Introdução / Introduction

Análise de dados dicotômicos / Analysis of dichotomic data

Há 36 sujeitos em cada grupo. There are 36 subjects in each group.

Náusea / Nausea

  • Sabe-se pela tabela 4 que, respectivamente nenhum paciente apresentou náusea no grupo unilateral e oito pacientes apresentaram no grupo Bilateral.

  • We know from table 4 that that, respectively no patient suffered nausea from the unilateral group and that 8 suffered nausea in bilateral group.

# A tabela deve parecer com essa:
# The table should look like this:
matrix(c(0,8,36,36-8),nrow=2)
##      [,1] [,2]
## [1,]    0   36
## [2,]    8   28
  • A presença de células com valor menor que 5 indica que o teste do qui-quadrado apresentará viés importante logo o teste exato de Fisher é melhor.

  • Cells with values less than 5 indicate that Fisher’s Exact test will be less biased than Chi-squared test.

fisher.test(matrix(c(0,8,36,28),nrow=2))
## 
##  Fisher's Exact Test for Count Data
## 
## data:  matrix(c(0, 8, 36, 28), nrow = 2)
## p-value = 0.005056
## alternative hypothesis: true odds ratio is not equal to 1
## 95 percent confidence interval:
##  0.0000 0.5098
## sample estimates:
## odds ratio 
##          0

Dor de cabeça / Headaches

  • Sabe-se pela tabela 4 que, respectivamente 2 pacientes apresentaram cefaleia no grupo unilateral e oito pacientes apresentaram no grupo Bilateral.

  • We know from table 4 that that, respectively two patients suffered headaches from the unilateral group and that 8 suffered headaches in bilateral group.

# A tabela deve parecer com essa:
# The table should look like this:
matrix(c(2,8,34,28),nrow=2)
##      [,1] [,2]
## [1,]    2   34
## [2,]    8   28

Portanto (therefore):

fisher.test(matrix(c(2,8,34,28),nrow=2))
## 
##  Fisher's Exact Test for Count Data
## 
## data:  matrix(c(2, 8, 34, 28), nrow = 2)
## p-value = 0.0847
## alternative hypothesis: true odds ratio is not equal to 1
## 95 percent confidence interval:
##  0.02018 1.17138
## sample estimates:
## odds ratio 
##     0.2101

Hipotensão / Hypotension

  • Sabe-se pela tabela 4 que, respectivamente nenhum paciente apresentou hipotensão no grupo unilateral e 6 pacientes apresentaram no grupo Bilateral.

  • We know from table 4 that that, respectively no patient suffered hypotension from the unilateral group and that 6 suffered hypotension in bilateral group.

# A tabela deve parecer com essa:
# The table should look like this:
matrix(c(0,6,36,36-6),nrow=2)
##      [,1] [,2]
## [1,]    0   36
## [2,]    6   30

Portanto (therefore):

fisher.test(matrix(c(0,6,36,36-6),nrow=2))
## 
##  Fisher's Exact Test for Count Data
## 
## data:  matrix(c(0, 6, 36, 36 - 6), nrow = 2)
## p-value = 0.02493
## alternative hypothesis: true odds ratio is not equal to 1
## 95 percent confidence interval:
##  0.0000 0.7819
## sample estimates:
## odds ratio 
##          0

Bradicardia / Bradicardia

  • Sabe-se pela tabela 4 que, respectivamente nenhum paciente apresentou bradicardia no grupo unilateral e 5 pacientes apresentaram no grupo Bilateral.

  • We know from table 4 that that, respectively no patient suffered bradicardia from the unilateral group and that 5 suffered bradicardia in bilateral group.

# A tabela deve parecer com essa:
# The table should look like this:
matrix(c(0,5,36,36-5),nrow=2)
##      [,1] [,2]
## [1,]    0   36
## [2,]    5   31

Portanto (therefore):

fisher.test(matrix(c(0,5,36,36-5),nrow=2))
## 
##  Fisher's Exact Test for Count Data
## 
## data:  matrix(c(0, 5, 36, 36 - 5), nrow = 2)
## p-value = 0.05389
## alternative hypothesis: true odds ratio is not equal to 1
## 95 percent confidence interval:
##  0.00 1.03
## sample estimates:
## odds ratio 
##          0