Analysis of 2 173 cases of pediatric inpatients with fracture-an application of log-linear model

Objective To investigate the interactions among factors influencing the fracture of hospitalized children in Hunan province, so as to provide targeted measures for the prevention of pediatric fractures. Methods The interactions among factors influencing children's fracture were screened and analyzed by the log linear model. Results There were significant difference in the cure rate ((口恶)²=51.036, P<0.001),the duration of hospital stay (H=67.101,P<0.001) and the hospitalization expenses (H=46.282,P<0.001) of different types of fracture of hospitalized children. There were interactions between the causes of fracture and types of fracture.The fall injury was easy to cause the fracture of the ulna or radius, and the traffic accident was prone to the fracture of the tibia or fibula. There were interactions among gender, age and causes of fracture.The most common causes of fracture among 0 to 3 years old boys were traffic accidents and falls. There were interactions among gender, age, and types of fracture. 0 to3 years old children, regardless of the gender, were less inclined to the fracture of ulna, radius or humerus. Conclusions Targeted measures should be taken to prevent children from fracture according to the age, gender, causes of fracture and types of fracture.The log-linear model can be used to analyze the interactions among multiple variables, and it is suitable for exploratory research.