Send to

Choose Destination
J Pediatr. 2016 Feb;169:93-7.e1. doi: 10.1016/j.jpeds.2015.10.050. Epub 2015 Nov 24.

Crib Bumpers Continue to Cause Infant Deaths: A Need for a New Preventive Approach.

Author information

BDS Data Analytics, Alexandria, VA (former CPSC project manager, Infant Suffocation Project). Electronic address:
US Department of Labor, Dallas, TX (former CPSC Corrective Actions Director).
Department of Pediatrics, Washington University, St. Louis, MO.



To assess whether clutter (comforters, blankets, pillows, toys) caused bumper deaths and provide an analysis of bumper-related incidents/injuries and their causal mechanisms.


Bumper-related deaths (January 1, 1985, to October 31, 2012) and incidents/injuries (January 1, 1990, to October 31, 2012) were identified from the US Consumer Product Safety Commission (CPSC) databases and classified by mechanism. Statistical analyses include mean age, 95% CIs, χ(2) test for trend, and ANOVA with a paired-comparisons information-criterion post hoc test for age differences among injury mechanisms.


There were 3 times more bumper deaths reported in the last 7 years than the 3 previous time periods (χ(2)(3) = 13.5, P ≤ .01). This could be attributable to increased reporting by the states, diagnostic shift, or both, or possibly a true increase in deaths. Bumpers caused 48 suffocations, 67% by a bumper alone, not clutter, and 33% by wedgings between a bumper and another object. The number of CPSC-reported deaths was compared with those from the National Center for the Review and Prevention of Child Deaths, 2008-2011; the latter reported substantially more deaths than CPSC, increasing the total to 77 deaths. Injury mechanisms showed significant differences by age (F4,120 = 3.2, P < .001) and were caused by design, construction, and quality control problems. Eleven injuries were apparent life-threatening events.


The effectiveness of public health recommendations, industry voluntary standard requirements, and the benefits of crib bumper use were not supported by the data. Study limitations include an undercount of CPSC-reported deaths, lack of denominator information, and voluntary incident reports.

Comment in

[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Elsevier Science
Loading ...
Support Center