有帖子惊呼说辉瑞最新发布的报告里疫苗有效率的计算只包括了10个样本。我来科普一下这个计算。这里不讨论到底有没有效，打还是不打。单纯的就是科普一下这个计算。 我就解释一下80% 的有效率是怎么来的吧。文章里面原话： Preliminary descriptive efficacy analyses of COVID-19 cases occurring at least 7 days post- Dose 3 included 376 BNT162b2 recipients and 179 placebo recipients 6-23 months of age and 589 BNT162 recipients and 271 placebo recipients 2-4 years of age. In these analyses, three COVID-19 cases occurred in participants 6-23 months of age, with 1 COVID-19 case in the BNT162b2 group compared to 2 in the placebo group, corresponding to an estimated VE of 75.6% (95% CI: −369.1%, 99.6%), and 7 COVID-19 cases occurred in participants 2-4 years of age, with 2 cases in the BNT162b2 group and 5 in the placebo group, corresponding to an estimated VE of 82.4% (95% CI: −7.6%, 98.3%). In a combined analysis of both age groups, VE was 80.4% (95% CI: 14.1%, 96.7%) with 3 cases in the BNT162b2 group and 7 cases in the placebo group. Cases were accrued during February 2022 through April 2022, when the Omicron variant was prevalent in the United States. Among all COVID-19 cases accrued from Dose 1 through the data cutoff of April 29, 2022, 1 placebo recipient 6-23 months of age and 7 participants 2-4 years of age (6 BNT162b2 recipients and 1 placebo recipient) met the protocol- specified criteria for severe COVID-19 during both blinded and open-label follow-up. Only one of these severe COVID-19 cases (in a BNT162b2 recipient 99 days post-Dose 2) resulted in hospitalization, and the remainder met criteria for severe COVID-19 based on vital sign findings that were not clinically significant in the opinion of the investigator and FDA. 实验组总共965人，对照组总共450 人。 实验组有3个cases， 就是有3人得了covid。 所以实验组的中奖率是3/965 ，约为0.31%吧。 对照组有7人中奖，所以中奖率是7/450， 约为1.56%吧。 那么实验组的中奖率比对照组低了1.25%，因为1.56-0.31 =1.25呀。 那么相对于对照组，就得到疫苗的有效率了： 1.25/1.56， 约80%。 我这里是约，因为统计软件计算时，小数位更精确。 帖子说80%只考虑了10个case是不正确的。 上面的计算还包括了所有covid 阴性的人，总共1415人。
我就解释一下80% 的有效率是怎么来的吧。文章里面原话：
Preliminary descriptive efficacy analyses of COVID-19 cases occurring at least 7 days post- Dose 3 included 376 BNT162b2 recipients and 179 placebo recipients 6-23 months of age and 589 BNT162 recipients and 271 placebo recipients 2-4 years of age. In these analyses, three COVID-19 cases occurred in participants 6-23 months of age, with 1 COVID-19 case in the BNT162b2 group compared to 2 in the placebo group, corresponding to an estimated VE of 75.6% (95% CI: −369.1%, 99.6%), and 7 COVID-19 cases occurred in participants 2-4 years of age, with 2 cases in the BNT162b2 group and 5 in the placebo group, corresponding to an estimated VE of 82.4% (95% CI: −7.6%, 98.3%). In a combined analysis of both age groups, VE was 80.4% (95% CI: 14.1%, 96.7%) with 3 cases in the BNT162b2 group and 7 cases in the placebo group. Cases were accrued during February 2022 through April 2022, when the Omicron variant was prevalent in the United States. Among all COVID-19 cases accrued from Dose 1 through the data cutoff of April 29, 2022, 1 placebo recipient 6-23 months of age and 7 participants 2-4 years of age (6 BNT162b2 recipients and 1 placebo recipient) met the protocol- specified criteria for severe COVID-19 during both blinded and open-label follow-up. Only one of these severe COVID-19 cases (in a BNT162b2 recipient 99 days post-Dose 2) resulted in hospitalization, and the remainder met criteria for severe COVID-19 based on vital sign findings that were not clinically significant in the opinion of the investigator and FDA.
实验组总共965人，对照组总共450 人。 实验组有3个cases， 就是有3人得了covid。 所以实验组的中奖率是3/965 ，约为0.31%吧。 对照组有7人中奖，所以中奖率是7/450， 约为1.56%吧。 那么实验组的中奖率比对照组低了1.25%，因为1.56-0.31 =1.25呀。 那么相对于对照组，就得到疫苗的有效率了： 1.25/1.56， 约80%。 我这里是约，因为统计软件计算时，小数位更精确。
帖子说80%只考虑了10个case是不正确的。 上面的计算还包括了所有covid 阴性的人，总共1415人。
哈哈。是的。一些人临床实验和统计里面的一些专业术语都没搞清楚，就敢评论别人的统计报告
也许是正骨科的博士？
这个科普简明易懂，赞。