seat. In 1997, researchers from the Insurance Institute on Highway Safety, using Fatality Analysis Reporting System data from 1988-1995,found that “32 percent of child deaths under 12 occurred in side impact crashes, the second largest group of fatalities.

The Final Rule is expected to save more than 18 children annually and cost child seat manufacturers about 50 cents worth of improvements – in wider side wings and better padding, for a total of $3.7 million.

The Proposal

The proposed sled test simulates a near-side side impact of a small passenger car, including the velocity of the striking vehicle, the struck vehicle, and an intruding door. The proposed test buck consists of a sliding vehicle seat mounted to a rail system, with a side door structure rigidly mounted to the sled buck structure. The child safety seat would be attached using LATCH, belt positioning boosters would be tested with a lap and shoulder belt. Child seats for children from 22-40 pounds would be tested with the Q3, a three-year-old side-impact dummy developed in Europe; infants would be represented by the 12-month-old CRABI dummy.

The proposed injury criteria for the Q3 would be chest deflection and Head Injury Criterion (HIC).  The CRABI would be used to measure the ability of the seat to contain the dummy’s head from contacting the intruding door.

The child safety seat would be required to maintain structural integrity, defined in the NPRM as “there could not be any complete separation of any load-bearing structural element of the CRS or any partial separation exposing surfaces with sharp edges that may contact an occupant. These requirements would reduce the likelihood that a child using the CRS would be injured by the collapse or disintegration of the system in a side crash or by contact with the interior of the passenger compartment or with components of the CRS.”

The structural integrity component could protect victims of poor child safety seat design like the Guadaloupe Pardo, a six-month-old infant who allegedly died in a low-speed fender-bender. In August 2000, Pardo was secured in a newly purchased Evenflo Discovery infant carrier that separated from its base in a crash. According to court documents, Victoriano Pardo was at the wheel of a 1999 Ford Escort, southbound and creeping across an intersection to make a left-hand turn, when a motorist trying to avoid a back-up by travelling northward in the bus lane, struck the Pardo Escort in the rear passenger door. Witnesses and Evenflo’s accident reconstruction expert agreed that both vehicles were traveling at very low speeds. No one in either vehicle was seriously injured and only the damage was to the exterior of the Escort’s rear passenger door.

But Guadaloupe Pardo occupied the rear right position. And the force of the crash allegedly caused the Discovery infant seat to detach from the base, forcing it upwards and flipping the seat over on to the chest of her mother, who was seated next to her daughter in the middle rear position. Guadaloupe Pardo remained strapped in her seat. She died the next day of massive head injuries, court documents said.  

The agency also requested comments on the merits of adding side-impact requirements to CRS models designed for children over 40 lbs., the oft-forgotten “Forgotten Child,” ages 4-8, and comments on whether the designs of belt-positioning booster seats might get in the way meeting these requirements. The agency did not include larger children in this proposal because, it reasoned, larger children would be protected by the side air curtains mandated, in essence, by the new FMVSS 214 Side Impact requirements.

Schmidtke, a biomechanical engineer with Safety Research & Strategies and the editor of Pediatric Injury Biomechanics: Archive & Textbook, challenged the agency’s decision to protect children only up until 40 pounds.

“It’s good, but it doesn’t go far enough,” she says. “The agency’s rationale – that the air curtains will protect the child in a belt-positioning booster, isn’t necessarily true because the large range of child sizes above 40 lbs. results in children positioned in many different places relative to the vehicle interior and different booster seat designs interact with the vehicle interior differently.”

She also questioned why only the middle of the weight range will be tested. Just like in the frontal impact test, the proposed side impact standard does not test the lower and upper weight limits – it uses a 32 lb. dummy for child restraints recommended from 22-40 lb. children.  Weight differences can affect the structural integrity of the seat in crashes,” Schmidtke says.

Weight ranges in child seat compliance testing has been a perennial issue in rulemaking, going back to 1974.  Seats recommended for children whose weight is in a wider range than can be met by a single dummy would be required to meet the standard with each dummy within the range of recommended weight.

Whitman also raised concerns about NHTSA’s rosy view that air curtains will protect older children.

“Not all vehicles have side curtains, and I’m not convinced that the six-year-old is going to be protected by the side curtain. Many of them are relatively short. If NHTSA wants to rely on that they need rulemaking on side curtains, because they’re relying on something that is not required and not necessarily designed to protect that population.”

Whitman said that the agency should also consider testing in the center position, which it has consistently touted as the safest position for children. In a nearside crash, the kinematics of a child safety seat in the center position are markedly different than those in the nearside outboard position.

“It’s a completely different environment,” Whitman said. “When the seat is on the nearside outboard position, the seat loads up against the door so side wings are prevented from moving. In the center the position, the child seat yaws and pitches toward impact and wings move out of the way of the head. And, in a small car with some intrusion that will be within range of a head impact. The head is free to swing toward the impact.”

Whitman said he was taken aback by the proposed effective date three years from the Final Rule.  

“What could possibly require three years for this to be enacted? All they have to do is fabricate the sled and learn how to conduct the test. Truly, that should take several weeks. Let’s give them a whole year, but not three years.”

Hold the Champagne

NHTSA gets a few points for introducing the first sled test to simulate both the acceleration of the struck vehicle and the vehicle door crushing toward the car seat. Unfortunately, it gets many more timeliness demerits.

The regulation is a requirement of the 2012 Moving Ahead for Progress in the 21st Century (MAP-21) Act, which mandated that a Final Rule be promulgated by July 2014. And NHTSA took pains here to reach over its institutional shoulder and give itself a little pat on the back:

“We interpret this provision of MAP-21 as providing us a fair amount of discretion,” the NPRM said. “NHTSA informed Congress in 2004 that enhanced side impact protection for children in child restraints was a priority for NHTSA.”

But, NHTSA first proposed side impact testing of child restraints in 1974. Yes, 40 years ago. The compliance test would have consisted of a 90 degree lateral impact at 20 mph, with a requirement that the child seat retain the dummy, that the dummy’s head motion would be limited to 19 inches in either lateral direction, and that the seat itself remain intact. But, as it turned out, virtually none of the child safety seats at the time could pass it. Only a seat with a tether made the cut and as the agency says “This was of concern because tethers were widely unused at that time.” Why was that? The NPRM neglects to mention that more than 70 percent of the child restraints on the market used tethers around that time.  But the tether usage rate was very low – mainly because automakers didn’t provide – nor were they required to provide anchorages to secure the tethers.

Adults have had a side-impact-crash-protection standard since 1996. That’s 18 years ago.

As for all the MAP-21-recognition-of-NHTSA’s-previous-work-yadda-yadda, in 2000 there was another law requiring NHTSA to develop a side-impact regulation for child safety seats. You may remember it: the Transportation Recall Enhancement, Accountability and Documentation (TREAD) Act of 2000. In 2002, NHTSA published an ANPRM, but gave up by 2003, saying it hadn’t the time to do the research to make a credible proposal. Exactly – because it had done virtually nothing in the previous three decades. That was 11 years ago.

In many respects, that ANRPM showed just how little had been accomplished on the issue of side impact and child safety seats. The agency said that it could not even propose a rule regarding side-impact protection for child safety seats, because there were too many uncertainties: “the determination of child injury mechanisms in side impacts, and crash characteristics associated with serious and fatal injuries to children in child restraints; development of test procedures, a suitable test dummy and appropriate injury criteria; and identification of cost beneficial countermeasures.” The agency said that it had analyzed crash data, developed a dynamic side impact test and identified some countermeasures, but had not had time to determine their feasibility and benefit.  Under a tight statutory deadline, the agency said that it was unlikely to resolve these issues in time.

In the meantime, we have proud moments in agency history like the Evenflo Discovery debacle. In January 2007, a controversial Consumer’s Report story claimed that sled-tests showed that some models of the infant carrier combination car seat had a tendency to separate from its base under the stress of crash forces. Discovery deaths in side-impact base separation crashes had already been the target of a 2004 NHTSA investigation that closed four months later with no defect finding. In the Preliminary Evaluation, the agency noted that Discovery infant seats had separated in 45 multi-vehicle and 11 single vehicle crashes: 29 involved side impact.

NHTSA and Evenflo torpedoed the Consumer Reports story by pointing out that its sled tests were conducted at a much higher rate of speed than the story claimed.  Consumers Union, the publisher of Consumer Reports printed a retraction, withdrew the story and apologized. One year later, NHTSA and Evenflo announced that 1.1 million Discovery infant carriers were being recalled because testing by both parties showed that it could be knocked off its base in a side impact.

NHTSA discovered that CU was right all along after conducting its own look-see by placing virtually the same population of child safety seats examined in the CU tests into their Side Impact New Car Assessment Program – (SINCAP).  (see The Safety Record Special Report: How Consumer’s Union Shocking Child Seat Tests Forced the Recall of the Evenflo Discovery)

A compliance sled test is better than nothing, but they can be gamed. Just look at the frontal crash test for child safety seats. One of the injury criteria is HIC, but the sled buck has no seat back for the child’s head to interact with.

A dynamic vehicle crash test gets you a lot closer to real-world performance. Euro NCAP and Australian NCAP have tested child restraints dynamically in vehicles before 1999 in both frontal and side impacts. That’s 15 years ago. Rear seat restraint systems need to be dynamically tested in vehicles with restrained child dummies and dummies in child restraints and assessed using the injury criteria and excursion limitations from FMVSS 213 and 208.

To be sure, neither NHTSA nor IIHS will be able to test every child seat in every vehicle – there are hundreds of possible combinations. But they could select the most popular models and focus there. In 2009, our dearly departed DOT Secretary Ray LaHood announced that he was ordering NHTSA to embark on a program to match child seats and vehicle rear seats. Then it was going to be a voluntary program for manufacturers. That bandwagon never made it out of the driveway.

NHTSA was testing child safety seats in NCAP and SINCAP, but have since stopped, for reasons unknown. This would have made for a much better standard.