目的 调查与探讨高龄髋部骨折患者术后谵妄（POD）的发生因素,并提出相关处理对策。方法 选取2019年8月—2022年12月择在南阳市中医院独山院区进行手术治疗的高龄髋部骨折患者82例为研究对象,所有患者在术前1 d进行机械痛阈评定,在术后7 d判定患者的POD发生情况,进行POD与术前痛阈水平的相关性分析,并提出相关的处理对策。结果 术后7 d,82例患者中发生POD 12例（谵妄组）,占比14.6%,未发生POD 70例（非谵妄组）,占比85.37%。谵妄组的性别、体质指数、骨折类型、骨折至手术时间与非谵妄组对比差异无统计学意义（P>0.05）,谵妄组的年龄、术前血红蛋白水平、术前白蛋白水平与非谵妄组对比差异有统计学意义（P<0.05）。谵妄组的术前1 d的痛阈水平低于非谵妄组（P<0.05）。Spearman分析显示,POD与术前痛阈、年龄、术前血红蛋白、术前白蛋白均存在相关性（P<0.05）。Logistic回归分析显示,术前痛阈、年龄、术前血红蛋白、术前白蛋白等都为导致POD发生的影响因素（P<0.05）,要积极加强预防性护理干预。结论 高龄髋部骨折患者POD的发生率较高,患者的术前痛阈、年龄、术前血红蛋白、术前白蛋白等均为导致POD发生的影响因素,要积极加强预防性护理干预。

Objective To investigate and explore the factors leading to postoperative delirium（POD）in elderly patients with hip fractures,and to propose relevant handling measures．Methods From August 2019 to December 2022,82 cases of elderly patients with hip fractures who underwent surgical treatment in Nanyang Hospital of Traditional Chinese Medicine Dushan District were selected as the research subjects．All patients underwent mechanical pain threshold assessment 1 day before surgery,and their postoperative delirium were determined 7 days after surgery,followed by correlation analysis,and relevant handling measures were proposed．Results Seven days after surgery,there were 12 patients（delirium group）of POD,accounted for 14.6%,and 70 patients（non delirium group）without POD,accounted or 85.37%．There was no significant difference in genders,body mass index,fracture types and fracture to surgery time compared between the delirium group and the non delirium group（P>0.05）． However,there were significant differences in ages,preoperative hemoglobin levels and preoperative albumin levels compared between the delirium group and the non delirium group（P<0.05）．The pain threshold level of the delirium group on the first day before surgery was significantly lower than that of the non delirium group（P<0.05）．Spearman analysis showed that POD was associated with preoperative pain threshold,ages,preoperative hemoglobin and preoperative albumin levels（P<0.05）．Logistic regression analysis showed that the preoperative pain threshold level,ages,preoperative hemoglobin and preoperative albumin levels were all independent risk factors for the development of POD（P<0.05）,preventive nursing intervention should be actively strengthened．Conclusions The incidence of POD is high in elderly patients with hip fractures．Preoperative pain threshold level,age,preoperative hemoglobin and preoperative albumin levels are all factors that contribute to the occurrence of POD．It is necessary to actively strengthen preventive nursing interventions．

术后谵妄(POD)指术后严重的注意力及神经认知障碍,其发病率高,且可致多种术后并发症的发生率增加,老年患者为其高危人群之一。相关研究显示:心率变异性(HRV)作为反映自主神经系统(ANS)功能的生物电指标,与老年患者POD的发生相关。本文综述了近年HRV指数与老年患者POD关系的研究,描述了老年患者POD的流行病学规律、ANS功能异常引发POD的可能机制以及HRV与神经认知功能及POD的可能联系,以期为POD的防治提供新的思路。

Postoperative delirium (POD) is a syndrome of severe postoperative attention and neurocognitive impairment, which has a high incidence and can lead to an increased incidence of various postoperative complications. Elderly patients are one of the high-risk groups for POD. Relevant studies have shown that heart rate variability (HRV), as a bioelectrical indicator reflecting the function of the autonomic nervous system (ANS), is associated with the occurrence of POD in elderly patients. This paper reviewed the recent studies on the relationship between HRV index and POD in elderly patients, described the epidemiological regularity of POD in elderly patients, the possible mechanism of POD caused by abnormal ANS function, and the possible connection between HRV and neurocognitive function or POD, in order to provide new evidence for the prevention and treatment of POD.

      近年来，人工智能技术（AI）的发展正在逐渐改变传统的医疗行业，机器学习作为人工智能技术中的主流被越来越多地应用于分析复杂的医学数据，为疾病的诊断、预后风险评估、诊疗决策的制定等方面提供了便利。文章对国内外机器学习算法在术后谵妄中的应用进行综述，以期为术后谵妄预测模型的构建提供新的思路，为临床早期评估术后谵妄提供新的依据。

    In recent years，the development of artificial intelligence（AI）is gradually changing the traditional medical industry．Machine learning，as the mainstream of artificial intelligence technology，is increasingly applied to analyze complex data in medical research．It provides convenience for disease diagnosis，risk assessment and diagnosis and treatment decision making．This paper reviews the application of machine learning algorithms in postoperative delirium at home and abroad，in order to provide a new idea for the construction of postoperative delirium prediction model and a new basis for early clinical evaluation of postoperative delirium．