Following the discussion on the significance of effect sizes, and selecting suitable data entry formats, this next step is to understand different software functionalities through appropriate case studies. The present article introduces the procedures of manual data entry in a spreadsheet through a case study.
Case study 1: cancer risk associated with cholecystitis
One of the leading causative factors of gallbladder carcinoma is gallstones. A number of case-control studies and prospective studies have explored and correlated the presence of gallstones with gallbladder carcinoma. Therefore the present meta-analysis review aims to ascertain the overall extent of carcinoma risk associated with gallstones. The present case study deals with one group, dichotomous data.
Inclusion and Exclusion Criteria
The table below shows the criteria for inclusion and exclusion of relevant published studies in the present case study.
Data collection procedure
For the purpose of completing the systematic review for the present case study, the NCBI and EMBASE databases were searched, using the keywords:
- gallbladder cancer,
- gallbladder stones,
- gallstones and
- gall bladder carcinoma.
The data extraction procedures lead to the collection of two types of data:
- Data pertaining to the number of events (dichotomous data) that is, number of patients exhibiting gallstone associated gallbladder cancer.
- Data pertaining to the odd ratio (dichotomous data) that is, odds of patients having gallbladder carcinoma due to gallstones. The subsequent article on Importing and Loading Data illustrates the same.
This article employees the number of events data. The number of events represents the favorable number of cases out of the total sample size. In the present study, it is the number of patients having gallstones and presenting gallbladder cancer.
Manual data entry procedure
The following steps will guide in entering the data for the number of events, to eventually meta-analyze the data from different studies.
Step 1: Inserting column for study data
The first column essential for initiating meta-analysis is the column for Study names (author or year). To insert the column for study names:
- Click on ‘Insert’.
- Bring cursor on ‘Column for’.
- Select ‘Study names’.
The Comprehensive meta analysis (CMA) software will insert the column for ‘Study names’ as shown below.
Step 2: Inserting column for effect size data
Following the insertion of column for ‘Study names’, the insert the column for effect size data. As shown below, ‘Effect size data’ option could be accessed by:
- Click on ‘Insert’
- Bring cursor on ‘Column for’
- Click on ‘Effect size data’.
The quick access options toolbar also provides access to ‘Effect size data’ option. The selection of effect size data option leads to the ‘Welcome’ dialog box. Upon selecting ‘Next’ leads to the dialog box showing ‘Types of studies included’, as shown below. Select ‘Estimates of means, proportions or rates in one group at one time-point’ (encircled), as the data for the events pertain to the patients having gallbladder cancer associated with gallstones.
Step 3: Inserting one group data
Within one group data entry category, the software interface offers multiple formats for data entry. The gallstone carcinoma case study data belongs to the Dichotomous type of Raw data, with number of events and total sample size (Number of patient with the disease). Hence:
- Click on ‘Dichotomous’.
- Under ‘Raw data’ select ‘Events and sample size’ (Figure below).
Spreadsheet and data entry
Step 4: Dichotomous raw data entry format columns
Eventually, the CMA software creates a spreadsheet with columns for data entry, effect size measures and measures of accuracy,as shown below. Here, the column for ‘Events’ takes up the number of favorable events (i.e. numbers of patients with gallstone associated with gallbladder cancer). Whereas, the column for total ‘Sample size’ takes up those with and without event reported. Furthermore, while entering the data, the software automatically calculates the yellow highlighted fields.
The spreadsheet display can be customized as per choice by clicking on the effect size column name tags (‘Event rate’ and ‘Logit event rate’) during data entry. Select between different options offered in the displayed menu, to customize the display as per his or her choice.
- Right click on the ‘Effect size’ name tag.
- ‘Menu’ appears.
- Select the desired option as shown in the menu.
Here, note that primary index could be anything (Event rate, or Logit event rate). ‘Event rate’ constitutes the primary index, for the present case research. The interface displays only event rate, the primary index, upon selecting (a). While, option (b) allows change of primary index, to other indices for example, ‘Logit event rate’. Whereas, in case of (c) customize the display of effect sizes as shown below (Summation of different options provided earlier in the menu).
Effect size indices dialog box allows selecting the primary effect size index and the type of indices to be displayed in data entry. Moreover, it also allows controlling the display of measures of accuracy (standard deviation and variance in the present case). As the data is entered into the entry columns, the CMA software simultaneously computes the effect size (event rate & logit event rate) and displays them in the columns highlighted as yellow. The effect size indices for the present data format are ‘Event rate’, ‘Logit event rate’, ‘Standard error’ and ‘Variance’. The image below shows the final spreadsheet with complete data.