[Influence of retrospective time period on reliability of epidemiological survey to injuries]

To analyze the influence of different recall time periods on epidemiological survey of injuries among children and adolescents.

A study on injuries for 12 months follow-up period was actively surveyed among 4523 pupils selected by cluster sampling from 12 kindergartens, 1 elementary school, 1 junior high school, 1 senior high school and 1 college in Maanshan city in November, 2007 to November, 2008. Injury events were actively surveyed by school doctors or class health workers simultaneously. Participants were divided into three groups, the first group including self-reported injuries by 3-months interval, the second group with self-recalled injuries by 6-months interval and the third group with self-reported injuries in the past 12 months. Validity and reliability of injuries between surveillance injuries and retrospective injuries were compared.

The overall rate of injury events was 74.2% and injury events and rate was higher in boys than that in girls (P < 0.01), In total, 6350 injury events were registered for the survey. Compare with surveillance findings, the injury rate declined 0.02% and the number of injuries had a 19 person/time reduction from the retrospective injuries by 3-months interval. The injury rate declined 0.02% and number of injuries reduced 58 person/times from the survey of retrospective injuries by 6-month interval. The rate of injuries declined 8.8% and frequencies had a 505 reduction from data of self-reported injuries in the past 12 months. The largest declines were found for the 3 to 6 year-old age group and for minor injuries.

The 12-month recall period on underestimate injury rates was compared to the 3-month recall periods and 6-month recall periods. Recall periods of 6 months were recommended to be used in injury survey in children and adolescents.