Background: There is a need for analytical tools predicting the risk of periodontitis. The purpose of this study was to estimate and evaluate a risk score for prediction of periodontitis.
Materials and methods: This case-cohort study included a random sample of 155 cases (with periodontitis) and 175 controls (randomly sampled from the study population at baseline) that were followed for 3-year. A logistic regression model was used with estimation of the risk ratio (RR) for each potential predictor.
Results: The risk model included the predictors "age > 53 years" (RR = 0.53), "smoking" (RR = 2.9), "gingivitis at baseline" (RR = 3.1), "subgingival calculus at baseline" (RR = 1.9), "history of periodontitis" (RR = 2.3), and "less than 2 observations in the first year of follow-up" (RR = 3.7). Patients were distributed into three risk groups based on the preanalysis risk: low risk, moderate risk, and high risk. The risk score discrimination (95% confidence interval (CI)) was 0.75 (0.70; 0.80) (p < 0.001, C-statistic).
Conclusions: The risk score estimated in the present study enabled to identify patients at higher risk of experiencing periodontitis and may be considered a useful tool for both clinicians and patients.
Keywords: epidemiology; periodontitis; risk; teeth.