基于COM-B与CFIR整合模型的社区家庭医生心脏康复服务实施行为形成机制结构方程模型研究

A Structural Equation Model Study on the Mechanisms Underlying the Implementation of Cardiac Rehabilitation Services by Community Family Physicians Based on an Integrated COM-B and CFIR Model

摘要: 目的：探讨社区家庭医生心脏康复(cardiac rehabilitation, CR)服务实施行为的影响因素及作用机制，为促进基层CR资源向服务行为转化提供依据。方法：2026年采用横断面调查法，对上海市浦东新区社区卫生服务中心家庭医生进行问卷调查，获得有效问卷334份。基于COM-B行为系统理论和实施研究综合框架(CFIR)构建理论模型，经文献分析、临床实践及15名专家咨询形成调查工具，内容包括CR知识、实施自我效能、实施能力、实施气候、实施动机、实施行为和机构CR建设水平。采用Cronbach’s α系数、验证性因子分析评价量表信效度，采用结构方程模型和Bootstrap法检验作用路径及中介效应。结果：调查对象年龄为(39.61 ± 7.13)岁，男性占42.81%，49.70%接受过CR相关培训。各潜变量Cronbach’s α系数和组合信度均为0.936~0.964，平均方差提取量为0.711~0.795；测量模型拟合良好(χ²/df = 1.131, CFI = 0.995, TLI = 0.994, RMSEA = 0.020, SRMR = 0.028)。CR知识正向影响实施自我效能(β = 0.429, P < 0.001)；实施自我效能(β = 0.421, P < 0.001)和实施气候(β = 0.379, P < 0.001)正向影响实施能力；实施能力(β = 0.310, P < 0.001)、实施动机(β = 0.130, P = 0.013)及机构CR建设水平(β = 0.305, P < 0.001)正向影响实施行为。模型分别解释实施自我效能、实施能力和实施行为18.39%、32.08%和21.48%的变异。CR知识通过“实施自我效能→实施能力→实施行为”产生链式间接效应(效应值 = 0.0098，95% CI: 0.0053~0.0144，P < 0.001)；实施自我效能和实施气候亦分别通过实施能力间接影响实施行为。结论：社区家庭医生CR实施行为受知识、自我效能、实施能力、组织实施气候、实施动机及机构建设水平等多层因素共同影响，实施能力是连接个体心理因素、组织环境与服务行为的重要中介。基层CR推广应采取“知识提升–能力形成–组织支持–行为转化”的综合实施策略。

Abstract: Objective: To identify factors influencing cardiac rehabilitation (CR) implementation behavior among community family physicians and explore the underlying mechanisms, thereby informing the translation of primary care CR resources into routine service delivery. Methods: A cross-sectional survey was conducted in 2026 among family physicians at community health service centers in Pudong New Area, Shanghai, yielding 334 valid responses. A theoretical model was developed based on the Capability-Opportunity-Motivation-Behavior (COM-B) system and the Consolidated Framework for Implementation Research (CFIR). The questionnaire, informed by the literature, clinical practice, and consultation with 15 experts, assessed CR knowledge, implementation self-efficacy, capability, climate, motivation, behavior, and institutional CR development. Cronbach’s α and confirmatory factor analysis were used to evaluate reliability and validity. Structural equation modeling and bootstrap analysis were performed to test hypothesized pathways and indirect effects. Results: Participants had a mean age of 39.61 ± 7.13 years; 42.81% were men, and 49.70% had received CR-related training. Cronbach’s α and composite reliability ranged from 0.936 to 0.964, and the average variance extracted ranged from 0.711 to 0.795. The measurement model demonstrated good fit (χ²/df = 1.131, CFI = 0.995, TLI = 0.994, RMSEA = 0.020, and SRMR = 0.028). CR knowledge positively predicted implementation self-efficacy (β = 0.429, P < 0.001). Self-efficacy (β = 0.421, P < 0.001) and implementation climate (β = 0.379, P < 0.001) positively predicted implementation capability. Implementation behavior was positively predicted by capability (β = 0.310, P < 0.001), motivation (β = 0.130, P = 0.013), and institutional CR development (β = 0.305, P < 0.001). The model explained 18.39%, 32.08%, and 21.48% of the variance in self-efficacy, capability, and implementation behavior, respectively. CR knowledge exerted a significant sequential indirect effect through self-efficacy and capability (effect = 0.0098, 95% CI: 0.0053~0.0144, P < 0.001). Self-efficacy and implementation climate also indirectly affected implementation behavior through capability. Conclusion: CR implementation behavior among community family physicians is jointly shaped by multilevel factors, including knowledge, self-efficacy, capability, implementation climate, motivation, and institutional development. Implementation capability is a key mediator linking individual psychological factors and organizational context to service delivery. Primary care CR promotion should therefore adopt an integrated strategy encompassing knowledge enhancement, capability building, organizational support, and behavioral translation.