Buyer Beware: When One Study Contradicts All the Rest

October 29, 2013

A recently published research article on booster seats caught my attention and caused concern, as it contrasted with findings from multiple prior studies. While there is a vast body of research supporting belt-positioning booster seats as protective for children in crashes,(1)(2) this most recent study found a higher risk of injury of any severity to the neck and thorax for children in boosters when compared to seat belts alone. As the leader of CIRP’s Child Passenger Safety Research team, I wanted to more carefully review this study and its surprising conclusions to be sure that we were providing health care providers and families with the best, most evidence-based recommendations. 

Upon further read of the article, there were two overarching concerns with the methods: the selection of cases, and the outcomes measured. Although the conventional ‘booster eligible’ age for children is from 4 to 8 years old (particularly in the study period of 1998-2009), children from 0 to 10 were included with some restrained in boosters using a lap-belt only system in order to increase the very small sample size. Additionally, the matching of children in boosters to those in a belt only or unrestrained was not completed with a high level of precision. These concerns call into question the appropriateness of the booster group selected, and the comparison between injured and controlled children. Perhaps most concerning is that very common and minor injuries to the neck (categorized as an Abbreviated Injury Scale severity scale of 1- or AIS 1) such as minor cuts, bruises, and strains were grouped with infrequent, severe injuries. This point is particularly important since nearly all studies of crash-related injury with Abbreviated Injury Scale (AIS) score severity as an outcome use a threshold of AIS 2+ or AIS 3+ to define a clinically significant injury such as skull fractures, brain bleeds, abdominal organ injuries, and large bone fractures. These concerns about both the methodology and the potential implications of the study were nicely outlined in a corresponding editorial in the same issue of the journal.

As scientists and safety advocates, we should continue to look at how to improve child occupant protection, but it is important to not change practice based on a single study with several limitations. There are two important points to consider. The first is that multiple studies have now demonstrated that belt-positioning booster seats are effective; they reduce the risk of significant injury by 45-59% when compared to children in belts. Furthermore, we know that policy and legislation help increase booster use (3) and decrease injury (4). Children should start to use a belt-positioning booster seat when they outgrow the weight or height limit of their forward-facing harness car seat. Children should then remain in a booster seat until the lap and shoulder belt properly fits, which is typically around the height of 57 inches and between the ages of 8-12 years.

The second is that we must always rigorously and critically review our own work and that of others. Although the authors of this study utilized the only available federal data source that contains a sampling of real-world crashes, the well-documented limitations of the dataset must be considered, and appropriate outcomes should be measured. Until there is a more effective system in place for collecting child-specific crash data, we must continue to advocate and carry out best practice child passenger safety guidelines- including the use of booster seats.

 

References
(1) Durbin DR,Elliott MR, Winston FK. Belt-positioning booster seats and reduction in risk of injury among children in vehicle crashes. JAMA. 2003 Jun 4;289(21):2835-40.


(2) Arbogast KB, Jermakian JS, Kallan MJ, Durbin DR. Effectiveness of belt positioning booster seats: an updated assessment. Pediatrics. 2009 Nov;124(5):1281-6.


(3) Winston FK, Kallan MJ, Elliott MR, Xie D, Durbin DR. Effect of booster seat laws on appropriate restraint use by children 4 to 7 years old involved in crashes. Archives of Pediatric Adolescent Medicine. 2007 Mar;161(3):270-5.


(4) Eichelberger AH, Chouinard AO, Jermakian JS. Effects of booster seat laws on injury risk among children in crashes. Traffic Injury Prevention. 2012;13(6):631-9.