目标效度视角下随机对照试验与真实世界研究证据整合的Meta分析方法OA

Randomized controlled trials(RCTs)and real-world studies(RWSs)provide complementary forms of evidence.The rigorous integration of these two evidence sources has become a key issue in health technology assessment(HTA)and value-based healthcare decision-making.From the perspective of"target validity,"this review classifies existing meta-analytic approaches into three categories:descriptive synthesis,internal bias adjustment,and target bias adjustment,with the latter two collectively referred to as quantitative bias adjustment methods.The strengths and limitations of these approaches in addressing internal and external validity are systematically evaluated.This review indicates that although methods such as quality weighting provide useful tools for quantifying bias,their robustness remains limited by the subjectivity of prior assumptions and the lack of a standardized framework for assessing external bias.While quantitative bias adjustment methods offer substantial methodological flexibility,their performance is likewise constrained by subjective bias assessment and the lack of standardized criteria for evaluating external validity.Future evidence integration studies should adhere to the principles of transparent reporting,multidimensional evaluation,and sensitivity analysis.In addition,efforts should be made to standardize tools for assessing external bias,thereby providing more targeted scientific evidence to support value-based healthcare security and health policy decision-making.