Skip to main content

Name your variables and predict your export columns

The naming rules Datacapt enforces, how section prefixes and suffixes build the final name, and the column names repeated measures, tables and checkboxes produce in an export

The variable name of a question becomes a column header in every export. Your statistician reads those headers for the whole life of the study. Decide the convention before you build the first form.

šŸ”‘ Prerequisite: eCRF Builder (Edit), and a study in Draft or Suspended status.


The rules Datacapt enforces

Rule

Detail

Characters

Letters, digits, hyphen and underscore. No space, no accent, no other symbol

Length

100 characters for the name you type

Uniqueness

One name per study, across the eCRF, the ePRO forms and the eConsent. weight and WEIGHT count as the same name

Left empty

Datacapt builds a name from the first five characters of the title, an underscore and six random characters, for example WEIGH_K3X9QA

When two questions share a name, the Builder refuses to save and says "One or multiple questions have the same variable. Please add a suffix or a prefix to continue".

šŸ’” Tip: type your own names. An auto-generated name such as WEIGH_K3X9QA tells a statistician nothing, and you will rename it later at the cost of a changed export.

The eCRF has no reserved words. The Subject Database does: datacapt_id, first_name, last_name, email, phone, language and status belong to the platform, and the Builder answers "You cannot use this variable name".


Use section prefixes and suffixes for repeated visits

The same form often appears at several visits: vital signs at V1, V2 and V3. Give each question one base name, and let the visit add its prefix or suffix.

šŸ› ļø Step by step guide

  1. In the eCRF Builder, select Structure.

  2. Open the edit window of the section (visit).

  3. Switch on Add prefix or Add suffix, then type the text, 25 characters at most. Each prefix and each suffix must be unique in the study.

  4. Click Save.

Datacapt builds the final name as PREFIX_VARIABLE_SUFFIX, in uppercase. With the prefix V2, the question SBP exports as V2_SBP.

šŸ’” Tip: build visit 1 with plain names, give it the prefix V1, then use Duplicate section and change the prefix to V2 on the copy. You get a second visit with unique names in under a minute.

The same window holds the Domain field, letters and digits, for teams that group forms by CDISC-style domain (DM, VS, AE). The Domains export option of the export window then produces one sheet per domain.

In a Calcul formula, a Data display or a condition, use the full name as the Builder shows it: {V2_SBP}.


Predict your export columns

Question

Columns in the export

Text, number, date, radio, dropdown, rating, slider, calcul

One column: VAR

Checkbox, export by Values

One column per option: VAR$1, VAR$2, and so on, using each option's Value. The "Other" comment field adds VAR$ followed by its label

Checkbox, export by Labels

One column, with the selected labels separated by a vertical bar

Repeated measures

One column per measure: VAR_1, VAR_2, up to the maximum number of measures. Tick RM in a separate sheet to get one row per measure instead

Table/Matrix

One column per row of the table: VAR_ROWTITLE

Two options on the Advanced tab of a question change what comes out:

  • Exclude from data exports removes the question from every export.

  • Hide in eCRF removes it from the site's screen and keeps it in the export.


Check your names before go-live

Run an export on your test subjects and open the Items Dictionary sheet. It lists every question with its visit, form, domain, variable name, type, units, permissible values, data validations and conditional logic. Send it to your statistician for review before you set the study Live.

āš ļø Warning: you can rename a variable while the study is Suspended. Answers stay, but the export column changes name from that day. Any SAS or R program written against the old name breaks.


A convention that works

  • Uppercase, 8 characters or fewer for the base name when your statistician exports to SAS transport files.

  • CDISC CDASH names when they exist: BRTHDTC, SEX, AESTDTC, VSORRES.

  • Numeric option values in a fixed order across the study: 0 = No, 1 = Yes.

  • No visit number inside the base name. The section prefix does that job.


šŸ‘‰šŸ» Next step

Did this answer your question?