摘要
目的探究慢性阻塞性肺疾病急性加重(acute exacerbations of chronic obstructive pulmonary disease,AECOPD)并呼吸衰竭患者的预后情况,分析影响预后的相关因素并构建预后的预测模型。方法选取2020年6月—2024年4月南阳市中心医院收治的满足纳排标准的AECOPD并呼吸衰竭患者的临床资料。按就诊时间,将前220例AECOPD并呼衰患者纳入建模组(n=220),后55例纳入验证组(n=55),以住院死亡率描述预后终点,基于建模组实际情况,将住院期间死亡的患者纳入预后不良组(n=42),好转出院的患者纳入预后理想组(n=178)。收集患者资料并进行logistics多因素回归分析,公式法构建预后预测模型,Hosmer-Lemeshow检验评估预测模型的拟合度,采用ROC曲线评估该模型的预测能力。结果建模组住院期间有42例患者死亡,死亡率为19.09%,178例患者好转出院,生存率为80.91%。呼衰严重程度以重度为主(52.38%),预后理想组呼衰程度以中度为主(57.87%);预后不良组年龄明显大于预后理想组,合并糖尿病、并发呼吸机相关性肺炎(ventilator-associated pneumonia,VAP)、并发多器官功能障碍综合征(multiple organ dysfunction syndrome,MODS)人数明显多于预后理想组,入院急性生理学与慢性健康状况评分II(Acute physiology and Chronic Health Score II,APACHEII)、白细胞计数(white blood cell count,WBC)、二氧化碳分压(partial pressure of carbon dioxide,PCO_(2))、血尿素氮(blood urea nitrogen,BUN)、B型脑钠肽(B-type natriuretic peptide,BNP)、D-二聚体(D-dimer,D-D)和C反应蛋白(C-reactive protein,CRP)水平明显高于预后理想组,pH值和Hb水平明显低于预后理想组,差异有统计学意义(P<0.05)。呼衰严重程度、并发VAP、并发MODS、PaCO_(2)、BUN水平、BNP水平和D-D水平是AECOPD并呼衰患者预后的独立影响因素(P<0.05)。风险预测模型概率P=1/[1+e^(-(-7.962+1.463×(呼衰严重程度)+0.997×(并发VAP)+1.692×(并发MODS)+1.013×(PaCO_(2))+0.978×(BUN)+0.899×(BNP)+1.106×(D-D)))],Hosmer-Lemeshowχ^(2)=3.974,P=0.416。通过ROC曲线分析模型的AUC为0.843(95%CI:0.788~0.888),灵敏度为83.33%,特异度为73.60%。利用验证组数据进行验证,通过ROC曲线分析模型在验证组的AUC为0.810(95%CI:0.682~0.903),灵敏度为81.82%,特异度为77.27%。结论呼衰严重程度、并发VAP、并发MODS、PaCO_(2)、BUN水平、BNP水平和D-D水平是AECOPD并呼衰患者预后的独立影响因素,以此建立的模型能够较好地预测患者预后。
Objective To explore the prognosis of patients with acute exacerbation of chronic obstructive pulmonary disease(AECOPD)and respiratory failure,to analyze the influencing factors of prognosis and construct the prediction model.Methods A retrospective study was performed on the clinical data of patients with AECOPD and respiratory failure who met inclusion/exclusion criteria in Nanyang Central Hospital between June 2020 and April 2024.According to different visit time,the first 220 cases were included into modeling group,while the last 55 cases were included into verification group.The outcomes were described by in-hospital mortality.According to prognosis,patients in modeling group were divided into poor prognosis group(n=42)and good prognosis group(n=178).The data of patients were collected and analyzed by logistics multivariate regression analysis.The prediction model for prognosis was constructed by formula method.The fit of the model was evaluated by Hosmer-Lemeshow test,and its predictive efficiency was evaluated by ROC curves.Results In modeling group,there were 42 cases(19.09%)died and 178 cases(80.91%)survived during hospitalization.The incidence of severe respiratory failure was high(52.38%),and incidence of moderate respiratory failure(57.87%)was high in good prognosis group.The age of patients in poor prognosis group was significantly elderly than that in good prognosis group,number of patients with diabetes mellitus,ventilator-associated pneumonia(VAP)and multiple organ dysfunction syndrome(MODS)was significantly more than that in good prognosis group,APACHEII score,WBC,PaCO_(2),BUN,BNP,D-D and CRP levels were significantly higher than those in good prognosis group,pH and Hb were significantly lower than those in good prognosis group(P<0.05).The severity of respiratory failure,VAP,MODS,PaCO_(2),BUN,BNP and D-D were independent influencing factors of prognosis in patients with AECOPD and respiratory failure(P<0.05).The probability of poor prognosis by risk prediction model was as follow:P=1/[1+e^(-(-7.962+1.463×(severity of respiratory failure)+0.997×(VAP)+1.692×(MODS)+1.013×(PaCO_(2))+0.978×(BUN)+0.899×(BNP)+1.106×(D-D)))],Hosmer-Lemeshowχ^(2)=3.974,P=0.416.According to ROC curves analysis,AUC,sensitivity and specificity of the model were 0.843(95%CI:0.788-0.888),83.33%and 73.60%.According to data verification,ROC curves analysis showed that AUC,sensitivity and specificity of the model in verification group were 0.810(95%CI:0.682-0.903),81.82%and 77.27%,respectively.Conclusion The severity of respiratory failure,VAP,MODS,PaCO_(2),BUN,BNP and D-D are independent influencing factors of prognosis in patients with AECOPD and respiratory failure,and the model constructed based on the above factors can better predict prognosis of patients.
作者
吴哲
董万里
李克芳
孙潺
李慧
WU Zhe;DONG Wan-li;LI Ke-fang;SUN Chan;LI Hui(Department of Respiratory and Critical Care Medicine,Nanyang Central Hospital,Nanyang,Henan 473000,China)
出处
《医药论坛杂志》
2025年第3期253-259,共7页
Journal of Medical Forum
基金
河南省中医药科学研究专项课题(2022ZY1200)。
关键词
慢性阻塞性肺疾病急性加重
呼吸衰竭
预后
影响因素
预测模型
Acute exacerbation of chronic obstructive pulmonary disease
Respiratory failure
Prognosis
Influencing factor
Prediction model
