Development and Internal Validation of a Self-Reported Oral Health Prediction Model for Periodontitis Risk Assessment

Abstract Background Self-reported oral health measures may provide a practical approach for identifying individuals at risk of periodontitis where comprehensive periodontal examination is not routinely available. This study aimed to develop and internally validate a self-reported oral health (SROH)-based prediction model for periodontitis risk assessment. Methods The study included 191 adults who completed a six-item SROH questionnaire and had full-mouth periodontal examination records. Periodontitis was defined based on the 2018 AAP/EFP classification. An SROH-only model and a combined model incorporating demographic and behavioural predictors were developed using multivariable logistic regression and internally validated using 1,000 bootstrap resamples. Model performance was assessed using discrimination, calibration, classification and decision curve analysis. Results Of the 191 participants, 92 (48.2%) had periodontitis. Self-reported gum disease, loose teeth and alveolar bone loss were independently associated with periodontitis, while age was the only additional predictor retained in the combined model. Apparent AUROC increased from 0.714 for the SROH-only model to 0.763, with optimism-corrected AUROC values of 0.702 and 0.745, respectively. The combined model achieved a sensitivity of 70.7%, specificity of 74.7%, positive predictive value of 72.2% and negative predictive value of 73.3%. The combined model showed generally satisfactory calibration and greater net benefit across a range of threshold probabilities. Conclusions A prediction model incorporating self-reported gum disease, loose teeth, bone loss, and age demonstrated acceptable internally validated performance for identifying individuals at risk of periodontitis. The model may support non-invasive risk assessment and referral for periodontal examination in community and non-dental settings. External validation is required before broader implementation.