论文部分内容阅读
目的分析临床试验中不同区域间疗效的方向性差异的影响因素,为设计更合理的临床试验提供建议。方法从试验设计角度分别在假设事实疗效一致与不一致的条件下,采用数学推导来推算观察到不一致临床试验结果的概率,用求导及SAS编程分析其每一个随机因素对不一致概率的影响,并分析该如何控制这些影响因素。结果不同区域间疗效的方向性差异的影响因素主要有试验样本量、各区域标准化事实疗效、各区域人数比例和区域数目。通过在原地区增加样本量、增加疗效绝对值小的区域人数比例和减少区域数目等方法可减小随机因素的影响,使得临床试验中观察到的结果更接近于真实状况。结论优化样本量、区域人数比例和区域数目等设计因素,可降低临床试验中随机因素造成的疗效不一致。
Objective To analyze the influencing factors of the differences in the direction of efficacy between different regions in clinical trials and provide suggestions for designing more rational clinical trials. Methods From the perspective of experimental design, the probability of inconsistent clinical trial results was estimated using mathematical derivation under the assumption of consistent and inconsistent clinical efficacy. The influence of each random factor on the probability of inconsistency was analyzed by using derivative and SAS programming. Analysis of how to control these factors. Results The main influencing factors of the directional differences of efficacy between different regions were the sample size of the trial, the actual effect of standardization in each region, the proportion of people in each region, and the number of regions. By increasing the sample size in the original area, increasing the proportion of patients with small absolute value of effect and reducing the number of areas, the influence of random factors can be reduced, and the results observed in clinical trials are closer to the true ones. Conclusion The design factors such as the optimal sample size, regional population proportion and the number of regions can reduce the inconsistent curative effect caused by random factors in clinical trials.