Adjusting for multiple comparison following fisher's exact test and t-test

[Madu Abuchi]

Hi Stata and Stats experts,

I am conducting analyses between two control groups and one treatment in 1:1:1 parallel trial. Say I have C1, C2 and T for the two control and one treatment. So my comparison was C1 vs T, C2 vs T and C1 vs C2 on a vaccine with 10 serotypes. I conducted fishers exact test for proportions reaching a cutoff point of 2microgram/dl and t-test for geometric mean concentrations (GMC) of the comparison pairs.

An example of one of my comparison C1 vs T yielded for ratio of GMC following t-test, the following table

	C1 vs T
Serotypes	Ratio	95%CI	p
A	9.99	(8.52, 16.87)	<0.001
B	9.02	(5.32, 13.47)	<0.001
C	5.93	(3.48, 6.98)	<0.001
D	0.99	(0.65, 1.54)	0.69
E	7.56	(4.73, 10.25)	<0.001
F	5.88	(3.34, 7.12)	<0.001
G	1.96	(1.24, 2.93)	0.0076
H	2.51	(1.76, 3.89)	<0.001
I	1.10	(0.85, 1.72)	0.35
J	1.25	(0.95, 1.80)	0.305

Questions
1. Does it make any sense controlling for multiple comparison? I did ask one of my lecturers in biostatistics and he responded that some do and some don't, the response wasn't very clear.
2. Is there any stata code that does bonferonni test after t-test or fishers exact test?
3. I coded the below based on formular for bonferroni

adjusted p-value = min(k * raw p-value, 1)

. The study participants were measured at 4 time point (month 1, 3, 6 and 9). I am a bit unsure if to put into consideration number of measurements, and number of groups in my k-value. However, for the illustration below I have used k=10 based on number of serotypes in the vaccine shot.

my code below is an illustrated p values of C1 vs T, C2 vs T and C1 vs C2 comparison groups

clear
input str6 serotypes str6 p_1 str6 p_2 str6 p_3
A    0.001    0.001    0.001
B    0.001    0.001    0.001
C    0.001    0.001    0.001
D    0.69    0.69    0.69
E    0.001    0.001    0.001
F    0.001    0.001    0.001
G    0.0076    0.0076    0.0076
H    0.001    0.001    0.001
I    0.35    0.35    0.35
J    0.305    0.305    0.305
end
gen k = 10
destring p_1 p_2 p_3, replace
foreach raw_p of any p_1 p_2 p_3 {
    gen adj_`raw_p' = min(k*`raw_p',1)
}

I will appreciate useful comments and contributions to my questions.

Regards