Temporal variation in childhood injury from common recreational activities

Injury. 2011 Sep;42(9):945-57. doi: 10.1016/j.injury.2010.02.009.


Background/purpose: To investigate the month and day of the week of injury in common childhood activities using the National Electronic Injury Surveillance System database.

Methods: All emergency department visits 2002–2006 from bicycles/tricycles, scooters, playground equipment, swimming/water activities, skiing/snowboarding, trampolines, and skating were analysed. The NEISS weighted and stratified data set was analysed using SUDAAN software. Weekday and month of injury, gender, race, anatomical location of the injury, geographical location of injury, and disposition were tabulated. Simple variation by month or weekday was analysed using cosinor analysis; combined variation for both month and weekday was analysed by topographical analysis.

Results: There were an estimated 4.61 million emergency department visits for injuries from these activities in children in the United States. The average age was 9.5 years; there were 1.65 million girls (35.9%) and 2.97 boys (64.1%). Cosinor analysis demonstrated significant single peaks for month of injury for snow activities (January 27), trampolines (June 10), scooters (June 24), cycling (July 6), and water (July 12) activities. Double cosinor peaks were noted for skating (April 13 and September 12) and playground (April 22 and September 21) activities. Cosinor analyses demonstrated that the peak week days of injury were Monday for trampoline and snow activities, Saturday/Sunday for skating activities, Sunday for cycling, and Wednesday for playground equipment. There was no peak injury day for scooter or water activities. Topographical representation of paediatric injuries demonstrated that injuries from slides most frequently occurred April–May on Wednesdays and Saturdays, and those on swings April–May all days except Tuesday. Monkeybar injuries were bimodal, with the spring peak on Wednesday–Thursday in April/May and the fall peak Tuesday–Friday in September. Rollerblade injuries occurred Saturday–Sunday from March/April; rollerskates on Saturday–Sunday in January–April, and skateboards Saturday–Monday in August–October and Sundays in April.

Conclusions: These findings can be used to further guide childhood injury prevention programmes/campaigns and especially track improvements after targeted prevention programmes. Public parks and schools should check/correct the status of playground landing surfaces just before the bimodal peaks. Education campaigns reinforcing the need for bicycle helmets could be concentrated immediately before the increase in cycling activity – March on weekends and April/May for weekdays.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Accident Prevention / methods
  • Athletic Injuries / classification
  • Athletic Injuries / epidemiology*
  • Athletic Injuries / prevention & control
  • Child
  • Data Interpretation, Statistical
  • Emergency Service, Hospital / statistics & numerical data*
  • Equipment Safety / standards
  • Female
  • Hospitals / statistics & numerical data*
  • Humans
  • Male
  • Play and Playthings / injuries*
  • Seasons
  • Sex Distribution
  • Software*
  • Sports Equipment / standards
  • Time Factors
  • United States / epidemiology