Users can build a cohort from a dataset by choosing the query icon from the dataset tab:
This opens the Cohort Builder.
The Cohort Builder has two main components:
- The Query Builder
- Visualisations
The Query Builder
The current query builder allows users to:
- Create multi-clause queries
- Create multiple sub-cohorts to compare
- Query across multiple dictionaries where the dataset has more than one associated with it
Building a query
Below is a simple example that explains how the query builder works.
Single clause query
If a user wants to build a cohort that covers males with a BMI of 25-30, they only require a single clause query, searching for males with a BMI in the range of 25-30.
Multi-clause query
If a user then wanted to add males with a BMI of 15-20 to this cohort they would need to add an additional clause, searching for males with a BMI in the range 15-20.
This will create one cohort which consists of males with a BMI between 15 and 20, and males with a BMI between 25 and 30.
Multiple sub-cohorts
If a user wanted to compare males with a BMI of 25-30 to males with a BMI of 15-20, then rather than opening a new clause they would have to create a new sub-cohort by clicking the '+', and creating a single clause query for males with a BMI of 15 to 20 in the new sub-cohort.
Visualisations
The beta version of the cohort builder offers four visualisations:
- Bar chart
- Stacked bar chart
- Box plot
- Histogram
Users can toggle between, and make changes to, these visualisations by using the drop down menus at the top of the display window. In addition, users can download their visualisations in a number of file types.