20/06/20 19:20:16.10 tzFG6qz7.net
公表された抗体陽性割合
8/7950 # 厚労省
7/1000 # 東大
191/44066 # ソフトバンク
一番高い0.7％を採用して
>今年１０月に数百人規模で治験
を実薬500人、偽薬500人として
偽薬の500人中4人が陽性、ワクチン投与で500人中0人にできたとすると
> prop.test(c(4,0),c(500,500))
2-sample test for equality of proportions with continuity correction
data: c(4, 0) out of c(500, 500)
X-squared = 2.259, df = 1, p-value = 0.1328
alternative hypothesis: two.sided
95 percent confidence interval:
-0.001808434 0.017808434
sample estimates:
prop 1 prop 2
0.008 0.000
Warning message:
In prop.test(c(4, 0), c(500, 500)) :
Chi-squared approximation may be incorrect
> Fisher.test(c(4,0),c(500,500))
Fisher's Exact Test for Count Data
data: cbind(hit, shot - hit)
p-value = 0.1242
alternative hypothesis: true odds ratio is not equal to 1
95 percent confidence interval:
0.6617083 Inf
sample estimates:
odds ratio
Inf
> poisson.test(c(4,0),c(500,500))
Comparison of Poisson rates
data: c(4, 0) time base: c(500, 500)
count1 = 4, expected count1 = 2, p-value = 0.125
alternative hypothesis: true rate ratio is not equal to 1
95 percent confidence interval:
0.660124 Inf
sample estimates:
rate ratio
Inf
どれでも、有意差がでない。
どういうデザインで治験するんだろうなぁ？