探讨影响股骨粗隆间骨折患者术后非感染性发热的相关危险因素，并建立一种预测其发生风险的模型。

回顾性筛选并纳入2019年1月至2021年5月苏北人民医院所收住的股骨粗隆间骨折行股骨近端抗旋髓内钉(PFNA)手术的患者。纳入标准：术前明确诊断为单侧闭合性股骨粗隆间骨折；拟行PFNA手术；既往无髋部开放性外伤、手术史；资料完整。排除标准：术前发热；术前及术后有明显的感染；术前明确诊断患有可致发热性疾病；多部位联合手术患者；病理性骨折。选取患者性别、年龄、身体质量指数(BMI)、高血压、糖尿病、吸烟史、饮酒史、麻醉方式、美国麻醉师协会(ASA)评分、手术时间、术中出血、导尿、围术期输血史、术前是否雾化、深静脉血栓、术前血红蛋白(Hb)、术前白细胞、术前白蛋白、术前C-反应蛋白，以及术后1 d的Hb、白细胞、白蛋白、C-反应蛋白等相关危险因素纳入研究。通过对纳入的指标进行单因素和多因素logistic回归分析来确定独立危险因素，并建立列线图风险预测模型预测术后并发非感染性发热的风险。

共收集316例患者，其中有103例发生非感染性发热。单因素分析结果发现性别(χ²＝8.509)、麻醉方式(χ²＝7.058)、手术时间(χ²＝90.225)、手术出血量(χ²＝24.497)、围术期输血史(χ²＝17.005)及术后1 d低白蛋白(χ²＝5.925)等两组之间差异有统计学意义(均为P<0.05)；多因素logistic回归分析发现男性、全身麻醉、手术时间长、手术出血量多、围术期输血史及术后1 d低白蛋白是导致术后非感染性发热的独立危险因素。建立列线图风险预测模型，校准曲线结果显示校正曲线与理想曲线拟合均较好，预测值同实测值基本一致。采用C-指数(C-index)评价校准曲线的效果，结果显示该列线图模型具有良好的预测能力[C-index＝0.870，95%置信区间(CI)(0.831，0.910)，校正C-index ＝0.858]。

通过研究发现PFNA术后并发非感染性发热的影响因素主要包括男性、全身麻醉、手术时间较长、手术出血量多、围术期输血史及术后D1低白蛋白，建立的列线图模型对PFNA术后发生非感染性发热风险具有良好的预测能力。然而，这些结论需要大样本、多中心的研究支持。

To explore the risk factors of postoperative noninfectious fever in patients with femoral intertrochanteric fracture, and to establish a model for predicting the risk.

The patients with intertrochanteric fractures who underwent proximal femoral nail antirotation (PFNA) surgery in Northern Jiangsu People’s Hospital from January 2019 to May 2021 were retrospectively screened. Inclusion criteria: preoperative diagnosis of unilateral closed intertrochanteric fracture of femur; PFNA operation was performed; no previous open hip trauma or surgical history; complete data. Exclusion criteria: preoperative fever; obvious infection before and after operation; preoperative diagnosis of febrile diseases; patients undergoing multi-sites combined surgery; pathological fracture. Sex, age, body mass index (BMI), hypertension, diabetes, smoking history, drinking history, anesthesia methods, American Association of Anesthesiologists (ASA)score, operation time, intraoperative blood loss, catheterization, the perioperative blood transfusion, preoperative nebulization, deep vein thrombosis, the preoperative hemoglobin, white blood cells, albumin, and C-reactive protein, as well as the postoperative hemoglobin( Hb), white blood cells, albumin, C-reactive protein on the first day after operation were selected and recorded. Uni-and multivariate logistic regression analyses were carried out to determine the independent risk factors, and a nomogram risk prediction model was established to predict the risk of postoperative noninfectious fever.

A total of 316 patients were collected, of which 103 had non-infectious fever. The results showed that male (χ²=8.509), general anesthesia (χ²=7.058), longer operation time (χ²=90.225), more blood loss (χ²=24.497), perioperative blood transfusion (χ²=17.005) and low albumin on the first day after operation (χ²=5.925) were independent risk factors leading to postoperative noninfectious fever (all P<0.05). Taking six factors into account, a nomogram risk prediction model was established. The result of calibration curve showed that the calibration curve fitted well with the ideal curve, and the predicted value was consistent with the measured value. C-index was used to evaluate the calibration curve, and the result showed the nomogram model had good prediction ability [C-index=0.870, 95% confidence interval(CI) ( 0.831, 0.910), the corrected C-index=0.858].

The influencing factors of noninfectious fever after PFNA mainly include male, general anesthesia, long operation time, large amount of bleeding, perioperative blood transfusion history and low albumin on the first day after operation. The nomogram model has a good ability to predict the risk of noninfectious fever after PFNA. However, these conclusions need the support of large samples and multi-center research.