Gong Yangfan, Zhang Kai, Chen Wei, Yang Qiqi, Shi Mingyue, Dong Zhao, Yin Zhuohao, Zhang Yuyu, Ge Wei
Department of General Practice, Xijing Hospital, Air Force Medical University, Xi'an, China.
Department of Biostatistics, School of Public Health, Cheeloo College of Medicine, Shandong University, Jinan, China.
Front Med (Lausanne). 2025 May 22;12:1532196. doi: 10.3389/fmed.2025.1532196. eCollection 2025.
The aim of this study was to develop a novel nomogram for predicting one-year mortality in the older adult patients with hip fracture and to further evaluate its effectiveness.
This retrospective cohort research analyzed the clinical data of 1,263 older adult patients with hip fractures who underwent surgery at the First Affiliated Hospital of Air Force Military Medical University from January 2014 to December 2022. Patients receiving surgical treatment during January 2014 to December 2019 (864 cases) for the model development and further, data from the same centre with same inclusion criteria from January 2020 to December 2022 (399 cases) for the external validation of the model. The univariate and multivariable logistic regression were utilized to identify independent risk factors linked to one-year mortality. A predictive nomogram was subsequently developed. The discriminatory power of the model and its accuracy were monitored by utilizing receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis. Furthermore, visual risk applications were developed to enhance usability.
The one-year mortality is 16.8%. A total of seven predictors, namely age, body mass index (BMI), fibrinogen (FIB), stroke, dementia, ASA (American Society of Anesthesiologists), intraoperative blood transfusion were identified by multivariate analysis from a total of 65 variables studied. The model constructed using these seven predictors displayed medium prediction ability, with an area under the ROC of 0.775 in the training set and 0.740 in the validation set. The calibration curve shows a good degree of fitting between the predicted and observed probabilities. The DCA curve showed that the nomogram could be applied clinically if the risk threshold was between 8 and 64%, which was found to be between 6 and 80% in the external validation.
Independent factors, including age, BMI, preoperative fibrinogen level, stroke, dementia, ASA, intraoperative blood transfusion are pivotal in influencing one-year survival rate for patients with hip fractures. This risk dynamic nomogram developed from these factors renders substantial predictive accuracy and clinical utility, providing a reliable basis for a reasonable and personalized treatment plan.
本研究旨在开发一种新型列线图，用于预测老年髋部骨折患者的一年死亡率，并进一步评估其有效性。
这项回顾性队列研究分析了2014年1月至2022年12月在空军军医大学第一附属医院接受手术的1263例老年髋部骨折患者的临床资料。2014年1月至2019年12月期间接受手术治疗的患者（864例）用于模型开发，另外，来自同一中心2020年1月至2022年12月符合相同纳入标准的数据（399例）用于模型的外部验证。采用单因素和多因素逻辑回归来确定与一年死亡率相关的独立危险因素。随后开发了一种预测列线图。通过使用受试者操作特征（ROC）曲线、校准曲线和决策曲线分析来监测模型的辨别力及其准确性。此外，还开发了视觉风险应用程序以提高可用性。
一年死亡率为16.8%。在总共研究的65个变量中，多因素分析确定了七个预测因素，即年龄、体重指数（BMI）、纤维蛋白原（FIB）、中风、痴呆、美国麻醉医师协会（ASA）分级、术中输血。使用这七个预测因素构建的模型显示出中等预测能力，训练集的ROC曲线下面积为0.775，验证集为0.740。校准曲线显示预测概率与观察概率之间具有良好的拟合度。DCA曲线表明，如果风险阈值在8%至64%之间，该列线图可在临床上应用，在外部验证中发现该阈值在6%至80%之间。
包括年龄、BMI、术前纤维蛋白原水平、中风、痴呆、ASA分级、术中输血在内的独立因素对髋部骨折患者的一年生存率有关键影响。由这些因素开发的这种风险动态列线图具有较高的预测准确性和临床实用性，为合理的个性化治疗方案提供了可靠依据。
