NHTSA study on effectiveness of child restraints

I'm not sure it this has been posted. I can't remember seeing it and nothing came up in a search.

link to full study

It's not letting me copy any text, but yeah... carseats work in reducing injury.

Wow okay. I only had time to skim, but there seemed to be some very interesting findings.
First - bowel injuries were substantially higher in the 4-7 age groups than in the under 1 or 1-3. This makes sense, since presumably, the 4-7's are in boosters. This seems to back up what we know.

Below-the-knee injury was substantially higher in the 1-3 and 4-7 groups than in the under 1 group..backing up what we know about rearfacing NOT being a huge risk for leg injury.

Rollover crashes were the most dangerous, by a large factor...which makes sense. Side impacts were next, then frontal then rewar...all of which we knew.

Skull base fracture was double in the 1-3 and 4-7 groups than in the under 1 group..which supports what we know about RFing protecting the neck. I did find it concerning however that the rates for the 1-3 and 4-7 groups were basically the same I'm wondering what the reasons are for that? I guess i would have expected it to be lower in the 4-7 group, right??

In general, injury rates were 1-2% for restrained kids, regardless of method of restraint, and 8% for unrestrained kids. This shows a GREAT reduction in injury when a restraint is used. BUT....
and here is the big, big kicker......according to table 5b and figure 3....in ALL of the age groups studied, 4-7, 1-3 and even under 1!!! the rates of injury for using a carseat and using a lap shoulder belt are essentially identical. It is *unrestrained* children that have 8 times the injury rate. SO..I'm wondering....what the heck? Are these the numbers the freakonomics people use to show that a lap shoulder belt is just a good as a car seat? Because..thats what these numbers are saying, right?

I think there will be much discussion of these findings. The study was done from '99-'08, so obviously times have changed and the seats that used to be available, the laws that used to be in effect, the societal thoughts of the time, all were different.

Anyway...I hope to steal some time later to really look at it in depth.

I'm only halfway through but around page 12 - there is the child safety seat and the lap and shoulder belt. Does lap and shoulder belt mean NO booster? Therefore, is booster included in 'child safety seat'? Is there no differentiation between booster and harness?

Also, for under 1 year old, certainly some of those children were FF but we have no way of knowing that. So, does the amount that were FF or unrestrained account for the more sever injuries in that age group?

From above:
""Skull base fracture was double in the 1-3 and 4-7 groups than in the under 1 group..which supports what we know about RFing protecting the neck. I did find it concerning however that the rates for the 1-3 and 4-7 groups were basically the same I'm wondering what the reasons are for that? I guess i would have expected it to be lower in the 4-7 group, right??""

Such a higher amount of the 4-7 kids were in lap and shoulder belts so I'm wondering if that took away any effect of stronger bones/occified spine?

This study obviously supports the use of any child safety seat but for people who care more about the specifics of what type of seat and RF/FF, it would be great to have it broken down into more categories.

Lemonade said:

I'm only halfway through but around page 12 - there is the child safety seat and the lap and shoulder belt. Does lap and shoulder belt mean NO booster. Therefore, is booster included in 'child safety seat'? Is there no differentiation between booster and harness?

Click to expand...

Child safety seat can refer to any kind of nhtsa approved child restraint: infant only, convertible, forward facing only, combination, 3-in-1, high back booster only, high back/backless booster, backless booster only, and travel vest.

bobandjess99 said:

In general, injury rates were 1-2% for restrained kids, regardless of method of restraint, and 8% for unrestrained kids. This shows a GREAT reduction in injury when a restraint is used. BUT....
and here is the big, big kicker......according to table 5b and figure 3....in ALL of the age groups studied, 4-7, 1-3 and even under 1!!! the rates of injury for using a carseat and using a lap shoulder belt are essentially identical. It is *unrestrained* children that have 8 times the injury rate. SO..I'm wondering....what the heck? Are these the numbers the freakonomics people use to show that a lap shoulder belt is just a good as a car seat? Because..thats what these numbers are saying, right?

Click to expand...

Yes, at least to some extent. Being "correctly" restrained [in back for kids] in whatever you are using and having an unimpaired/undistracted driver are what reduces your risk by huge amounts. That has been well known for a long, long time. That's also why the vast majority of our transportation safety dollars are spent in those types of programs and research. Other factors (like which seating position, which type of restraint system, which brand of carseat, etc) probably have measurable effects on injury and fatality risk too, but not by nearly as much as these big two issues.

Incidentally, I believe the original Freakonomics reports only looked at data from the FARS database. That data is generally regarded as the most reliable and impartial data from crash scenes, but only records crashes with at least one fatality. Freakonomics later added injury data to support their claim. Best I can tell, this NHTSA report also uses injury data, rather than the FARS fatality data.

In the end, no one ever refuted the Freakonomics claims directly. So, I assume they were right, at least within the paramaters and data sets they chose to discuss. Others, like CHOP, did post studies that appear to contradict Freakonomics, but never using exactly the same data. They always used a different source, or used the same source but included/omitted different variables. So, the debate may come down to who twisted the data to support their hypothesis. Perhaps they both did? Who knows:-(

Anyway, for this study, I would not draw too many conclusions beyond what they mention in their introduction and conclusion. For anything more specific to CPS issues, it is quite possible that the data they present isn't correctly adjusted to make anything more than general statements.