Research Workspace
Turn your own records into studies you can explain. Defined, counted, versioned, on the record.
State the question and the unit, build the patient group criterion by criterion, prepare the data in named steps and analyse it with the methods clinical studies use. Every result stays traceable to its source.
What this gives you
- Define the purpose and the unit before touching any data.
- See how each criterion narrows your patient group.
- Freeze a version and repeat the same steps on newer records.
How a project runs
-
01Define the study
-
Purpose and unit
Six purposes, from audit to thesis, and a unit of analysis from the patient down to a tooth surface.
-
Cohort
Nested criteria with live counts, showing what each step removed.
-
Data sources
Demographics, conditions, medications, measurements, history answers, visits, treatment plans and tracking, or your own spreadsheet.
-
Purpose and unit
-
02Prepare the data
-
Inspect
A read-only grid that keeps six kinds of missing value apart, from “not asked” to “declined”.
-
Transform
Recode, derive, filter or correct from a fixed menu. Every step is named, undoable and logged; no cell is edited in place.
-
Freeze
Freeze a version, then refresh it by re-running the recorded steps on current records.
-
Inspect
-
03Analyse and explain
-
Statistics
Table 1, group comparisons, correlation, regression, survival curves and ROC analysis.
-
Figures
Ten chart types, from forest plots to tooth-surface heatmaps, each captioned with its source, version and missing data.
-
Results you can repeat
Saved analyses record how they were made, and tell you when their source has changed.
-
Statistics
Also in the workspace
- Dentistry built in
Tooth, surface, site, implant and radiograph as units of analysis, with DMFT and DMFS built in.
- Practice reports alongside
Visits, recalls, case mix and record quality, where every figure opens the named patients behind it.
- Activity on the record
Every view and change in a project is recorded, with notes and tasks kept beside the data.
- Erasure reaches every version
When a patient’s data is erased, it leaves every version of a dataset, frozen ones included.
- More statistical methodsIn development
Cox models, mixed models, agreement and power analysis.
- Research exportsIn development
Table 1, the variable dictionary, the methods log and figures, released once research consent can be checked.
- Patient research consentIn development
Patients grant, refuse or withdraw consent for each category of data.
Safeguards in the Workflow
Research with its limits in plain view
Research is for doctors, and each project has one owner working from their own patients’ records. Groups of 1 to 10 are hidden on the server, or fewer than 20 for sensitive categories, and an analysis will not run if it would quietly leave out a small group.
Mood, cycle and reproductive-health entries are kept out of research. Research data is not anonymised, so it stays inside the project and its access rules; anonymised export is planned.
- Reports & InsightsEveryday reports answer day-to-day questions; the workspace goes deeper.
- TrustThe same access model governs research and care.
Help shape research tools worth trusting
If you work with clinical data, show us where today’s tools hide assumptions, lose the trail back to the source or make privacy harder to manage.