老年住院患者衰弱指数不同临界值与出院预后分析

Relationship of different cut-off values of frailty index and prognosis after discharge in elderly inpatients

摘要:

目的 评价衰弱指数(FI)不同临界值确定衰弱对患者出院6个月后发生不良健康事件的预测能力。方法 选取2015年7月至2016年12月在首都医科大学附属复兴医院综合科收治的≥70岁老年住院患者312例,进行衰弱评估计算FI。采用3种不同的FI临界值(FI-1:0.20～0.45；FI-2:0.20～0.35；FI-3:0.12～0.25),将患者分为无衰弱、衰弱前期和衰弱；随访6个月以上,记录不良结局(跌倒、骨折、再入院、死亡),据此将患者分为发生不良结局组及未发生不良结局组。采用SPSS 18.0 统计软件进行分析。根据数据类型,分别采用t检验或χ²检验进行组间比较。采用Cox回归分析和受试者工作曲线(ROC)下面积(AUC)评价不同FI临界值对不良结局的预测能力。结果 312例患者FI值范围0.06～0.61,不同FI临界值(FI-1:FI≥0.45、FI-2:FI＞0.35、FI-3:FI≥0.25)确定衰弱的发生率分别为13.5%、25.3%及54.2%。出院6个月后,有 146例发生不良结局(占46.8%),其中跌倒31例(9.9%)、骨折12例(3.8%)、再入院123例(39.4%)、死亡16例(5.1%)。发生不良结局组患者的年龄、共病数量(≥4)及冠心病、慢性肾脏病、房颤、慢性阻塞性肺病的患病率显著高于未发生不良结局组(P＜0.01)。校正年龄、性别和共病数量后,Cox回归分析显示,采用FI-1临界值确定的衰弱是老年住院患者不良结局的独立预测因子(HR=2.38,5%CI 1.35～4.19, P=0.003),FI-2及FI-3确定的衰弱与不良结局无关(HR=1.63及HR=1.54,95%CI 0.96～2.77及0.56～4.25,P=0.068及P=0.406)。FI-1、FI-2和FI-3预测老年住院患者出院6个月后不良结局的AUC分别为0.685、0.673及0.644(P＜0.05),预测死亡的AUC分别为0.747、0.731及0.647(P＜0.05)。结论 衰弱指数可预测老年住院患者出院>6个月的不良结局,预测能力与FI确定衰弱的临界值有关,FI≥0.45和FI>0.35预测能力优于FI≥0.25,两者对死亡风险预测效果更佳。

Abstract:

Objective To evaluate and compare the abilities of different cut-off points of frailty index (FI) in the prediction of adverse outcomes for patients who have been discharged for 6 months. Methods A total of 312 elderly inpatients (≥70 years old) hospitalized in our hospital from July 2015 to December 2016 were enrolled, and their frailty was evaluated to calculate FI. Based on 3 different cut-off points of FI (FI-1:0.20-0.45; FI-2:0.20-0.35; FI-3:0.12-0.25), the patients were assigned into non-frailty, pre-frailty and frailty groups. All of them were followed up for at least 6 months, and were divided into adverse outcome group and no adverse outcome group according to the adverse outcomes (falls, fractures, readmission and death). SPSS statistics 18.0 was used to perform the statistical analysis. Student′s t test or Chi-square test was employed for comparison between groups based on different data types. Cox regression model and area under receiver operating characteristic (ROC) curves (AUC) were used to evaluate the predictive abilities of different cut-off points of FI for adverse outcomes. Results The range of FI values was 0.06 to 0.61 in the 312 patients, and the incidence of frailty with different cut-off points of FI (FI-1:FI≥0.45; FI-2:FI＞0.35; FI-3:FI≥0.25) was 13.5%, 25.3% and 54.2%, respectively. For those having been discharged for over 6 months, 146(46.8%) of them experienced adverse outcomes, including falls in 31(9.9%), fractures in 12(3.8%), readmission in 123(39.4%), and death in 16 patients (5.1%). The patients with adverse outcomes had older age, larger number of comorbidities, and higher incidences of coronary heart disease, chronic kidney disease, atrial fibrillation, and chronic obstructive pulmonary disease when compared with those without adverse outcomes (P<0.01).After adjustment of age, gender, and number of comorbidities, frailty determined by critical value of FI-1 was an independent predictor for adverse outcomes in the elderly inpatients (HR=2.38,5%CI 1.35-4.19,P=0.003), while that determined by FI-2 (HR=1.63,5%CI 0.96-2.77,P=0.068) and FI-3 (HR=1.54, 95%CI 0.56-4.25,P=0.406) was not associated with the adverse outcomes. The AUC of FI-1,FI-2 and FI-3 which were used to predict adverse outcomes of elderly inpatients who were discharged after 6 months was 0.685,0.673 and 0.644, respectively (P＜0.05), and was 0.747,0.731 and 0.647, respectively when they were used to predict the mortality (P＜0.05).Conclusion FI can predict adverse outcomes in elderly inpatients who have been discharged for over 6 months.The predictive ability is associated with the cut-off point of FI that determines the frailty, with those of FI≥0.45(FI-1) and FI＞0.35(FI-2) better than that of FI≥0.25(FI-3). The former 2 cut-off points also are good at predicting mortality.