Key points are not available for this paper at this time.
OBJECTIVE: Mortality and morbidity data on childhood injury are used to construct developmentally appropriate intervention strategies and to guide pediatric anticipatory counseling on injury prevention topics. Effective anticipatory guidance depends on detailed injury data showing how risks change as children develop. Conventional age groupings may be too broad to show the relationship between children's development and their risk of various causes of injury. Previous studies revealed differences in overall rates and specific causes of injury by year of age. However, single year of age rates for children younger than 4 years may not reflect the variations in risk as a result of rapid developmental changes. This study was designed to analyze injury rates for children younger than 4 years by quarter-year intervals to determine more specifically the age period of highest risk for injury and for specific causes. METHODS: We used data from 1996-1998 California hospital discharges and death certificates to identify day of age and external cause of injury (E-code) for children younger than 4 years. The number of California residents for each day of age was estimated from US Census of estimates of California's population by year of age for the midpoints (1996-1998). Rates were calculated by 3-month intervals. We grouped the E-codes into major categories that would be particularly relevant for developmentally related risks of injury specific to young children. The categorization took into account physical, motor, behavioral, and cognitive developmental milestones of children 0 to 3 years. RESULTS: There were a total of 23,173 injuries; 636 resulted in death. The overall annual rate for children aged 0 to 3 years was 371/100,000. Beginning at age 3 to 5 months, the overall rate of injury rapidly increased with increased age, peaking at 15 to 17 months. The mean injury rate calculated for each single year of age did not reflect the variation and the highest rate of injury by quarter year of age for children younger than 1 year, 1 year, and 2 years. The leading major causes of injury in descending order were falls, poisoning, transportation, foreign body, and fires/burns. The overall rate of the major category of falls exceeded poisoning, the second leading cause of injury, by a factor of 2. Age-related differences were detected within each major cause of injury. For children 0 to 12 months of age, there was a different leading cause of specific injury for each 3-month period: other falls from height (0-2 months), battering (3-5 months), falls from furniture (6-8 months), and nonairway foreign body (9-11 months). Hot liquid and vapor injuries were the leading specific causes for children 12 to 17 months. Poisoning by medication was the leading specific cause of injury for all age groups from 18 to 35 months and exceeded poisoning by other substances. Pedestrian injury was the leading specific cause of injury for all age groups from 36 to 47 months. Fall from furniture has the highest rates of specific causes of falls from age 3 to 47 months. Fall from stairs peaked at age 6 to 8 months and 9 to 11 months. Fall from buildings was highest at 24 to 26 months. Poisoning by medication peaked at age 21 to 23 months, but poisoning by other substances peaked at 15 to 17 months. The motor vehicle occupant injury rates were fairly stable over the age span of this study. The pedestrian injury rate increased beginning at age 12 to 14 months and by 15 to 17 months was double that of the motor vehicle occupant. Foreign body had a marked peak at age 9 to 11 months. Both battering and neglect rates were highest among infants 0 to 2 and 3 to 5 months. Bathtub submersions had a narrow peak at age 6 to 11 months. Other submersions peaked at 12 to 14 months and remained high until 33 to 35 months. CONCLUSIONS: We departed from usual groupings of E-codes and devised groupings that would be reflective of age-related developmental characteristics. Differences in rates by narrow age groups for young children can be related to developmental achievements, w can be related to developmental achievements, which place the child at risk for specific causes of injury. We found marked variability in both rates and leading causes of injury by 3-month interval age groupings that were masked by year of age analyses. Children aged 15 to 17 months had the highest overall injury rate before age 15 years. This coincides with developmental achievements such as independent mobility, exploratory behavior, and hand-to-mouth activity. The child is able to access hazards but has not yet developed cognitive hazard awareness and avoidance skills. A remarkable finding was the high rate of battering injury among infants 0 to 5 months, suggesting the need to address potential child maltreatment in the perinatal period. Poisoning was the second major leading cause of injury; more than two thirds were medication. Cultural factors may influence views of medications, storage practices, use of poison control system telephone advice, and risk of toddler poisoning. The pedestrian injury rate doubled between 12 and 14 months and 15 and 17 months and exceeded motor vehicle occupant injury rates for each 3-month interval from 15 to 47 months. Pedestrian injury has not received sufficient attention in general and certainly not in injury prevention counseling for children younger than 4 years. Anticipatory guidance for pedestrian injury should be incorporated before 1 year of age. Effective strategies must be based on the epidemiology of childhood injury. Pediatricians and other pediatric health care providers are in a unique position to render injury prevention services to their patients. Integrating injury prevention messages in the context of developmental assessments of the child is 1 strategy. These data can also be used for complementary childhood injury prevention strategies such as early intervention programs for high-risk families for child abuse and neglect, media and advocacy campaigns, public policies, and environmental and product design.
Agran et al. (Sun,) studied this question.