RCT/队列研究+分类变量
2017-10-23 来源: 医咖会 作者: 张耀文

一、问题与数据

NEJM上的一个研究[1]在Protocol中是这样描述如何计算样本量的: The power for the primary endpoint weight change is calculated based on a two sided t-test with a significance level of 5%. The power with regard to the co-primary dichotomous endpoints proportion of subjects with a weight loss larger than of at least 5% and or more than 10%, respectively, is calculated based on a two-sided chi-square test. With a sample size of 2400 subjects treated with liraglutide and 1200 subjects treated with liraglutide placebo, the trial will have more than 90% power to detect a difference between liraglutide and liraglutide placebo in the proportion of subjects with a weight loss greater than 10%, given that the probabilities to achieve this weight loss is 10% for liraglutide placebo and 14% for liraglutide. 大致意思是: 比较主要结局(体重变化)时,按照P=0.05进行双侧t检验。对于另一个主要结局(二分类变量)——体重下降5%及以上、10%以上的人数比例,采用双侧卡方检验比较。假设对照组体重下降10%以上的人数比例为10%,利拉鲁肽组的这个比例为14%。当利拉鲁肽组和对照组的样本量分别为2400例、1200例时,可以有超过90%的把握度发现这种差异。 以这个研究为例,如何使用PASS软件来计算样本量呢。 参考文献: 1. N Engl J Med. 2015 Jul 2;373(1):11-22.