Abstract:

Objective To analyze the risk factors for gingival invagination after orthodontic treatment for extraction of the first premolar and to construct a relevant prediction model. Methods A total of 272 patients (890 tooth extraction sites) who received orthodontic treatment in Taizhou Fourth People's Hospital were retrospectively collected as research objects. By means of the set apart method, they were randomly divided into the modeling cohort (n=218, including 692 tooth extraction sites) and the internal validation cohort (n=54, 198 tooth extraction sites). Another 133 patients (with a total of 502 tooth extraction sites) who received orthodontic treatment in Nanjing Stomatological Hospital during the same period were selected as the external validation cohort. LASSO regression analysis was used to screen key variables. With the occurrence of gingival invagination as the dependent variable, multi-factor Logistic regression analysis was carried out. A nomogram prediction model was constructed based on independent risk factors, and its prediction performance was verified. Results In this study, 405 patients with 1 392 extraction sites were included, and gingival invagination occurred in 862 extraction sites, with an incidence rate of 61.9%. A total of 9 key variables were screened in the LASSO regression analysis. Results of multifactorial analysis showed that the initiation of orthodontic treatment at ≥2 weeks after extraction, thin gingival biotypic gingiva, the buccal bone thickness 4 mm from the apical CEJ root tip and the buccal bone thickness of the mid-root section of the root were independent risk factors for gingival invagination after orthodontic treatment of the first premolar (P<0.05). Results of the subjects' work characteristic curves showed that the AUCs of the modelling cohort, the internal validation cohort and the external validation cohort were 0.813 (95%CI: 0.777-0.850), 0.816 (95%CI: 0.745-0.888) and 0.815 (95%CI: 0.782-0.847), respectively. Results of the Hosmer- Lemeshow test showed that the χ² for the modelling cohort, internal validation cohort and external validation cohort were 5.768 (P=0.673), 5.719 (P=0.685) and 5.727 (P=0.680), respectively, indicating good model fitting. Kappa analysis results showed that the Kappa values of the modeling queue, internal validation queue and external validation queue were all greater than 0.6, and the predicted results were highly consistent with the actual results. Conclusion Gingival recession after orthodontic treatment of the first premolar extraction is related to various factors such as orthodontic treatment time, gingival biotype and cheekbone thickness. The column-line diagram prediction model constructed based on the related risk factors can predict the occurrence of gingival invagination more accurately, which can provide a reference for the early prevention and treatment of gingival invagination.

Key words: orthodontics, corrective, first premolar, gingival invagination, factor analysis, nomogram