Epidemiological Cross tabulation (2 by 2)

1.1
Other Commands How to...

Epitables outcome exposure [variable3 ...][/T or /Table][/Q or /Quiet]
ETab

The EPITABLES command does a cross-tabulation of the specified variables (Outcome and Exposure). Values of the first variable (Outcome) will appear across the top of the table, and those of the second variable (exposure) will be on the left margin of the table. Normally cells contain counts of records (and percentages) matching the values in the corresponding marginal labels.

The two first variables are mandatory, the rest, if any (until a total maximum of 3 variables), are used as the basis to create strata. EpiData Analysis will produce an independent table for each value in as many strata as all possible combinations.
Outcome and Exposure must be dichotomous variables (Yes/No or with only two different values).
Epitables will calculate odd ratios and risk ratios

Output:


Ill?
Sex:	Y	%	N	%	Total	%
M	16	{51.6}	15	{48.4}	31	{100.0}
F	30	{68.2}	14	{31.8}	44	{100.0}
Total	46	{61.3}	29	{38.7}	75	{100.0}
Outcome: ILL=Y
Exposure: SEX=M
Chi2= 2.105 p= 0.147
Fisher exact: p= 0.1588
Crude OR = 0.498 (95% CI: 0.17 - 1.44)
MLE: OR= 0.503 (Exact 95% CI: 0.17 - 1.43)
p (MLE OR <= 0.50 if population OR = 1.0)= 0.1132
RR= 0.757 (95% CI: 0.51 - 1.12)

The number (count) of records matching each category is given in each cell. Optionally the row and column percents can be included.
Impact and asociation measures are displayed beside p values.
The real output can be different depending on the options used (see below)


Examples and Hints
EPITABLES SEX ILL                      (A cross tabulation of SEX by ILL)
EPITABLES FOOD ILL SEX                 (A cross tabulation of FOOD by ILL stratified by SEX)
Notice:
No more than five variables are allowed
EPITABLES and ETAB are equivalent
If Outcome and Exposure are not boolean variables (yes/no) use Set table rotate if necesary (see below)

Related Commands:
Tables
Options:
/T or /Tables
/Q or /Quiet
If stratified analysis, tables for each of the different strata are shown, otherwise only a summary will be displayed
Do not show tablas at all
Output variables:
$Chi#
$pChi#
$Fisher#
$OR#
$ORCiL#
$ORCiH#
$MleOR#
$MleCiL#
$MleCiH#
$MLEp#
$RR#
$RRCiL#
$RRCiH#
Chi square value for table # (1 por first table an so on)
p value corresponding to the Chi square value in table #
p value for Fisher exact test in table # (1 for first table and so on)
OR: Odds Ratio (point estimator) in table # (1 for first table and so on)
Lower limit of the 95% IC of the OR in table # (1 for first table and so on)
Higher limit of the 95% IC of the OR in table # (1 for first table and so on)
MLE (Maximum Likelihood Estimator) in table # (1 for first table and so on)
Lower limit of the 95% IC of the MLE in table # (1 for first table and so on)
Higher limit of the 95% IC of the MLE in table # (1 for first table and so on)
p value of MLE OR <=0.5 if population OR=1 in table # (1 for first table and son on)
RR: Risk Ratio (point estimator) in table # (1 for first table and so on)
Lower limit of the 95% IC of the RR in table # (1 for first table and so on)
Higher limit of the 95% IC of the RR in table # (1 for first table and so on)
Output definition:
set table design =line[box][filled][system]
set table design summary =line[box][filled][system]
set table definition
set table rotate
Main tables design will be acording with this command. The default value is line
Summary table design will be acording with this command. The default value is line
See the set table definition help for more information about Column and Row percents
Will transpose variables such that the highest value of outcome and exposure are shown in the top left 2X2 cell
References:
No specific references
Top