Postoperative sleep disturbance is a common occurrence with significant adverse effects on patients including delayed recovery,impairment of cognitive function,pain sensitivity and cardiovascular events.The developmen...Postoperative sleep disturbance is a common occurrence with significant adverse effects on patients including delayed recovery,impairment of cognitive function,pain sensitivity and cardiovascular events.The development of postoperative sleep disturbance is multifactorial and involves the surgical inflammatory response,the severity of surgical trauma,pain,anxiety,the use of anesthetics and environmental factors such as nocturnal noise and light levels.Many of these factors can be managed perioperatively to minimize the deleterious impact on sleep.Pharmacological and non-pharmacological treatment strategies for postoperative sleep disturbance include dexmedetomidine,zolpidem,melatonin,enhanced recovery after surgery(ERAS) protocol and controlling of environmental noise and light levels.It is likely that a combination of pharmacological and non-pharmacological therapies will have the greatest impact;however,further research is required before their use can be routinely recommended.展开更多
目的使用logistic回归构建适合于关节置换术患者术后睡眠障碍(postoperative sleep disturbance,PSD)的风险预测模型。方法回顾性收集四川省成都市某三甲医院2017年1月1日–2021年9月30日行关节置换术的4286例患者资料,其中3001例作为...目的使用logistic回归构建适合于关节置换术患者术后睡眠障碍(postoperative sleep disturbance,PSD)的风险预测模型。方法回顾性收集四川省成都市某三甲医院2017年1月1日–2021年9月30日行关节置换术的4286例患者资料,其中3001例作为训练集,1285例作为测试集,在Matlab中使用逻辑回归算法筛选预测因子建模,采用列线图展示术后睡眠障碍预测风险。使用受试者工作特征曲线下面积(area under the curve,AUC)、准确率、精度、召回率、F1值和校准曲线进行模型效果评价。结果本研究最终纳入入院后术前是否睡眠障碍、病房类型、体质量指数、是否吸烟、疾病范围、关节活动度(屈曲)、关节活动度(伸)、术前末次血红蛋白以及手术类型等9个预测因子进入模型构建。预测模型的AUC值为0.708(95%置信区间:0.677~0.740),准确率为75.20%,精度为65.80%,召回率为43.70%,F1值为0.525,校准曲线显示预测概率与实际一致性较好。结论本研究构建的模型预测效能良好,且列线图简便易操作,医护人员可根据预测因子在关节置换术患者术前预测PSD的发生风险,便于及早进行预防,降低患者发生PSD的风险。展开更多
文摘Postoperative sleep disturbance is a common occurrence with significant adverse effects on patients including delayed recovery,impairment of cognitive function,pain sensitivity and cardiovascular events.The development of postoperative sleep disturbance is multifactorial and involves the surgical inflammatory response,the severity of surgical trauma,pain,anxiety,the use of anesthetics and environmental factors such as nocturnal noise and light levels.Many of these factors can be managed perioperatively to minimize the deleterious impact on sleep.Pharmacological and non-pharmacological treatment strategies for postoperative sleep disturbance include dexmedetomidine,zolpidem,melatonin,enhanced recovery after surgery(ERAS) protocol and controlling of environmental noise and light levels.It is likely that a combination of pharmacological and non-pharmacological therapies will have the greatest impact;however,further research is required before their use can be routinely recommended.
文摘目的使用logistic回归构建适合于关节置换术患者术后睡眠障碍(postoperative sleep disturbance,PSD)的风险预测模型。方法回顾性收集四川省成都市某三甲医院2017年1月1日–2021年9月30日行关节置换术的4286例患者资料,其中3001例作为训练集,1285例作为测试集,在Matlab中使用逻辑回归算法筛选预测因子建模,采用列线图展示术后睡眠障碍预测风险。使用受试者工作特征曲线下面积(area under the curve,AUC)、准确率、精度、召回率、F1值和校准曲线进行模型效果评价。结果本研究最终纳入入院后术前是否睡眠障碍、病房类型、体质量指数、是否吸烟、疾病范围、关节活动度(屈曲)、关节活动度(伸)、术前末次血红蛋白以及手术类型等9个预测因子进入模型构建。预测模型的AUC值为0.708(95%置信区间:0.677~0.740),准确率为75.20%,精度为65.80%,召回率为43.70%,F1值为0.525,校准曲线显示预测概率与实际一致性较好。结论本研究构建的模型预测效能良好,且列线图简便易操作,医护人员可根据预测因子在关节置换术患者术前预测PSD的发生风险,便于及早进行预防,降低患者发生PSD的风险。
