基于倾向性评分的中医复杂干预临床疗效评价

摘要

The success of clinical research, depends on the ability to control the bias. In the non-randomized observational study, propensity score can reduce the contrast differences between groups and narrow bias. Compared with design stage through matching control bias, propehsity analysis frees from experimental design (randomized controlled trial design) approach limits, with no need for randomization in the design stage. The result is closer to "real world" aitual effectiveness rather than the clinical trial results. Therefore, propehsity analysis applying to personalized medicine treatment efficacy evaluation of complex interventions is appropriate, has an important prospects and use value.%临床研究的成功与否,与是否有效控制偏倚有关.在非随机化的观察性研究中,倾向性评分能减少对比组间的差异,缩小偏倚.与试验设计阶段进行配比控制偏倚相比,倾向性分析法不受试验设计(随机对照试验设计)方法的限制,无须在试验设计阶段进行随机,其结果更接近“真实世界”的实际干预效果而非临床试验的效果.因此,倾向性分析应用于中医个性化治疗的复杂干预疗效评价是适合的,具有重要的研究前景和使用价值.