# epiR epi.2by2 command with cohort.count versus case.control

I am running a stratified analysis in epi.2by2. I need help in understanding why the cohort.count option works but the case.control does not (as shown below). Any help is greatly appreciated!

# compute odds ratio adjusted for sex

alltable <- table(prob1\$coffeedrinker, prob1\$case, prob1\$sex, dnn = c("Coffee", "Case", "Sex"))
alltable
, , Sex = 1

``````  Case
``````

Coffee 1 0
1 140 280
0 11 56

, , Sex = 0

``````  Case
``````

Coffee 1 0
1 207 275
0 9 32

adjtable <- epi.2by2(alltable, method = "cohort.count", digits = 2, conf.level = 0.95, units = 100,

• ``````                 interpret = FALSE, outcome = "as.columns")
``````

print(adjtable)
Outcome + Outcome - Total Inc risk *
Exposed + 347 555 902 38.47 (35.28 to 41.73)
Exposed - 20 88 108 18.52 (11.69 to 27.14)
Total 367 643 1010 36.34 (33.36 to 39.39)

## Inc risk ratio (crude) 2.08 (1.39, 3.11) Inc risk ratio (M-H) 2.00 (1.33, 2.99) Inc risk ratio (crude:M-H) 1.04 Odds ratio (crude) 2.75 (1.66, 4.55) Odds ratio (M-H) 2.60 (1.57, 4.32) Odds ratio (crude:M-H) 1.06 Attrib risk in the exposed (crude) * 19.95 (11.97, 27.94) Attrib risk in the exposed (M-H) * 18.53 (-0.68, 37.73) Attrib risk (crude:M-H) 1.08

M-H test of homogeneity of IRRs: chi2(1) = 0.008 Pr>chi2 = 0.928
M-H test of homogeneity of ORs: chi2(1) = 0.008 Pr>chi2 = 0.928
Test that M-H adjusted OR = 1: chi2(1) = 14.540 Pr>chi2 = <0.001
Wald confidence limits
M-H: Mantel-Haenszel; CI: confidence interval

• Outcomes per 100 population units

adjtablecc <- epi.2by2(alltable, method = "case.control", digits = 2, conf.level = 0.95, units = 100,

• ``````                 interpret = FALSE, outcome = "as.columns")
``````

Error in `.rowNamesDF<-`(x, value = value) : invalid 'row.names' length
In addition: Warning message:
In cbind(c1, c2, c3, c4) :
number of rows of result is not a multiple of vector length (arg 1)

This topic was automatically closed 42 days after the last reply. New replies are no longer allowed.

If you have a query related to it or one of the replies, start a new topic and refer back with a link.