Application of the Poisson Regression Model with Zero Accumulation in the Analysis of Dental Data in 12- and 13-year-old adolescents

Additional zeros in some counting data are a common phenomenon that limit the use of the traditional poisson regression model for analysis. The aim of the present study was to use the zero-inflated poisson regression model for analyzing dental data in 12- and 13-year-old adolescents.
In this descriptive study, the dental data of 12- and 13-year-old students in Pakdasht and Pishva provinces of Varamin, who were randomly selected from a list of urban and rural schools, were selected. Overall, 764 samples were selected and their data for dental caries was evaluated. The DMFT index was used to evaluate the teeth condition of the students. At first, the consent of the participants was received for participation in the project. Given the fact that the response variable and zero accumulation were numbered in the number of drawn, filled, and rotten teeth, the index was modeled with a poisson regression model accumulated at zero. Data analysis was performed using the R software (3.2.2).
Overall, 414 Students were females. Mean and standard deviation of DMFT was 1.02 ± 1.35. In total, 54% (389 students) had dental caries. For the intervention variable (such as: preventive treatment or routine treatment, rapid treatment including scaling, immediate pain treatment, or tooth infection) (P = 0.04), dental health variable (P = 0.003), variable of pain (P = 0.049), brushing variable (P = 0.111), and dental floss (P = 0.040) had a significant relationship with the DMFT index.
The prevalence of dental caries in 6- and 12-year-old students in the city of Varamin was greater than the World Health Organization standards. For modeling DMFT students, the poisson regression accumulated at zero was the most appropriate model. Therefore, in order to improve the existing situation, proper planning and education of families about oral health are suggested.