ObjectiveThis study aimed to identify independent risk factors for head and neck squamous cell carcinoma (HNSCC) based on the surveillance, epidemiology, and end results (SEER) database and to develop a nomogram model for predicting patient survival outcomes. MethodsPatients diagnosed with HNSCC from 1975 to 2021 were selected from the SEER database. After applying inclusion and exclusion criteria, 2 271 patients were included and randomly divided into a training cohort and a validation cohort in a 7∶3 ratio. Independent prognostic factors were identified using LASSO regression, Cox regression analysis, and the Akaike information criterion (AIC). A nomogram model was constructed, and its discrimination and calibration were assessed using the concordance index (C-index), time-dependent area under the curve (time-dependent AUC), and calibration curves. The nomogram model was compared with the American Joint Committee on Cancer (AJCC) staging system using decision curve analysis (DCA), net reclassification index (NRI), and integrated discrimination improvement (IDI) to evaluate clinical utility and risk stratification performance. ResultsFive independent prognostic factors (age, marital status, N stage, tumor stage, and radiotherapy) were selected to build the nomogram model for HNSCC. The C-index values of the model were 0.731 4 (95%CI 0.714 5 to 0.748 5) in the training cohort and 0.735 1 (95%CI 0.709 1 to 0.761 0) in the validation cohort. The time-dependent AUC values were all above 0.7, indicating good discriminatory ability. Moreover, decision curve analysis showed that the nomogram model provided higher clinical net benefits at different threshold probabilities and performed better than the AJCC staging system in identifying high-risk patients. ConclusionThis study develops a nomogram model based on the SEER database to predict survival outcomes in patients with HNSCC. The model demonstrates high discrimination and clinical utility, offering a personalized prognostic tool for clinicians.
ObjectivePeriodontal disease may increase the risk of cardiovascular disease(CVD). Periodontal inflammation can contribute to CVD through both local and systemic inflammatory pathways. The Periodontal Inflamed Surface Area (PISA) quantifies the extent of periodontal inflammation; however, evidence based on large, representative populations remains limited. This study aimed to evaluate the association between PISA and CVD using data from the National Health and Nutrition Examination Survey (NHANES). MethodsThis study utilized data from six cycles of NHANES, including a total of 8 925 participants. Three machine learning algorithms—Random Forest, XGBoost, and Boruta—were used for key feature selection. A logistic regression model was constructed to assess the association between PISA and CVD prevalence, and further machine learning prediction models were developed to explore the additional predictive value of PISA in identifying CVD status. ResultsParticipants in the highest quartile of PISA showed a higher prevalence of CVD. Multivariable logistic regression analysis and restricted cubic spline analysis demonstrated that PISA score was positively associated with CVD in a linear fashion (OR=2.85, 95%CI 1.75 to 4.62, P<0.001). ConclusionIn individuals with periodontitis, the PISA index is linearly associated with the prevalence of cardiovascular disease events and their subtypes. As a continuous measure of periodontal inflammatory burden, PISA may help identify people at higher risk of CVD.