Edit Checks
Edit Checks are validation rules that are associated with a data-point. The rules verify, confirm or authenticate data entered into the data-point. Edit checks can also be used to verify the occurrence of an event and trigger an action, such as, displaying a query message.
The Medidata Rave EDC system has extensive user programmable, Edit Check functionality:
- Value range checking
- Cross-checking values within a Form and between Forms
- Global checks across all data
- Internal programming language or direct import of C#, VBS, SQL code
Value Range Checks
Question/RangeChecks are used to test a collected value meets a simple set of criteria and are suitable for:
- Value = X ; Value !=X
- X<= Value <=Y
- Value = X, Y, Z ; Value != X,Y, Z
Note: A Value Range Check only applies to the Question (ItemDef) it is attached to.
From the Properties tab under Range checks, click Add.
Specify the comparator. For example, less than (LT), greater than (GT), equals (EQ)
Select Add and enter the data value (IN, NOTIN allows multiple data values)
Note: This Formal Expression tab is not supported by Rave
Select the Error Message tab
Set Soft/Hard:
Soft: checks allow the value to proceed
Hard: checks do not allow the value to proceed.
Enter the Query message text
Click Validate & Update
Scroll down to the Medidata Rave Properties and click Add Range Check Custom Attributes group
Complete the following:
- Bypass During Migration: used to specify if the Range Check will be bypassed during CRF Migration. Defaults to “false”
- Active: used to specify if the Range Check is active. Defaults to “true”
- Marking Group: used to specify the Marking Group for the Query. The options for Marking Group are configured in the Master Marking Group property at Asset Group Level
- Require Response: used to specify if the Query raised by the Range Check requires a response. Defaults to “false”
- Require Manual Close: used to specify if the Query raised by the Range Check requires to be manually closed. Defaults to “false”
When the ALS is exported, an Edit Check is created
Field Edit Checks
Field Edit Checks are the simplest type of Edit Check. The following Field Edit Checks can be configured for Questions/Fields:
- Auto-Query for required data entry
- Auto-Query for non-conformant data
- Auto-Query for future date/time
- Auto-Query for data out of range
- Mark non-conformant data out of range
Auto-Query for required data entry
This Edit Check generates a query when the selected field is left blank and submitted by the user. To set this property select the ‘Mandatory’ checkbox next to the Question in the Section Editor. This populates the ‘IsRequired’ column in the ALS Fields tab to TRUE.
Note: this property can also be set at Field (ItemRef) Level properties by selecting the ‘Auto-Query for required data entry’ property.
Auto-Query for non-conformant data
This Edit Check generates a query when the user enters data in a format other than what has been defined for the variable. For example, if text is entered into a numeric field.
This property can be set at Question Level properties and Field (ItemRef) Level properties.
Auto-Query for future date/time
This Edit Check generates a query when the user enters a date in the future into a date field.
This property can be set at Question Level properties and Field (ItemRef) Level properties.
Note: The control type for the field must be defined as ‘DateTime’.
Auto-Query for data out of range:
This Edit Check generates a query when the user enters data outside a defined range into a field.
At Question Level, this Edit Check is set using Value Range Checks. At Field (ItemRef) Level properties, it is set using the ‘Auto-Query For Data Low’ and ‘Auto-Query For Data High’ properties.
Note: this Edit Check only applies to numeric variables (i.e. with data type ‘integer’ or ‘float’).
Mark non-conformant data out of range:
This Edit Check automatically sets the field status to non-conformant when the data entered in the field is outside a defined range. At Field (ItemRef) Level properties, it is set using the ‘Mark Non-Conformant Data Out of Range: Low’ and ‘Mark Non-Conformant Data Out of Range: High’ properties.
These values populate the “NCLowerRange”, “NCUpperRange” columns in the ALS Fields Tab.
Note: this Edit Check only applies to numeric variables (i.e. with data type ‘integer’ or ‘float’).
Edit Check Programming
Medidata Rave has an internal programming language consisting of common arithmetic and logical operators, combined with function library. Rave Edit Checks are implemented as a series of programming statements, each returning a boolean value.
These statements are implemented in “PostFix notation” or “Reverse Polish notation (RPN)”, common to scientific calculators. “PostFix” notation places the operators after the operands. For example,
2 5 + , 10 2 ÷ , X Y AND
This removes the need for parenthesis and provides an unambiguous formula suitable for machine processing.
Edit Checks are defined through Conditions within the ODM at Metadata Version level. These are built up from a series of Check Steps and Check Actions. See Condition Level properties.
Edit Check Example
This example shows how to create an Edit Check which opens a query on the 'Age' field if the value entered is less than ‘18'.
Create a Condition - details
Create a Data Value Check Step for the Age Variable
Create a Constant Check Step for the value ‘18’
Create a Check Function Check Step for ‘Is Less Than’
Create a Check Acton to display a Message against the Age Question
The 'Folder' property within a Data Value, or within a Data Value Derivation Step, should not be populated if the 'Form' property specifies the Primary Form.