Logistic regression model for identifying factors affecting hospitalization of children with pneumonia

Abstract

Pneumonia is a lung infection that could happen in babies, children, adults and older people. However, pneumonia in infants and older adults is more serious. Several studies found that infants are more likely to get pneumonia if they live in low-income families. The study aimed to identify factors that cause children to be hospitalized for pneumonia. The binary logistic regression analysis was performed to build a full model regardless of the significance of the variables. The forward selection approach was used to select the significant variables. It was found that the age of the mother, cigarette smoked by the mother during pregnancy, duration (in months) of the children on solid food, and the age when the child had pneumonia with the p-value of 0.0009, 0.0010, 0.0003 and less than 0.0001, respectively. The odds ratio of mother's age, cigarette smoked by mother during pregnancy, how many months the child on solid food, and children’s age when they had pneumonia are 0.69, 6.22, 0.40 and 0.60, respectively.