四肢创伤骨折病人术后低蛋白血症的危险因素分析及风险预测模型的构建

目的 探讨四肢创伤骨折病人术后低蛋白血症的危险因素，并构建四肢创伤骨折病人术后低蛋白血症的风险预测模型。方法 选取2018年3月至2022年8月池州市人民医院手术治疗的226例四肢创伤骨折病人作为研究对象。应用Lasso分析筛选四肢创伤骨折病人术后低蛋白血症的预测因素，应用logistic回归分析筛选四肢创伤骨折病人术后低蛋白血症的危险因素，应用R软件构建四肢创伤骨折病人术后低蛋白血症的列线图模型，并对四肢创伤骨折病人术后低蛋白血症的列线图模型进行验证。结果 logistic回归分析结果显示，女性、年龄≥65岁、术中出血量≥300 mL、胃肠功能损伤、肺部感染、术前白蛋白<35 g/L、衰弱、输注胶体液及蛋白等是四肢创伤骨折病人术后低蛋白血症的危险因素（P<0.05）。四肢创伤骨折病人术后低蛋白血症的列线图模型的校正曲线显示预测值与实际值的拟合度较好；模型的AUC是0.84[95%CI:(0.80,0.87)]；决策曲线显示阈值概率在2%～75%范围内时，应用该列线图预测四肢创伤骨折病人术后低蛋白血症净获益值较高。结论 女性、年龄≥65岁、术中出血量≥300 mL、胃肠功能...

Analysis of risk factors for postoperative hypoalbuminemia in patients with traumatic fractures of extremities and construction of risk prediction model

Objective To investigate the risk factors of postoperative hypoalbuminemia in patients with limb traumatic fractures, andto construct a risk prediction model for postoperative hypoalbuminemia in patients with limb traumatic fractures.Methods A total of 226 patients with limb trauma fractures who entered Chizhou People''s Hospital for surgical treatment from March 2018 to August 2022were selected as the research subjects.The predictive factors of postoperative hypoalbuminemia in patients with extremity trauma frac.tures were screened by Lasso analysis, and the risk factors of postoperative hypoalbuminemia in patients with limb trauma fractureswere screened by logistic regression analysis. The R software was used to construct postoperative hypoalbuminemia in patients withlimb trauma fractures. The nomogram model of post-operative hypoalbuminemia in patients with limb traumatic fractures was validated. Results Logistic regression analysis showed that female, age≥65 years, intraoperative bleeding≥300 mL, gastrointestinal function in.jury, pulmonary infection, preoperative albumin<35 g/L, weakness, colloid fluid and protein infusion were the risk factors of postopera.tive hypoproteinemia in patients with traumatic fracture of extremities (P<0.05). The correction curve of the nomogram model of postop.erative hypoproteinemia in patients with traumatic fracture of extremities showed that the predicted value was well matched with the ac.tual value; AUC of the model is 0.84[95%CI:(0.80,0.87)]; When the decision curve shows that the threshold probability is within the range of 2%-75%, the application of this nomograph to predict the net benefit of postoperative hypoproteinemia in patients with traumat.ic fracture of extremities is higher.Conclusions Female, age≥65 years, intraoperative bleeding≥300 mL, gastrointestinal function inju.ry, pulmonary infection, preoperative albumin<35 g/L, weakness, colloid fluid and protein infusion are risk factors for postoperative hy.poproteinemia in patients with traumatic fracture of extremities.The nomogram constructed based on the above eight risk factors in thisstudy can help prevent and treat postoperative hypoproteinemia in patients with traumatic fracture of extremities.