Mastrorilli Davide, D'Oria Mario, Lepidi Sandro, Mezzetto Luca, Calvagna Cristiano, Taglialavoro Jacopo, Bruno Salvatore, Veraldi Gian Franco
Department of Vascular Surgery, University of Verona School of Medicine, University Hospital of Verona, Verona, Italy.
Division of Vascular and Endovascular Surgery, Cardiovascular Department, University Hospital of Trieste, Giuliano Isontina University Health Authority, Trieste, Italy.
J Vasc Surg. 2023 Mar;77(3):804-810.e3. doi: 10.1016/j.jvs.2022.10.011. Epub 2022 Oct 18.
The aim of the present study was to develop and validate a risk prediction model for the prediction of long-term mortality for patients with severe asymptomatic de novo carotid stenosis undergoing carotid endarterectomy (PREMYSE-CEA).
Data were collected retrospectively from a dedicated database of consecutive patients who had undergone elective CEA for severe (>70% using the NASCET [North American Symptomatic Carotid Endarterectomy Trial] criteria) asymptomatic carotid stenosis at two Italian University Hospitals from 2008 through 2016. Internal validation of the score was performed after random sampling in a 3:1 fashion. The primary end point of the PREMYSE-CEA risk score was the 5-year mortality.
Of the 1214 patients, 901 were included in the derivation cohort and 313 in the validation cohort. Using multivariable logistic regression with backward elimination, a parsimonious model was derived. A risk score incorporating eight risk factors was generated and found to be highly predictive of long-term mortality in the derivation (odds ratio [OR], 1.38; 95% confidence interval [CI], 1.28-1.41; P < .001) and validation (OR, 1.29; 95% CI, 1.21-1.37; P <.001) cohorts. The discrimination power in the receiver operating characteristic curve analysis was C = 0.775 (95% CI, 0.74-.80), and the optimism-corrected area under the curve in the bootstrapped samples was 0.761 (P < .001). A strong correlation was found between the predicted and actual mortality rates in the validation cohort (r = 0.71; P < .001).
In the present study, we have described the development, evaluation, and validation of a risk prediction model (PREMYSE-CEA) for long-term mortality after CEA in asymptomatic patients. Physicians could use the PREMYSE-CEA risk scoring tool to complement their estimates of life expectancy and prompt selective consideration of prophylactic CEA to improve the long-term benefits of interventions.
本研究旨在开发并验证一种风险预测模型，用于预测接受颈动脉内膜切除术的重度无症状初发颈动脉狭窄患者的长期死亡率（PREMYSE-CEA）。
回顾性收集2008年至2016年期间在两家意大利大学医院因重度（采用北美症状性颈动脉内膜切除术试验 [NASCET] 标准，狭窄>70%）无症状颈动脉狭窄接受择期颈动脉内膜切除术的连续患者的专用数据库中的数据。以3:1的方式随机抽样后进行评分的内部验证。PREMYSE-CEA风险评分的主要终点是5年死亡率。
在1214例患者中，901例纳入推导队列，313例纳入验证队列。使用逐步向后排除的多变量逻辑回归，得出一个简洁的模型。生成了一个包含八个风险因素的风险评分，发现其在推导队列（优势比 [OR]，1.38；95%置信区间 [CI]，1.28 - 1.41；P <.001）和验证队列（OR，1.29；95% CI，1.21 - 1.37；P <.001）中对长期死亡率具有高度预测性。受试者工作特征曲线分析中的鉴别力为C = 0.775（95% CI，0.74 - 0.80），自抽样样本中经乐观校正的曲线下面积为0.761（P <.001）。在验证队列中预测死亡率与实际死亡率之间发现了很强的相关性（r = 0.71；P <.001）。
在本研究中，我们描述了一种用于无症状患者颈动脉内膜切除术后长期死亡率的风险预测模型（PREMYSE-CEA）的开发、评估和验证。医生可以使用PREMYSE-CEA风险评分工具来补充他们对预期寿命的估计，并促使对预防性颈动脉内膜切除术进行选择性考虑，以提高干预措施的长期效益。
