文献速递医学教育与科研方法流行病学与生物统计

系统综述与现实主义综述结合:混合综述设计分类与方法整合策略的范围综述

When systematic review meets realist review: proposing a hybrid review design to understand complex intervention effectiveness and mechanisms.
PubMed 文献 · J Clin Epidemiol · 2026-03-25

该范围综述检索2005年1月至2025年5月五个数据库,纳入129项明确整合系统综述与现实主义综述的混合综述,归纳为4种设计类型与8种方法整合策略,并总结其优势与挑战,为方法学指南制定提供基础证据。

要点

  • 混合综述可分为分离式、结合式、现实主义知情系统综述和多阶段混合综述4类。
  • 分离式与结合式中共识别8种方法整合策略,最常见为用系统综述启发现实主义综述的初始程序理论。
  • 报告优势包括证据更强、见解可操作、助力复杂干预与理论开发及流程精简;挑战包括哲学与方法学张力及劳动强度大。

核心结论

  • 该范围综述绘制了混合综述的基础地图与分类,阐明系统综述与现实主义综述方法在实践中如何整合。
  • 研究结果为方法学指南制定奠定证据基础,并支持研究者设计此类综述。

研究设计与样本

  • 采用遵循Joanna Briggs Institute方法学的范围综述。
  • 检索PubMed、Web of Science、Embase、PsycINFO和CINAHL数据库2005年1月至2025年5月文献,并辅以引文检索。
  • 纳入标准为明确以协调方式整合系统综述与现实主义综述的混合综述,共纳入129项研究。
  • 研究特征与方法学特征采用描述性分析,整合策略及报告的优势与挑战采用定性内容分析合成。

主要发现

  • 129项纳入研究中,分离式系统综述与现实主义综述(SR|RR)48项(37.2%)。
  • 结合式系统综述与现实主义综述(SR+RR)15项(11.6%)。
  • 现实主义知情系统综述47项(36.4%),多阶段混合综述19项(14.7%)。
  • SR|RR类型中最常见的整合策略为「用系统综述启发现实主义综述的初始程序理论开发」,48项中27项(56.3%)。
  • SR+RR类型中最常见的整合策略为「共享检索策略」,15项中11项(73.3%)。

关键数值与推荐

  • 混合综述设计分为4类:SR|RR、SR+RR、现实主义知情系统综述、多阶段混合综述。
  • 在SR|RR与SR+RR类型中共识别8种方法整合策略。
  • 报告的优势包括证据更强、见解可操作、助力复杂干预/理论开发、流程精简。
  • 报告的挑战包括哲学/方法学张力与劳动强度大。

临床怎么用

  • 研究者设计混合综述时可参照4类设计分类选择协调整合系统综述与现实主义综述的方式。
  • 开展SR|RR或SR+RR时可参考8种方法整合策略,如用系统综述启发初始程序理论或共享检索策略。
  • 研究结果可帮助期刊与资助方为此类创新综述设计制定报告标准。

局限性

  • 原文未报告局限性。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

BACKGROUND: Systematic reviews (SRs) and realist reviews (RRs) are increasingly conducted in a coordinated manner with intersecting methodological processes, forming what we term hybrid review designs. However, little is known about how these hybrid reviews are conducted and structured. OBJECTIVES: This study aimed to characterize existing hybrid reviews by examining their features, analyzing their method intersection strategies, and synthesizing the reported advantages and challenges, thereby informing future methodological guidance. METHODS: A scoping review following the Joanna Briggs Institute approach was conducted. We searched the following databases from January 2005 to May 2025: PubMed, Web of Science (Clarivate), Embase (Ovid), PsycINFO (Ovid), and CINAHL (EBSCOhost), supplemented by citation searches. Eligible studies were hybrid reviews that explicitly integrated SRs and RRs in a coordinated manner. Data on study characteristics and methodological features were analyzed descriptively, while intersection strategies and the reported advantages and challenges were synthesized using qualitative content analysis. RESULTS: Of 129 included studies, we classified hybrid review designs into 4 types based on methodological features: SR and RR in separation (SR|RR, 48 of 129, 37.2%); SR and RR in combination (SR + RR, 15 of 129, 11.6%); realist-informed SR (47 of 129, 36.4%); and multiphase hybrid review (19 of 129, 14.7%). Within the SR|RR and SR + RR types, 8 method intersection strategies were identified, with the most common being "using SR to inform the initial program theory development of RR" in SR|RR (27 of 48, 56.3%) and "sharing search strategies" in SR + RR (11 of 15, 73.3%). Reported advantages included stronger evidence, actionable insights, aiding complex intervention/theory development, and streamlined processes; challenges included philosophical/methodological tensions and labor intensity. CONCLUSION: This review provides a foundational map and classification of hybrid reviews. The findings illuminate how SR and RR methods are integrated in practice, establishing an evidence base to inform methodological guidance and supporting researchers in designing such reviews. PLAIN LANGUAGE SUMMARY: This study looked at how 2 types of research reviews-systematic reviews, which focus on whether interventions work, and realist reviews, which focus on how and why interventions work-can be conducted in a coordinated manner. We reviewed 129 published hybrid reviews and identified 4 types of integration, as well as 8 common method strategies for linking the 2 approaches. We also summarized the reported benefits and challenges of these hybrid reviews. Our findings show that hybrid reviews have the potential to provide a more comprehensive understanding of interventions by combining evidence of effectiveness with insights into underlying mechanisms. This information can help researchers plan future reviews that better inform policy and practice and guide journals and funders in setting reporting standards for these innovative review designs.

来源:J Clin Epidemiol,查看原文。版权归原作者所有。

引用

  1. Meng M, Zhou Y, Han S, et al. When systematic review meets realist review: proposing a hybrid review design to understand complex intervention effectiveness and mechanisms. J Clin Epidemiol. 2026. PMID: 41895671. doi:10.1016/j.jclinepi.2026.112253

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。