Supplementary Materials and Analysis Code for: Individualized Nomogram-Based Model for Predicting the Risk of Gingival Recession after Microsurgical Endodontics

Abstract Background: Microsurgical endodontic surgery is a standard tooth-preserving treatment. However, postoperative gingival recession, particularly in aesthetically critical areas, remains a common complication. A reliable tool for preoperative risk assessment is currently lacking. Objective: To identify key risk factors and develop an individualized nomogram for predicting gingival recession risk following microsurgical endodontics. Methods: This retrospective study analyzed 200 patients, including 50 with postoperative recession. Potential predictors were screened using Least Absolute Shrinkage and Selection Operator (LASSO) regression, and significant variables were incorporated into a multivariate logistic regression model to construct a nomogram. The model’s discrimination was assessed by the C-index, calibration was evaluated via calibration curves and the Hosmer-Lemeshow test, and internal validation was performed using 1,000 bootstrap resamples. Results: The study identified seven independent predictors: thin gingival biotype (OR=5.74), sulcular incision (OR=3.55), presence of restoration (OR=3.29), smoking (OR=2.96), and advanced age (OR=1.05). Conversely, greater keratinized gingiva width (OR=0.57) and the use of floss or water flosser (OR=0.35) were identified as protective factors. The final nomogram demonstrated excellent discrimination with a C-index of 0.87 (95% CI: 0.81; 0.92) and good calibration, supported by a Nagelkerke R2 of 0.46, a non-significant Hosmer-Lemeshow test (P=0.917), and high agreement in calibration curves (Mean Absolute Error=0.026). Conclusion: A validated nomogram integrating anatomical, surgical, and behavioral factors was developed. This tool facilitates preoperative identification of high-risk patients, enabling targeted preventive strategies to optimize aesthetic outcomes. Description of supplementary files This repository contains the analysis code and supplementary tables for the study. Analysis_Code_R.docx: The R code scripts used for LASSO regression, multivariate logistic regression, and nomogram construction. Supplementary Table 1.docx: Verification of statistical assumptions for the final multivariate logistic regression model. Supplementary Table 2.docx: Internal validation of the final model using Bootstrap method (1,000 resamples).摘要 背景：显微外科牙髓手术是标准的保牙治疗手段，但术后牙龈退缩仍是常见并发症，在美学敏感区域尤为突出，目前仍缺乏可靠的术前风险评估工具。 目的：明确显微牙髓术后牙龈退缩的关键危险因素，并构建个体化列线图（nomogram）以预测该风险。 方法：本回顾性研究共纳入200例患者，其中50例出现术后牙龈退缩。采用最小绝对收缩和选择算子（Least Absolute Shrinkage and Selection Operator, LASSO）回归筛选潜在预测因子，并将具有统计学意义的变量纳入多因素logistic回归模型，进而构建列线图。通过C指数（C-index）评估模型区分度，采用校准曲线与Hosmer-Lemeshow检验评估模型校准度，并通过1000次Bootstrap重抽样进行内部验证。 结果：本研究共确定7项独立预测因子：薄牙龈生物型（比值比OR=5.74）、沟内切口（OR=3.55）、存在修复体（OR=3.29）、吸烟（OR=2.96）及高龄（OR=1.05）；反之，较大角化牙龈宽度（OR=0.57）与使用牙线或水牙线（OR=0.35）为保护因素。最终构建的列线图展现出优异的区分度，C指数为0.87（95%置信区间CI：0.81~0.92），校准度良好，Nagelkerke R²为0.46，Hosmer-Lemeshow检验无统计学差异（P=0.917），校准曲线一致性高（平均绝对误差=0.026）。 结论：本研究构建了整合了解剖学、手术学及行为学因素的验证型列线图。该工具可辅助术前识别高危患者，从而制定针对性预防策略以优化美学治疗效果。 补充文件说明 本仓库包含本研究的分析代码与补充表格。 Analysis_Code_R.docx：用于LASSO回归、多因素logistic回归及列线图构建的R代码脚本。 Supplementary Table 1.docx：最终多因素logistic回归模型的统计学假设验证。 Supplementary Table 2.docx：采用Bootstrap法（1000次重抽样）对最终模型进行的内部验证。