Summary
A dashboard is built on a dataset, by placing widgets on a grid. Each widget is a plugin configured through a form — key indicator, chart, descriptive table, statistical test — and filters declared once apply to every widget at once.
No code is required. The demo dashboard tracks the activity of an intensive care unit: how many patients, what ages, which units, what mortality.
The dataset
A dashboard does not read the database directly: it builds on a dataset.
This one, icu_activity.csv, holds
1,810 rows describing intensive care stays — identifiers, sex, age, admitting
unit, dates.
Dashboards
A project's dashboards are listed under Lab → Dashboards. A project can hold several — one per purpose, per unit, or per audience.
When opened, the dashboard is organised into tabs — here a single one, Demographics — each holding its own grid of widgets. The Edit button switches to edit mode, Export exports the dashboard, and the fullscreen icon turns it into a presentation.
Adding a widget
In edit mode, Add widget opens the plugin library. Each one answers a specific display need, and all of them ship with the application.
The Custom code tab, next to Plugin, lets you write your own widget when no plugin fits — but most needs are covered without going there.
Once a plugin is chosen, its configuration form opens with a live preview on the right, computed on the real data.
An indicator therefore comes down to three choices: the column, the statistic (mean, median, count, sum…) and the formatting. The Unique per option counts distinct values of a column — that is what turns 1,810 rows into 100 unique patients.
Filters
The Filters panel on the right is what turns a collection of charts into a dashboard. A filter is declared once — here an age range from 18 to 92 — and applies to every widget at the same time.
The All tabs label means the filter follows the reader from one tab to the next. A widget can also be frozen so it ignores filters, which keeps a reference value beside a filtered view.
Going further
- Analyse the data — where the datasets feeding a dashboard come from.
- Dashboard documentation — the full reference for widgets and filters.
- Collect and analyse a survey — the same widget engine, applied to eCRF data.