Has the time come to reverse our stance on extended rear-facing and turn children forward-facing at age 1 like we used to in the olden days?
The simple answer, for the moment is, no.
UPDATED JULY, 2018
What’s going on?
Dorel Juvenile Group, the parent company of Safety 1st, Maxi-Cosi, Cosco and other juvenile brands recently issued an updated position statement on their website explaining why they reversed their position on a 2-year age minimum mandate for forward-facing in their convertible carseats. The short story is that they hired a statistician, Jeya Padmanaban, to replicate the original 2007 Henary study upon which all our assumptions of rear-facing (RF) safety statistics in the USA are based. Not only was Padmanaban unable to replicate the results using the same data set as the original authors of the study, her conclusions actually led to opposite findings. She presented her findings to NHTSA and to the journal Injury Prevention. This prompted some of the original authors of the 2007 study to re-examine their analyses. When their attempts to replicate the analysis also fell short, it became apparent that there were real flaws in the study. In August 2017, the journal Injury Prevention issued an “Expression of Concern” regarding the original study. From the statement: “Specifically, they believe that survey weights were improperly handled in the initial analysis, which caused the apparent sample size to be larger than the actual sample size. This resulted in inflated statistical significance.” UPDATE — In February 2018, Injury Prevention published an official retraction of the original study. “Because of serious concerns regarding the magnitude, significance and replicability of the findings reported in this paper, the journal made the decision to retract it.”
What do we know at this point?
The anatomy of the developing pediatric cervical spine predisposes children to injury of the upper cervical spine. In general, the younger the child, the more likely an upper cervical spine injury will occur. The neural arches in the pediatric cervical spine fuse posteriorly by 2–3 years of age. Until that time, the vertebrae are made of cartilage and bone and held in place by ligaments; it’s all very pliable and elastic. Traveling in the rear-facing position is inherently safe and is critical for babies less than 1-year-old. (Please also read Why Rear-Facing Is Better: Your RF Link Guide, an evidence-based justification for rear-facing.)
Even though the statistics from the original 2007 study have been proven to be inaccurate, there is a consensus that rear-facing carseats cradle the head, neck, and spine to protect them in frontal and side-impact crashes. We know it’s safe from basic physics, an understanding of crash dynamics and from results from other countries, notably Sweden.
What’s changed?
Since 2007 when the Henary, Sherwood, Crandall, et. al. study was first published, child passenger safety advocates have been told that rear-facing is 500% (or 5 times) safer than forward-facing for children under age 2. Now we know that statistic isn’t true, at least not based on the data used in this one study which analyzed injuries to fewer than 300 kids between 1988-2003. Having such a small sample size makes drawing broad conclusions very difficult. Large sample sizes generally result in more accurate and reliable conclusions. We have always had our own concerns about the original study and how the “5x safer” figure is presented to parents. Now we know that we were right to be concerned. Unfortunately, we still don’t know exactly how rear-facing compares quantitatively to forward-facing in most situations.
There are other methods, but it can also be difficult to draw broad conclusions from specific case studies or proprietary crash testing done by manufacturers. All of this underscores the need for a more modern crash test sled and better studies on the subject. Modern vehicles simply don’t have a back seat that’s a flat bench seat of a ’70s Chevy Impala with lap-only seat belts and no floor like the standard crash test bench does. Modern vehicles have very different back seat cushions, front seats that crowd the back seat, lap/shoulder seatbelts, and they all have floors too!
What are the risks to a rear-facing child?
We know that legs and feet are NOT at high risk for injuries in frontal crashes (which is actually what many parents and caregivers worry about). We also know that rear-facing carseats do an excellent job protecting the vulnerable head, neck and spine. However, while rare, moderate and serious head injuries can happen to babies and children in rear-facing carseats. There are typically 3 possible causes.
- The top of a child’s head might move out of the protection of the carseat shell and strike a front vehicle seat, pillar or console. This may be more of a concern with rear-facing only infant carriers, especially when fully reclined with a baby near the height limit.
- The carseat shell strikes part of the vehicle interior, possibly resulting in the back or side of baby’s head hitting the inside of the carseat with enough force to cause a contusion or other injury.
- In unusually energetic frontal crashes, the carseat could rebound with enough energy that the front of baby’s head may strike part of the vehicle interior.
FAQs:
Does this mean there is no benefit to rear-facing beyond 12 months?
No, the implications of improper weighting in this study have resulted in “inflated statistical significance.” That doesn’t mean there is no safety benefit to children 12 – 23 months riding rear-facing in a crash. But it does means that the “Rear-Facing is 5x Safer” statistic is not accurate. How much safer is RF when compared with FF using the limited data in this study? At this point, we still don’t know and the limited sample sizes may not support a compelling conclusion.
We also don’t know how this relates to side-impact safety. Dorel’s internal testing indicated potential issues with head injury criterion (HIC) scores for RF convertibles when they conducted frontal crash tests using a fixture that simulated contact with the front seat in a vehicle but they didn’t mention anything about the difference between RF and FF in side-impact testing, something they have focused on for the last several years. Side impacts, which are less common than frontal impacts, cause more serious injuries because of intrusion and the proximity to the occupants inside the vehicle.
What about Sweden? They have an excellent track record with rear-facing seats.
Sweden does indeed have very low rates of injury and death because they typically RF their children until age 3-4. However, Sweden has a very different educational awareness and driving environment compared to the U.S. as well as different vehicles and carseats. Their RF carseats are installed very differently (sometimes in the front passenger seat with the airbag disabled), they drive fewer miles on less crowded roadways with fewer impaired drivers. Their vehicle fleet lacks the sheer numbers of large pickups and SUVs we have in the USA. Also, Swedish seats do not pass federal safety standards in North America because of how different they are and how they are meant to be installed (with load legs, multiple tethers and bracing). It’s just not a fair comparison by any stretch. Most importantly, they don’t use forward-facing carseats with a 5-point harness system, so they have no RF vs FF comparative data from their own crashes.
How could the flaws in this study have gone unnoticed for so long? Wasn’t the Henary et al study peer-reviewed before it was published in 2007?
Good question. We were wondering the same thing about the diligence of those conducting peer review.
As for the flaws, various advocates expressed concerns as early as 2008. For example, some surprising results regarding forward-facing infants drew immediate concerns over sample sizes. Other concerns have been discussed at conferences and online for many years, but we are happy this topic is finally receiving renewed interest from researchers.
Where is the data? Where are all the studies?
There aren’t any. Other than looking at the original study again, no one is doing any new studies on this important topic, at least not as far as we know. If you look at the sample size of the original Henary study, the sample size is quite small which makes it difficult to pull an accurate conclusion. Nevertheless, we as an industry clung to it because it showed such a significant difference and it made logical sense based on crash physics and experience in other countries.
What are the takeaways, is rear-facing still safer than forward-facing?
There is no evidence that the American Academy of Pediatrics’ (AAP) current recommendation, to rear-face until at least 24 months, is harmful. RF to age 2 isn’t an unsafe idea by any stretch, even if there was an error in the way the initial analysis was handled. We simply don’t have conclusive numbers to back up when to turn forward-facing or to tell people how much safer RF may be for a child of any particular age. The good news here is that if installed and used correctly according to instructions, both RF and FF child restraints provide a great degree of protection. The risks, either way, may be much lower than any of us realized.
According to Dorel’s statement:
It is important to note that both the Henary study and the Padmanaban study demonstrate that car seats are very effective at preventing injury for children. In fact, the differences in safety for rear-facing or forward-facing children between 12 and 24 months are slight. For example, in the Padmanaban study 0.4% suffered serious injury rear-facing and 0.2% suffered serious injury forward-facing. Using the same group of children in the same study, that means fewer than 1% of children were seriously injured when restrained either forward-facing or rear-facing. This perspective is very important to understand when discussing the subject of direction with parents of a child between 12 and 24 months. As often indicated by car seat technicians, the most important factors of car seat safety is the right car seat is selected for the child, positioned and installed correctly in the vehicle, and that the child is harnessed right every time. The NHTSA curriculum and Car Seat Technicians do an excellent job focusing on those factors so that families correctly use car seats. Dorel proudly supports their efforts.
Now, with all that said, if you have a toddler or older child who is rear-facing and you want to keep them that way but now you find yourself worried about the potential for head injuries in RF seats, here are some specific suggestions:
- Install your carseat correctly following manufacturer’s instructions so that it moves less than 1″ at the belt path and properly snug the harness so it passes the pinch test.
- Follow your carseat manufacturer’s guidelines on height and weight requirements for RF.
- Switch to a rear-facing convertible carseat before your child actually outgrows the infant seat by height.
- If allowed, install your carseat more upright when your older baby or toddler can sit up without support, has good head and neck control and can safely tolerate riding in a more upright angle.
- If your rear-facing carseat offers a load leg or anti-rebound feature, use it.
- Choose a carseat that has a layer of energy-absorbing foam (EPP or EPS, typically) both behind the head and on the sides of the shell near the head. Deep side wings can also be advantageous.
Conclusion:
The ultimate takeaway for both parents and advocates is and has always been the same:
Any current FMVSS213 certified child safety restraint used according to instructions and relevant state laws (some car seats and some state laws don’t allow FF before age 2) is very effective at reducing the risk of severe and fatal injuries. Simply using an appropriate car seat correctly in the back of a newer vehicle and then driving unimpaired and undistracted will reduce any risk to very low numbers.
Here at CarseatBlog, we will continue to support the current AAP recommendations which encourage rear-facing until at least 24 months. If at some point in the future we feel the need to revise our recommendations based on new data, we won’t hesitate to do so. Being open-minded and receptive to new information can be hard sometimes, especially if that new information questions or disputes long-held beliefs. But we owe it to our kids and kids everywhere to approach whatever comes of this with a scientific mind and not with a purely emotional response.
All my children were very large for their age. At 2 years old they were about the size of a normal 4 year old. Now I hear talk in parenting forums that you should keep children rear-facing until age 4, because of bone development. My children would have been about the size of 6 year olds at four, and I can’t imagine having any kind of car seat they would reasonably be able to sit in that was rear facing by then.
What is your opinion of the “rear-facing until 4” advice? What about unusually large children? Has there been any studies about being in a rear facing seat if a child is especially large (both for crashes, but also I’m wondering about the effects of everyday riding in a seat that doesn’t fit on other things, like posture/proper bone development).
Hi Gale. It’s an important distinction that you hear about RF to age 4 in online parenting forums. It’s only from online where you’ll hear that recommendation. We simply don’t have any data that tells us RF to age 4 is any better or safer than RF to age 2 or even to age 1. It also doesn’t mean it’s less safe; it means we don’t know. NHTSA guidelines show RF to age 3 if possible, while the AAP recommends RF to the limits of a child’s carseat (assuming a convertible), which gets most kids to age 4. This AAP age recommendation is new as of 8/30/18, though they’ve always recommended RF to the limits of the carseat. Some carseats now RF to 50 lbs.; my daughter didn’t hit 50 lbs. until 4th grade, lol.
While we don’t have studies telling us which direction is beneficial to toddlers, one thing we do know is that regardless of size, you can’t speed up how fast the spinal column ossifies. Age dictates that, not size.
Hi Gale,
Bone and connective tissue development are often a red herring when used in regard to extended rear-facing discussions. Yes, horrific injuries to the neck and spine can and do occur to children because their bodies are not as developed as adults, but injuries like internal decapitation are quite rare. In nearly every study, including the major study on extended rear-facing, head injuries are by far the greatest risk for severe injury to children in motor vehicle crashes. That is the main reason we believe rear-facing to be safer. It’s also the same reason why an older, taller baby or toddler near the height limit of their rear-facing seat can also be at risk. Rear-facing to age 4 or higher is probably the safer way to transport a child, but only if they are properly restrained according to the owners manual, with their head well contained within the limits of the child restraint. As Heather said, we just don’t have statistically significant real-world data to prove it.
Hi Carrie,
We did not mean to imply that Swedish carseats are not meeting standards in Sweden or any European country. Our apologies. I have made a minor change in the wording to clarify this.
Hi Sarah. I updated the comment previously. Sometimes, edits can take longer to appear for some than others, due to the way internet service providers and browsers cache our articles. I have just cleared our local server cache and that may help also.
Thank you very much for the clarification, Juune! Of course, we didn’t mean to imply that Swedish seats were any less safe because they have not been tested or approved for the USA market. In fact, in some regards, European standards are more strict than those in the USA, but mainly just different. These differences are why it’s impossible to directly compare statistics from Sweden to the USA. For example, one could speculate that if something like a Britax Frontier ClickTight was available and used in Sweden, statistics on injury and death rates would also be extremely low for forward-facing children in a 5-point harness, due to all the other fundamental differences. Really, it’s a sad commentary on antiquated USA federal regulations and standards that we do not have access to some of the rear-facing carseats that are available in Sweden!
My post was only to clarify the sentence I quoted, because the way I read it it sounded like “Swedish seats” are certified to some separate standard of their own, which they are not 😉 [Not a native English speaker here, so maybe I misread.]
Yes, I’ve always wondered why R44 [and R129] does yet the U.S. car seat standard apparently doesn’t mandate means to manage downward rotation on RF seats. All current Euro RF seats either mandate bracing, or have have a load leg, or have the diagonal part of the belt go around the back of the car seat [what is often called “Euro-routing” here, super inconvenient on convertibles in practice if you ask me 😛 ], or have an Australian style tether.
Thank you again! We hope someday the USA standards will be updated to allow us to use seats and features that countries like Sweden have had for over a decade.
“Also, Swedish seats do not pass federal safety standards in the U.S., Canada or even in Europe because of how different they are and how they are meant to be installed (with load legs, multiple tethers and bracing).” Umm… the statement about not begin tested in the U.S. and Canada is obviously true, but all child car seats sold in Europe [be it Poland or France or Sweden] must be certified according to either UNECE R44/04 or R129 meaning all current “Swedish” seats are certified according to UNECE R44 or R129 😉
If I’m not mistaken Sweden used to have it’s own standard [called “T standard” (?)] up to until it joined EU, but that was quite a while ago already.
Hi Mark and thank you for your comments,
The physics of rear-facing child restraint performance in crashes is very well studied. One of the main risks understood for decades by manufacturers and researchers is over-rotation of the carseat that leads to the child ramping up the back of the carseat in a frontal crash, such that their head is no longer protected by the shell. Factors that increase the risk of head injury in this scenario include loose installation, loose harness, exceeding the height limit and an over-reclined installation. In that many carseats are installed and used incorrectly, it is not an outlandish hypothesis to speculate that many or even most of the injuries observed in the study may have incorrect usage as a contributing factor.
Unfortunately, we at CarseatBlog do not have the budget or means to conduct such analysis or crash testing, so we do rely upon other agencies to do so. For example, in this instance, we reported upon the finding from a new study sponsored by Dorel Juvenile Group, linked in the first sentence of the first paragraph of our post. As this data conflicts with the only other previous study on the topic in the USA, of course it is reasonable to suggest that their conclusions are not data-driven. There simply isn’t enough data to make a compelling conclusion! We agree, and in fact, that’s one of the main points we make in our article.
As you suggest, we could simply pull this post, stick our heads in the sand and let parents draw their own conclusions when they see that “1-year-old children were significantly safer in FFCRs than in RFCRs in all impact types.” Parents are already finding Dorel’s statement and results and are asking questions about the conflicting information. Until we know more, we believe the correct response to reduce any risk of injury to children in motor vehicle crashes is to take the steps we outlined and have endorsed for a decade. Most importantly, use an appropriate car seat correctly in the back of a newer vehicle and drive unimpaired and undistracted. In addition, we continue to support the American Academy of Pediatrics in regard to their policy on rear-facing. If you have a dissenting opinion with our conclusion in that regard, what would you suggest parents do instead when they find this new data?
If it is the data itself that concerns you, then we are also in agreement. We are skeptical of the results as well. As we did not fund or participate in either study, specific concerns over their methodology or their results would be best forwarded to the principals of the studies that we mention in our article. Of course, anyone interested is welcome to have a general discussion about the results of these studies here, with the realization that CarseatBlog did not author or participate in this research. We are merely the messenger of this information to our readers and hope to provide a basis for education to parents on this topic.
Hi,
After reading this post, I’m afraid I’m a dissenting opinion regarding the value of this posting. It lacked a crisp hypothesis or a data-driven conclusion. Presenting an inconclusive evaluation as a basis to inform decisions is misleading to say the least. In laymen terms, this isn’t doing people any favors in informing the path they should take.
As an example, under the risks to rear facing section of this post, the 3 risks outlined assumed one or more of the following conditions being met:
1. Incorrect installation of the seat itself.
2. Incorrect harnessing of the child in the seat.
3. Child has outgrown the seat and is exceeding the manufacturer guidelines (outside their SLA).
How is this a comparable comparison to a test scenario where the state conditions are setup correctly? What is the intended conclusion?
What about down the middle scenarios besides frontal impact? For example, the car is stationary at a stop light/sign and is rear-ended. Assuming all criterion for installation are correctly met, what conclusions can be drawn both for frontal as well as rear impact?
I am 100% behind doing further research to gather more precise data (it’s time to index on precision vs. accuracy). Gathering such data may very well blow up the hypotheses that have been accepted as the standard since 2007, but this post was really poorly executed, based on what was presented Basing tests on the same test scenarios, and using statistical regression to form an educated hupotheses is what is most salient here. I suggest finding people that specialize in data science to perform the testing and have them present the data.
Until that time, my recommendatiom is to pull this post in its current form and simply state that there is research going on to further understand the benefits and drawbacks of rear facing and front facing car seats. Once all data has been collected, analyzed and agreed-upon hypotheses are formed, it will be published. That would make much more sense.
My 2 pesos worth.
Cheers
Mark
Good info indeed! We’re still deciding when to turn our 26 month old front facing and what the advantages/disadvantages would be in doing so.
Hi James, it’s most important to remember that either rear-facing or front-facing will provide great protection for a 26-month old when the carseat is installed and used correctly. If you do turn them forward-facing, just be sure to use the top-tether! That’s a critical feature, and combined with a tight installation and harness adjustment will minimize any possible risks of head injury. We hope to have more information later this year from these battling statisticians…
Thank you! Since the Dorel Position Statement is out there, parents and advocates are already finding these new claims about facing forward and asking questions. In fact, the Dorel statement about the new study and related risk of injury is the top link on the Cosco Kids and Safety 1st Service knowledge base page right now. We hope that critics will not only read our full article above, but also the Dorel Position Statement in its entirety.
We felt that this topic deserved discussion and a quick response from child passenger safety advocates. This helps parents understand the issues involved and we give them some specific suggestions on how to prevent this type of injury. It also gives other advocates and technicians a reference to use when educating parents and caregivers.
We could have ignored this new research. We could have decided not to discuss it because it shows the original research may be wrong and because the corrected results may actually conflict with the original statistics about the benefits of rear-facing. We could have been less sensational with our title or facebook promotion. The fact is that we WANT advocates and caregivers to read this information and learn how to keep their babies safe from any possible risk of head injury!
We don’t want to stick our heads in the sand and pretend this new study doesn’t exist. What if it eventually turns out to be right? And critics who are extended rear-facing advocates should consider the more immediate worst case scenario. Without articles like CarseatBlog’s, parents are still going to find the results of the new study and Dorel Position Statement on the minimum age for facing forward. Without a reference like this one, many parents would be inclined to simply turn their children forward-facing at 12 months. I’m sure that rear-facing advocates don’t want to see that happen until there is conclusive data on the topic!
Yes, we at CarseatBlog are also skeptical of the new research. As we said in the article conclusion, we will continue to support the current AAP recommendations which encourage rear-facing until at least 24 months. We also understand that this new research may have merit, if only in terms of head injury risks, and to ignore it could be putting kids at risk.
Great article! I really appreciate the measured and thoughtful response here.