Abstract: Objective　To analyze the death of chronic heart failure（CHF）patients during 2-year-integrated management，to observe the outcome of integrated management，and to analyze the influence factors for death.Methods　A total of 381 CHF patients who arrived at Yuetan Community Health Service Center or Fuxing Hospital at least once medical consultation every 3 months from 2012 to 2013，were divided into management group（194 cases） and control group（187 cases）according to random principle and the patients’ desire.Patients in management group received the comprehensive management，which included patients self-management，management by GPs，and regular instruction by specialists.Patients in control group received routine treatment.Patients were followed up for 2 years，the death and cause of death were recorded，Cox proporational hazards regression model was used to analyze influence factors for death.Results　There were significant differences in history of renal insufficiency，levels of Cr，eGFR，albumin，TC and TG at baseline，taking rates of ACEI/ARB and β-receptor blocking agent at follow-up between two groups of patients（P<0.05）.During the follow-up period，130 cases（67.0%）in management group completed 2-year-follow-up，39 cases（20.1%）died，among whom 13 cases died from cardiac diseases，11 cases died from infectious diseases，15 cases died from other causes；104 cases（55.6%）in control group completed 2-year-follow-up，56 cases（29.9%）died，among whom 18 cases died from cardiac diseases，25 cases died from infectious diseases,13 cases died from other causes.There was significant difference in all-cause mortality between two groups（χ2=7.36，P=0.007），and there was significant difference in infection-induced mortality between two groups（χ2=8.47，P=0.004）.However，there were no significant difference in cardiac and other causes mortality between two groups（χ2=2.15，χ2=0.03，P>0.05）.According to Cox proporational hazards regression model analysis，the independent factors for cardiac mortality included aged，lower LVEF level，higher FPG level，lower HDL-C level at baseline and no use of β-receptor blocking agent during follow-up（P<0.05）；the independent factors for infection-induced death included non-comprehensive management,aged，COPD，history of cerebrovascular disease，and higher Cr level at baseline（P<0.05）；aged and lower LVEF level at baseline were independent factors for other causes mortality（P<0.05）.Conclusion　Non-cardiac mortality is more common among CHF patients.CHF management can reduce all-cause mortality，mainly can reduce infection-induced mortality.To reduce mortality，comprehensive management for CHF patients should be carried out.

Key words: Heart failure, Disease management, General practitioners, Mortality, Root cause analysis

摘要： 目的　分析慢性心力衰竭（慢性心衰）患者综合管理2年死亡情况，以探讨综合管理效果，并分析其影响因素。方法　选取2012—2013年在月坛社区卫生服务中心或复兴医院至少每3个月就诊1次的慢性心衰患者381例，按照随机原则及患者意愿将其分为管理组（194例）和对照组（187例）。管理组采取患者自我管理为基础、全科医生管理为主、专科医生定期指导的全方面管理，对照组为患者常规就诊。对患者随访2年，记录患者死亡情况及死亡原因，并采用Cox比例风险回归模型分析影响患者死亡的因素。结果　两组患者基础肾功能不全病史、肌酐（Cr）、估算肾小球滤过率（eGFR）、清蛋白、总胆固醇（TC）、三酰甘油（TG）水平及随访中血管紧张素转化酶抑制剂（ACEI）/血管紧张素Ⅱ受体阻滞剂（ARB）服用率、β-受体阻滞剂服用率比较，差异均有统计学意义（P<0.05）。随访过程中，管理组有130例（67.0%）患者完成2年随访，39例（20.1%）患者死亡，其中心源性死亡13例，感染性死亡11例，其他原因死亡15例；对照组有104例（55.6%）患者完成2年随访，56例（29.9%）患者死亡，其中心源性死亡18例，感染性死亡25例，其他原因死亡13例。两组全因病死率比较，差异有统计学意义（χ2=7.36，P=0.007）；两组感染性病死率比较，差异有统计学意义（χ2=8.47，P=0.004），两组心源性病死率、其他原因病死率比较，差异均无统计学意义（χ2=2.15、0.03，P>0.05）。Cox比例风险回归模型分析结果显示，高龄、基础低左心室射血分数（LVEF）、高空腹血糖（FPG）、低高密度脂蛋白胆固醇（HDL-C）及随访中不服用β-受体阻滞剂是心源性死亡的独立危险因素（P<0.05）；未进行综合管理、高龄、慢性阻塞性肺疾病、脑血管疾病、基础高Cr是感染性死亡的独立危险因素（P<0.05）；高龄、基础低LVEF是其他原因死亡的独立危险因素（P<0.05）。结论　对于慢性心衰患者，死亡原因主要为非心源性死亡；综合管理能够降低全因死亡，主要为降低感染性病死率。应重视对慢性心衰患者全方位综合管理以降低病死率。

关键词: 心力衰竭, 疾病管理, 全科医生, 死亡率, 影响因素分析