analysis / Tool 6.2
Example of analysis performed using EpiData Analysis
The EpiData Analysis programme (EpiDataStat) can perform a range of analyses on questionnaire datasets. Here we’ll give a brief guide to a few descriptive and analytical analyses which can quickly be made – based on a simple dataset; the one that is also presented under Tool 5. The output that EpiData Analysis can make can for instance be used to insert in an outbreak situation report (as exemplified by the links in this document to an example of a report template.
You will need to download the most recent version of EpiDataStat ( Please note, however, that the version provided is still under development. It may therefore be advisable to check if new versions of the programme have been released. This can be done through the Help menu of the programme or by consulting the EpiData (
The dataset used here is called SalmoCaCoFruits.epx. There are 40 persons entered in the dataset and it focuses on exposure data regarding fruits. Download this file and paste it in the ’Analysis’ folder of your programme.
1. Descriptive analyses of data
Open EpiDataStat and open the dataset.
1.1 View the data
To get an overview of the data “browse” and choose the variables you want to see. Choose “all” to see the whole dataset.
The browse output:
The basic data analysis for a short outbreak report is describing time, place and person. If a risk factor study has been conducted, also analysis of attack rates and relative risks (or odds ratio) could be included. How this is done in epidata is shortly described below.
1.2 Describe the number of cases
It is easy to create frequency tables describing for example the total number of cases.
The frequency tables are generated as shown below and you can choose one or more variables to be shown:
The outcome of a frequency on cases:
In a similar way, a table with frequency of cases by country can be made and copied into areport template.
1.3 Creating an epidemic curve
In this version of EpiData you can create a very nice epicurve to describe the time of onset of illness among your cases:
There is a range of possibilities to edit the lay-out of the epicurve. This is done by selecting “Edit graph” in the “Graph/Axis” section:
In the editing window you can change colours, titles and many other things to improve your epicurve:
Example of an epicurve output:
The epicurve can then be copied into the report template.
1.4 Age and gender distribution
As you can see in the screen print below it is possible to describe your data e.g. age and gender using different types of graphs, plot and figures. The example given below is creation of a histogram:
Outcome of the histogram when choosing sex by case status:
Age and sex distribution graph copied into the report template.
Similar graphs and tables on symptoms and outcome should be made if relevant and presented in the report template.
1.5 Generating 2 x 2 tables
When you want to generate 2 x 2 tables you use the “Tables” option in the analysis step and enter one variable for the row and one for the column:
An example of the 2 x 2 table output is shown here:
2. Analytical data analyses
2.1 Generating attack rates
If the data is from a closed cohort it can be useful for you to get an overview of the attack rates and relative risks for different types of exposures. This is done by using the “Tables” option and selecting the Table Type “Compact”:
This creates a very nice output of your exposures as seen in the example below (although the example data used are case control data and thus should not be analysed this way!):
The risk factor table is placed here in the report template.
2.2 Generating Odds ratios
Another option is to generate odds ratios which you can also choose from the “Tables” option:
The output of the odds ratio analysis is shown below:
ECDC toolbox for FWD outbreak investigations
- 1 -