Purpose : To investigate state- and trait-like risk factors leading to childhood eye injuries controlling for the between-subject difference. This study measured socioeconomic, environmental, behavioral, and injury event characteristics to identify eye injury protective and risk factors. Methods : A retrospective case-crossover study including patients aged 0-18 years old (y.o.) with severe eye trauma treated at the Canton Hospital Zenica between 2011 and 2017 was conducted. One case time point was at the time of injury, and two control time points 1 month before the injury and a month before the survey. Results : Of 36 patients meeting the criteria, four were excluded, resulting in 32 cases and 64 controls. The mean age was 10.79 in males (77.8%) and 11 y.o. in females (22.2%). In univariate GEE logistic regression unusual activity had odds of 17.25 (95%CI = 6.97, 42.70), working/chores vs.running activity odds of 6.60 (95%CI = 1.71, 25.46), very active level vs.an intermediate child activity level odds of 5.26 (1/0.19, 95%CI = 1.75, 16.67) no supervision odds of 2.63 (1/0.38, 95%CI = 1.45, 4.76) and less than 7 hours of sleep odds of 4.69 (95%CI = 1.06, 20.77) of sustaining an eye injury. Using the quasi-likelihood approach and QICu as an indicator, the best model yielded odds of getting eye injured = 0.59 + 19.35*engaging in unusual activity+0.21*supervised by an adult person+0.84*playing+3.04*working within the households+0.22*other activity. Conclusions : Giving the best model to predict injuries, the combined strategies of teaching, modifying the environment, and the watchful supervision present a preventive triad that needs to be further explored and encouraged in practice.
展开▼
