RGTables: the TFL dashboard
Tables, Figures and Listings Inside the EDC
Most EDC systems stop at data export. ResearchGuru EDC includes a TFL dashboard, so your team can produce numbered tables, figures, listings and statistical analyses from live study data without leaving the database.
Pick the variables, choose how to stratify the columns and group the rows, and the output is drawn with headings and footnotes written for you. Save it and it re-runs on the current data every time it is opened, from the first participant to database lock.
| Arm A N = 71 | Arm B N = 71 | Total N = 142 | ||
|---|---|---|---|---|
| Systolic blood pressure (mmHg) | ||||
| n | 70 | 69 | 139 | |
| Mean (SD) | 128.4 (14.2) | 131.1 (15.0) | 129.7 (14.6) | |
| 95% CI | (125.0, 131.8) | (127.5, 134.7) | (127.3, 132.2) | |
| Median (Q1–Q3) | 127 (118–138) | 130 (121–141) | 129 (119–139) | |
| Sex, n (%) | ||||
| Female | 38 (53.5%) | 35 (49.3%) | 73 (51.4%) | |
| Male | 33 (46.5%) | 36 (50.7%) | 69 (48.6%) | |
Illustrative example
How it works
From live data to Table 14.1.1 in four steps
No programming and no export-import cycle. A study team member with the right permission builds each output on screen.
Choose the study and the variables
The dashboard reads the project's items, codes and labels directly. Tick the variables to analyse.
Stratify and group
Put any categorical variable across the columns, such as treatment arm or case and control, and group the rows by visit. Add a Total column, p-values and filters.
Edit and number
Every title and heading is editable on screen. Saved outputs carry TFL numbers in the usual style: Table 14.1.x, Figure 14.2.x, Analysis 14.3.x and Listing 16.2.x.
Export
Send one output, or every saved output with a contents page, to Word, RTF, CSV or PDF. Figures also export as PNG and SVG.
Statistics
Descriptive tables and inferential analyses
The dashboard covers the tables most studies need during data collection and at reporting, from baseline characteristics to adjusted models. Each method was checked against an independent implementation on a demonstration study.
Outputs are for study monitoring, exploration and reporting. For a regulatory submission, final analyses are reproduced in the validated statistical software named in the SAP.
Available methods
Traceability
Every number can be traced back to the database
An output is stored as a definition, not as a set of numbers. When it is opened or exported it runs again on the current data, and its footnote states the project, the extraction time and the number of records included.
With the audit trail active, the footnote also carries the audit-trail entry at extraction, and each export and each change to a saved output is written to the trail.
What this gives you
| Time | User | Field | Change |
|---|---|---|---|
| 10:42:07 | j.okafor | sbp_v2 | blank → 212 |
| 10:58:31 | j.okafor | sbp_v2 | 212 → 122 |
| 11:05:12 | r.patel | ae_ser | Query raised |
Illustrative example
Worked example (dummy data)
Forty outputs on one case-control study
Our demonstration case-control project has 1,000 dummy records and 40 saved outputs: ten tables, ten figures, ten listings and ten analyses. The analyses run from chi-square tests and odds ratios with confidence intervals to a multivariable logistic regression with a forest plot.
A twelve-minute video walks through all ten analysis tables. Every figure shown is dummy data, generated for demonstration.
Frequently asked questions
TFL dashboard: common questions
Something else? Get in touch and a senior researcher will answer.
No. It removes the wait between data entry and the first look at the data, and it gives statisticians and study managers the same numbered outputs to discuss. The statistical analysis plan and its interpretation remain with your statistician.
They are designed for monitoring, exploration and reporting. For a submission, final analyses are reproduced in the validated statistical software named in your SAP; the dashboard's outputs give an independent cross-check.
Only users with the permission you assign for that project. A blinded study can keep the arm variable empty until unblinding.
See the TFL dashboard on your own study design
Send us your questionnaire/form and study requirements. We'll review the proposed structure and come back to you with:
Or email admin@edcstat.com · call +44 7484 816484