| Oracle® Clinical Creating a Study Release 5.1 E52787-01 |
|
|
PDF · Mobi · ePub |
| Oracle® Clinical Creating a Study Release 5.1 E52787-01 |
|
|
PDF · Mobi · ePub |
An Investigator collects all study data as responses to clinical study Questions.
This chapter contains the following topics:
These activities take place within Oracle Clinical's Global Library subsystem. For information on other Global Library functions, see Chapter 8, "Global Library Activities" and Chapter 9, "DCMs and DCIs."
This section contains the following topics:
Oracle Clinical Questions correspond to the Questions on a CRF for a patient visit. You must define an Oracle Clinical Question for each Question on a CRF.
Question definitions are reusable.
In the Oracle Clinical Global Library, you can create Question Groups that contain multiple logically related Questions to be collected in the same section of a CRF. You then create DCMs (Data Collection Modules) that contain multiple logically related Question Groups to be collected together. You can create DCIs (Data Collection Instruments) that contain multiple logically related DCMs. In general, a DCI corresponds to a CRF.
The same Question can be included in many Question Groups. The same Question Group can be included in many DCMs, and so on. You can create standard DCMs and DCIs in the Global Library for reuse in multiple studies, and modify them as necessary for particular studies.
There are three windows in the Global Library in which you see the same Question definitions. Which of the three windows you can see depends on your user privileges:
Query Questions (from the Glib menu, select Questions, then Qry Questions). You can query and view all Global Library Questions but cannot make any changes.
Provisional Questions (from the Glib menu, select Questions, then Prov Questions). You can view all Global Library Questions and create and modify Questions with a status of Provisional, but you cannot change a Question's status to Active. You can delete only provisional Questions, and only when they are not included in a Question Group or DCM.
Questions (from the Glib menu, select Questions, then Questions). You can do all Question-related Global Library tasks, including all the activities possible in the Provisional Questions window, plus changing a Question's status to Active or Retired, and changing some attributes of Active Questions.
Global Librarian To see the Questions window you must have the database role Global Librarian. The Global Librarian is responsible for organizing and cleaning the set of Questions in the Global Library: defining Question categories and medical evaluation qualifiers to apply to each Question, determining which provisional Questions to promote to Active status for use in production, and which Active Questions to retire.
The Global Librarian is also responsible for defining values in two installation references codelists (under the Admin menu) that affect Questions: Question Category Type Code and Medical Eval Type Code. The system uses values defined in these codelists to populate the lists of values for the Question Categories and Medical Evaluation Type attributes for Questions.
The Maintain Questions window has two viewing modes: Multi displays multiple Questions and their attributes, using one row per Question; and Single displays one Question at a time so that you can see all its attributes without scrolling. The Multi window has several buttons the Single window lacks; see "Assigning Question Categories", "Defining Details for Complex and Question Set Questions", and "Defining Extended Attributes".
To create a Question:
Complete these fields:
Name: A unique name within the Question's domain, up to 20 characters.
Domain: The Question's domain. See "Maintaining Domains".
Status: There are three valid statuses:
Provisional (P). The default status for new Questions. All attributes of provisional Questions can be modified, but they cannot be assigned to Question Groups or used in any way. Provisional Questions can be deleted.
Active (A). Active Questions are available for use: they can be assigned to Question Groups and DCMs, and used in test and production studies.
Retired (R). Retired Questions cannot be assigned to Question Groups, but Questions already in use in Active Question Groups can continue to be used.
Note:
You can change a Question's status from Active to Retired at any time.You cannot change a Question's status from Active to Provisional if the environment is replicated; that is, if there are more than one source locations set to Active in the installation reference codelist SOURCE LOCATION CODE; or if the Question is used by any other object:
included in an active Question Group
has Lab Ranges defined
has Preferred Lab Units in a Preferred Conversion Group
has Lab Test Units defined
is used as a DCM Qualifying Question
is used as a DCM Question
Medical Evaluation Type: A classification for how to evaluate medical responses.
Intent: A unique description of the meaning or intended use of the Question. The text you enter here is included in the list of values for the Questions field in the Question Groups window; it helps identify the Question when you are selecting Questions to include in a Question Group.
Question Type: From the list of values, select a type. Depending on the type you select, the system enables or disables other Question attribute fields for the Question you are defining. The type determines what data type is allowed for the Question, as noted. For example, if you select a Question type of Char, Unit, or Thes Validated, all of which must have a data type of CHAR, the fields for date/time, decimal places and lower bound and upper bound are disabled. Choices include:
Unit (Data type must be CHAR.) Informational datatype useful for identifying Questions used to capture unit information related to another Question, such as a lab Question.
Lab Test (Data type must be CHAR or NUMBER.) Defines the Question as collecting lab data and requires collection of the lab name for each received DCM. A Question of type LAB TEST can have a lab range defined in the Lab Subsystem; see "Defining Textbook Ranges". If a DCM Question is based on a lab test Question and the DCM Question is used in a Procedure, lab range information for that Question is automatically available for use in the procedure.
If you are using normalized lab data, you must define a Question to display the lab test name with a Question type of Lab Test; see "Using Normalized Lab Data" for details.
Date Time (Data type must be DATE or TIME.) Use for a date or time Question.
Non-Lab (Data type must be CHAR or NUMBER.) Non-lab Questions with a data type of NUMBER can have Lower Bound and Upper Bound values defined at the DCM Question attribute level as well as in the global library.
Complex. (Data type must be CHAR, NUMBER, DATE or TIME.) Complex Question values are derived by a macro from other Question responses and appear only in data extract views. They do not have to be defined as a DCM Question. Selecting Complex activates the Details button. See "Defining Details for Complex and Question Set Questions".
Char (Data type must be CHAR.) Use for character-data type Questions.
Thes(aurus) Validation (Data type must be CHAR.) For use with the Oracle Thesaurus Management System (TMS) only. Thesaurus Validation Questions are not collected at a patient visit, but receive data derived from TMS, such as the dictionary term to which the verbatim term is classified. If a DCM Question is based on a Thes Validated Question and the DCM Question is used in a Procedure, the derived TMS information for that Question is automatically available for use in the Procedure. You must select the Derived? check box for the Question and include it as a child Question in a Question Set; see "Defining Details for Complex and Question Set Questions".
Question Set (Data type must be CHAR, NUMBER, DATE or TIME.) For use with the Oracle Thesaurus Management System (TMS) only. Use for Questions you will use to collect TMS verbatim terms—responses collected at a patient visit that will be classified to a standard dictionary term in TMS. For example, you might collect a response to the Question CONCOMMITANT_MEDICATION at a visit and check the response against the dictionary WHO-Drug.
You must include the Question as the parent Question in a Question Set, with a child Question, or variable, for each term you want to derive from TMS back into Oracle Clinical; see "Defining Details for Complex and Question Set Questions" and the Oracle Thesaurus Management System User's Guide for more information.
Selecting Question Set activates the Details button.
Extended_Text (Data type must be CHAR.) Extended Text Questions can store a response of 10,000 characters instead of 200 characters, which is the maximum length for all other character-based Question types. See "Extended Text Questions" for restrictions and additional information.
Data Type: Data Type is a database classification: Character, Number, Date, and Time. The system generates a univariate discrepancy for responses to this Question that are of the wrong data type. For information about discrepancies, see the chapter "Using the Discrepancy Database" in the Oracle Clinical Conducting a Study manual.
Date Time Type: Specifies the expected precision of the response for a Question with Data Type of DATE or TIME.
For Questions with a data type of DATE, select DMY to include the day, month, and year; MY to include only the month and year; or Y to include only the year.
For Questions with a data type of TIME, select HMS to include hours, minutes, and seconds; or HM to include only hours and minutes.
Date Time Fmt (Format): The Date/Time Format must specify at least as much precision as specified by the Date/Time Type field.
Len (Length): Set the maximum length for responses to the Question. You can make the length longer when you add it to a study DCM, but never shorter. The system generates a univariate discrepancy for responses to this Question that are too long. For information about discrepancies, see the Oracle Clinical Conducting a Study manual.
Note on Extended Text Question Length:
All Extended Text Questions have a predefined length of 10,000 characters. However, the system cannot display the full text of the response in most places in the user interface. The length you define determines the number of characters displayed in these places in the user interface, including Oracle Clinical RDC Onsite discrepancy details.Note that the system reserves the first 24 characters for the attachment prefix, so if you define a length of 30, the system displays the prefix plus the first six characters of the response.
Dec Plc (Decimal Places): Specifies the expected maximum number of digits to the right of the decimal point for a response to a NUMBER Question. The number in this field counts toward the total number you specify in the Length field. Enter 0 to specify an integer response. This value is the Question's default Decimal Place value when you copy the Question to Question Groups or DCMs. You can change this field only for a provisional Question, or, for an active Question, only to increase it.
SAS Name: The Question's unique identification, within the Domain, for to access data through the SAS package; up to 8 characters in length, cannot end with a number, and can be comprised only of uppercase letters, numbers, underscores, or hashes. If the Question has multiple occurrences in a single Question Group, this field can be only up to 7 characters. You can change this field only for a provisional Question, or, for an active Question, only to increase it. See Chapter 14, "Data Extract" for more information.
SAS Label: SAS variable label for data extraction. The Intent field is the default value for this field. You can change this field only for provisional and active Questions.
Extract Macro: Specify the Extract Macro associated with this Question, if any. A list of values is available.
Question Set Name: For Questions of type Question Set, enter the name of the Question Set to which you want to link this parent Question.
Safety Question?: Check if this is a safety-related Question.
Derived?: Select if this Question serves as the recipient of a response generated by a Derivation Procedure. All Derivation Procedures must have a target derived Question, which cannot collect a response in any other way. (See "Defining Procedure Questions" for more information.)
Select also for all Questions to be used as TMS Question Set child Questions, because their responses are derived from TMS. You can also define a TMS Question Set parent Question as derived if its value is to be generated by an Oracle Clinical derivation procedure. See the Oracle Thesaurus Management System User's Guide for information on using parent Questions this way.
Note:
Oracle supports only the amount of derivation that can be accomplished within a single batch validation. Oracle does not support using an Oracle Clinical derivation procedure to populate the value of another derived parent Question with the value of the original derived parent Question.Upper Case?: If selected, the system records the response in upper case characters.
Sight Verify?: Select this to display the Pass-1 response during Pass-2 data entry.
Default Prompt: The free form text (up to 60 characters), you enter is the prompt text associated with the response field when the Question is rendered in a data entry form. You can change this value later.
The DVG and Alpha DVG fields: You can limit the Question's responses to a static list of values, or Discrete Value Group (DVG). The system generates a univariate discrepancy for responses to this Question that are not included in the DVG. See "Assigning DVGs to a Question".
Even if you do not want to assign a DVG to this Question now, you can select DVG Modifiable? (and/or Alpha DVG Modifiable?) to allow the option to add a DVG to this Question at a later time, even after the Question's status is Active, within a particular Question Group.
Lower and Upper Bound: fields (Optional; applicable only to Questions of data type Number): Enter an appropriate upper and lower limit for the response to this Question. The system generates a univariate discrepancy for responses to this Question that are above the upper bound or below the lower bound.
Valid(ation) Failure Type: During data entry, if a response to the Question fails validation, the discrepancy record the system generates receives this classification.
Choices include: Verify, Normal, Critical, Monitor Alert, Safety, Thesaurus-PT, Thesaurus_SCT.
Upper Case?: If you check this field the characters in the response to a DVG-associated Question are forced to uppercase. You can change this field only for Provisional and Active DCMs.
Pass 2 Sight Verify: If you check this field, the system completes the Second Pass Data Entry response field with the response from the first pass, allowing the response to be visually verified instead of being rekeyed. If this field is unchecked, the second pass field is left blank and requires reentry. You can change this field only for Provisional and Active DCMs.
Status Comment: If you change the Question's status, enter the reason.
Retirement Reason: If you change the Question's status to Retired, select the reason from the list of values.
Save.
A Question cannot have Active status unless you associate it with at least one Question category. A Question category is a broad search classification, such as Demography or General Physical, that you can use to find existing Questions. You define these values in the QUESTION CATEGORY TYPE CODE installation reference codelist; see the Reference Codelists chapter of the Oracle Clinical Administrator's Guide for editing instructions.
Question Categories: A Question cannot have Active status unless you associate it with at least one Question category.
With the Question selected, click Question Categories. If the Medical Evaluation Type of the Question is also a valid Question category, it is automatically assigned to the Question.
From the list of values in the Question Category Code field, select as many Question categories as appropriate.
Save.
Extended Text Questions support lengthy freeform text responses; for example, a description of a serious adverse event or a patient case history.
In RDC Onsite, an Extended Text Question field is denoted by a yellow note icon. When a user clicks the icon, double-clicks within the field or right-clicks and selects Show Extended Text from the right-click menu, a pop-up window appears in which the user can enter or paste up to 10,000 characters.
The following restrictions apply:
Responses to Extended Text Questions can be entered and updated only in RDC Onsite. You cannot use Oracle Clincial, Mass Changes or Batch Data Load to enter or update values for Extended Text Questions. (However, Batch Data Delete can delete Extended Text Question responses.)
The system cannot display the full response text in many parts of the user interface. Instead, it displays the value in the RESPONSES table VALUE_TEXT column: the beginning of the response plus information about the version of the attachment that includes the full text; see "Database Storage". This truncated response is displayed in Oracle Clinical, Discrepancy Details in RDC Onsite, Audit History in RDC Onsite, and some sections in Patient Data Reports. In the case of the RDC Onsite Audit History pane and Patient Data Reports, the full text is available in a separate window or page.
Validation Procedures have access only to the truncated response value text stored the the RESPONSES table; see "Database Storage".
Data Extract does not support Extended Text Questions.
Do not use an Extended Text Question as an indicator Question, as a trigger Question in a flexible study, or as a source Question in conditional branching.
Special characters entered as part of an extended text response may not be displayed properly in the user interface.
The system enforces the following restrictions:
You cannot assign a Discrete Value Group (DVG) to an Extended Text Question.
You cannot define a default response for an Extended Text Question.
You cannot define an Extended Text Question as Mandatory.
You cannot define an Extended Text Question as a derived Question.
You cannot define a Question Group as repeating if the group contains an Extended Text Question.
Oracle Clinical automatically sets the Enterable check box to unchecked for an Extended Text Question. You cannot change it. However, although the Question is identified an non-enterable, you can use RDC Onsite to enter and update data for an extended text field. You cannot use Oracle Clinical, Mass Changes, or Batch Data Load to enter or update data in an extended text field.
In the places where the system may not be able to display the full response to an Extended Text Question—Oracle Clinical, Discrepancy Details in RDC Onsite, Audit History in RDC Onsite, and some sections in Patient Data Reports—the system displays a prefix:
ATTACHMENT V<X.Y>:
X is the version number; the number of consecutively different changes saved by a user in RDC Onsite. The first time a user enters and saves a response for a patient, the version number is 1. If a user updates the response and saves again, the version number is 2 unless the actual value has not changed; the system compares the existing value against the value being saved and increments the version number only if there is a difference. The version number can have up to four digits.
Y is the subversion number; the system increments it each time the focus leaves the extended text field but before the user saves the record. It is used for internal auditing.
The system reserves 24 characters, the maximum possible, for the attachment prefix string whether all 24 characters are used or not.
Oracle Clinical stores responses to Extended Text Questions in two database tables; a truncated version in the RESPONSES table and the full text in the RESPONSE_LOBS table. The truncated version is displayed in several places in the user interface where the system cannot display the full text.
The RESPONSES table VALUE_TEXT column stores the 24-character Attachment Prefix plus the response entered, up to the defined Question length, truncating any remaining response text. The number of response text characters stored depends on the Question length:
Number of characters in truncated response text = Question length - 24
For example, if you define an Extended Test Question to have a length of 30 and a patient response to the Question begins "A 72-year-old male patient," the system stores (and displays in the user interface) the attachment prefix and the first six characters of the response, for example:
ATTACHMENT V<2.3>: A 72-y
You can limit the valid responses to a Question to a DVG. For information about DVGs, see "Creating and Using DVGs".
Questions used as Qualifying Questions to distinguish multiple occurrences of a DCM in a DCI must have a DVG. See "Configuring Qualifying Questions".
The first set of fields, labeled DVG, refer to both internal and thesaurus DVGs. The second set of fields refer to alpha DVGs. Otherwise, the functionality of the two sets of fields is the same:
(Alpha) DVG Modifiable? This box controls whether or not any changes can be made to the DVG settings for this Question within the context of a particular Question Group or DCM. If the box is deselected here, it is also deselected whenever the Question is included in a Question Group or DCM, and no changes can be made within those contexts.
If the box is selected here, it will be selected by default for the Question within a Question Group, in the Global Library and in a DCM. Within those contexts, it is then possible to change the DVG or deselect the box. See "Adding DVGs to Active Questions, Question Groups, and DCMs".
The Global Librarian can always change this setting here, even after the Question is set to Active. When the box is selected, it is possible to change the DVG assignment.
Note:
The Global Librarian can assign a different subset of the same DVG at any time, even if this box is not deselected.(Alpha) DVG Name To assign an internal or thesaurus DVG or an alpha DVG to a Question, choose a DVG from the list of values in the DVG Name field.
(Alpha) DVG Domain The system enters the name of the DVG's domain.
(Alpha) DVG Subset # Specify the subset of the DVG that you want to assign to this Question. The default is Subset 1. See "Creating a DVG Subset".
Note:
The base subset, which contains the complete list of values defined for the DVG, is Subset 0 (zero).The Details button in the Maintain Questions window is enabled only for Question Set Questions (used only in TMS) and Complex Questions (used only in Data Extract). In both cases you associate Questions with variables already defined for the current Question.
Question Set Questions In this window, associate the variables defined in the current (parent) Question's Question Set with the derived Questions that will receive the variables' values from TMS. You must first define the Question Set, the parent Question (of Question type Question Set), and each derived Question (of Question type Thes Validated).
The system populates a row for each variable defined for the Question Set.
In the Question Name column for each variable, from the list of values select the name of the derived Question to receive the variable value from TMS. The list of values includes only Questions whose Question type is Thes Validated.
Save.
Complex Questions In this window, associate the Extract macro variables with the Questions whose values the macro combines to derive the value of the current (Complex) Question. You must first define the Extract macro, the Complex Question, and the Questions to be used as variables in the macro.
The system populates a row for each variable defined for the current (Complex) Question's Extract macro.
In the Question Name column for each variable, from the list of values select the name of the Question whose value will be used as the variable value.
Save.
Complex Questions are not DCM Questions; they are created for the purpose of data extract. They are not part of default definition, so they have to be defined in the Global Library. Once a Complex Question is defined, it references other Questions as well as the Extract macro that combines them into a single value.
When you write an Extract macro with more than one variable, you accommodate that Extract macro by creating a Complex Question, which can be broken down into simple Questions, each of which can then be associated with one of the variables.
Oracle Clinical provides additional attributes that can be associated with Questions to increase the information available in data extract views. The attributes are columns in the RESPONSES table. By including them in a Question definition, you make it possible to extract the column value for a Question response.
By default, the system creates one attribute: VALIDATION_STATUS, which contains information on whether any discrepancies are logged against the response, and if so, their current status.
The other attributes available are:
AUDIT_COMMENT_TEXT. If the record was updated after it was made accessible in the Update mode of data entry, the comment, if any, added by the operator. Value is populated on the version prior to the new version, or, in the case of deletions, on the final version.
DATA_CHANGE_REASON_TYPE_CODE. Code indicating reason a new version of the response was created. Value is populated on the version prior to the new version, or, in the case of deletions, on the final version.
DATA_COMMENT_TEXT. Investigator comment, if any, for the response.
DISCREPANCY_INDICATOR. Box indicating whether a univariate discrepancy, a manual discrepancy, or both exist for the response. Possible values are U, M, and B, respectively.
DVG_LONG_VALUE. Applicable only to Questions associated with a DVG.
DVG_NUMBER. Applicable only to Questions associated with a DVG.
DVG_SHORT_VALUE. Applicable only to Questions associated with a DVG.
EXCEPTION_VALUE_TEXT. The full value of the response is stored in this column if a discrepancy has been created of the type that indicates that the value is inconsistent with the database use of the DCM Question. In particular:
Data type discrepancies store the value here with the value text null.
Length discrepancies store the full value here with the value text containing null for numbers and containing the text truncated to the DCM Question length for characters.
FULL_VALUE_TEXT
TERM_COL1. Applicable only to Questions associated with a thesaurus DVG.
TERM_COL2. Applicable only to Questions associated with a thesaurus DVG.
TERM_COL3. Applicable only to Questions associated with a thesaurus DVG.
You can remove a Question from the Global Library if it meets the following criteria:
It is a provisional Question. You cannot remove active or retired Questions.
The Question is not part of a Question Group. If it is, you must first remove the Question from the list of Questions in the Question Group. See "Managing Question Groups".
The Question is not part of a DCM.
From the Glib menu, choose Questions, then select (Prov) Questions. Click the Delete Record button, then save your work. If you have not yet saved the Question, there is nothing to delete and you can clear the record by clicking the Clear Record button.
This section contains the following topics:
Question Groups organize Questions in a study and collect related data by grouping related Questions. For instance, Question Group DEMOG may collect demographic data about patients and contain the Questions: BIRTHDATE, RACE, and SEX.
You can group multiple Question Groups in a single Data Collection Module (DCM) and multiple DCMs in a single Data Collection Instrument (DCI), which generally corresponds to a CRF. You can define a Question Group either in the Global Library or in a particular study DCM, but to promote consistency and reusability, define Question Groups in the Library and then copy them to study DCMs, where you can modify them for the needs of the study. Library Question Groups should be as general and complete as possible so that designers can easily modify them to work in any clinical study.
For example, a Library Question Group ALLERGIES might contain many Questions, each collecting data medically useful in various studies. If a study designer copied and modified ALLERGIES into a cardiovascular study study DCM, the designer might confine it to four Questions that gather a patient's history of allergic reactions to medication (HX_ALLERGIC_SENS, MED_NAME_II, ALLERGIC_REACTION, and FIRST_INCIDENT_YEAR). The designer would designate the other Questions in ALLERGY as "Not Collected."
Some choices of attributes you make in the library version of Question Groups impact what you are able to do when you include a Question Group in a study DCM and DCM subset. This section addresses some of these attributes:
If you disable COLLECTED at the library level, you cannot enable it in the study DCM Question Group.
If you enable both REQUIRED and COLLECTED at the library level, you must enable both in the study DCM and enable COLLECTED at the subset level.
If you enable REQUIRED but disable COLLECTED at the library level, you do the same at the study level, and you must enable COLLECTED in the DCM subset.
If you disable both REQUIRED and COLLECTED at the library level, then you cannot enable COLLECTED at the study level, but COLLECTED is enabled at the subset level.
If you disable REQUIRED and enable COLLECTED at the library level, but do not enable COLLECTED at the study level, then COLLECTED is also not enabled at the subset level.
If DERIVED is enabled, then ENTERABLE and MANDATORY must be disabled.
If MANDATORY is enabled, then ENTERABLE must be enabled. If ENTERABLE is enabled, then DISPLAYED must be enabled.
The maximum number of Questions in a Question Group is equal to 255, minus the number of fields in the key template. If the system does not present the Prov Question Groups menu item to you, see "Assigning Questions to a Question Group".
To create a new Question Group:
From the Glib menu, choose Question Groups, then select Prov Question Groups or Question Groups.
Enter values as follows:
Name – Must be unique within the given domain. Query only to view listings of existing Question Groups for guidance in naming.
Domain – The system supplies the first domain in your domain list. Use the List function to choose another, if necessary.
Status – There are three valid statuses:
Provisional (P). The default status for new Question Groups. All attributes of provisional Question Groups can be modified, but they cannot be assigned to DCMs or used in any way. Provisional Question Groups can be deleted.
Active (A). Active Questions are available for use: they can be assigned to DCMs and used in test and production studies.
Retired (R). Retired Question Groups cannot be assigned to DCMs but Question Groups already in use in Active DCMs can continue to be used.
Note:
You can change a Question Group's status from Active to Retired at any time.You can change a Question Group's status from Active to Provisional only if the Question Group is not used as a DCM Question Group. In addition, you cannot change a Question Groups's status from Active to Provisional in a replicated environment; that is, if there are more than one source locations set to active in the installation reference codelist SOURCE LOCATION CODE.
QG Type – The kind of data gathered by the Question Group. A list of values is available.
Expand? – If checked, you can add more Questions to a saved Question Group.
Description – A description of the Question Group to help other people know if they should use it.
Status Comment – (Optional) If you change the Question Group's status, provide a reason.
Retirement Reason Type – Required when you change the Question Group's Status to R. Select a value from the list. These values are set in the RETIREMENT REASON TYPE CODE installation reference codelist.
Save your work, or click the Exit button to discard it.
Click the Question Group Questions button to populate the Question Group. See "Assigning Questions to a Question Group" for details. The other fields in this window are for Question Group maintenance:
Use the status comment to explain changing a Question Group's status.
Select a retirement reason if you choose to prevent a Question Group from being used in new studies.
Click the Standards Affiliations button to assign the standards affiliations to the Question Group. For information on standards affiliations, see the Oracle Clinical Getting Started manual.
Once you create a new provisional Question Group, assign Questions to it. A Question must meet the following conditions to be assigned to a Question Group:
The Question must exist in the Global Library. If it does not, create the new Question first. See "Creating a New Question".
The Question Group must be provisional. (To add Questions to Active Question Groups, see "Managing Question Groups".)
To assign Questions to a Question Group:
From the Glib menu, select Question Groups, then choose Prov Question Groups.
Query and select the Question Group where you want to assign Questions.
Click the Question Group Questions button to display the Question Group Questions window.
Enter the name of the Question in the Question Name field. The system fills in the Question's current values when you select another field.
Fill in all mandatory fields. Depending on your computer's display configuration, you may have to perform several scroll actions to see all fields.
Note:
You can assign the same Question to the Question Group multiple times if needed. If there is only one occurrence of a Question in the Question Group, the system give it an Occurrence Sequence Number (Occ SN) of zero (0). If there are multiple occurrences, the system gives the first one an Occ SN of 1, the second an Occ SN of 2, and so on.Indicator branching allows you to control whether or not Questions in a Question Group are collectible for a particular patient depending on that patient's response to the first Question in the Question Group.
You mark the first Question in the Question Group as the indicator Question and specify the response value that, if entered for a patient, makes the rest of the Questions in the Question Group collectible.
If a patient response to the indicator Question matches the specified indicator value, the remaining Questions in the Question Group are collectible. Otherwise, the remaining Questions in the Question Group remain disabled, and the branch is to the first Question outside of the Question Group.
In RDC Onsite the Questions are enabled and the data entry form branches to the next Question.
In Oracle Clinical, there is no change in the appearance of the Questions but the remaining Questions in the Question Group become enterable.
If a patient response to the indicator Question does not match the specified indicator value:
In RDC Onsite the remaining Questions in the Question Group remain disabled and the data entry form branches to the first Question outside the Question Group.
In Oracle Clinical, there is no change in the appearance of the Questions but the remaining Questions in the Question Group remain unenterable.
For example, define the first Question, SEX, to collect the patient's gender. If a patient's response to SEX is FEMALE, then in RDC Onsite the remainder of the Questions in the Question Group, which cover pregnancy, become collectible for the patient. If the patient is male, these Questions are not collectible for the patient. The next Questions available in data entry are in the next Question Group in the DCM, if any, or the next DCM, if any.
You define indicator Questions and values in the Question Group Questions window in the Global Library.
Restrictions:
indicator Questions must be the first Question (according to sequence number) in the Question Group. If you do not define it first, on Save the system repositions the Question as first.
Only one value can make the rest of the Questions in the group collectible
An indicator Question cannot also be a conditional branch source Question, and should not be an Extended Text Question.
To create a new Question Group similar to an existing Question Group, you can copy an existing one. You can then add or change attributes to make the new Question Group fulfill all its purposes. Note that as soon as you make a copy, it exists as a distinct record; changes to the copy do not affect the original Question Group.
Navigate to Glib, then Question Groups, and Prov Question Groups.
Select Copy Ques Group from the Special menu to view the Locate Question Groups window.
Query to identify an existing Question Group with the attributes you want.
Select the Question Group.
Click the OK button to copy the Question Group.
The system closes the Copy window, redisplaying the Question Groups window. Its first field contains the name of a new record the system generates by prefixing COPYOF to the name of the original Question Group.
Change the name of the copied Question Group. If you do not change the name, the system retains the COPYOF prefix.
Save.
You cannot delete an Active or Retired Question Group. To delete a Provisional Question Group:
From the Glib menu, select Question Groups, then choose Prov Question Groups.
Query and select the Question Group you want to delete.
Click the Delete Record button.
Save.
Question Group management concerns setting the status of Question Groups, rejecting provisional Question Groups, defining sets of Question Group types, and other administrative tasks.
You can find the Question Group you want to work with by selecting Locate QG by SA (Standards Affiliation) from the Special menu. (See the Oracle Clinical Getting Started for information on standards affiliations.) For example, the standards affiliation type CARDIOVASCULAR may locate the Question Groups ECG_TRACE, ECG_ABNORM, EXER_STAGE, TREADMILL, and STOP_REASON.
Only the Global Librarian, who has access to the full-privilege Question Groups, can set the status of a Question Group to Active or Retired. A Question Group must have one or more Questions before you can make it Active. When you retire a Question Group, the system prompts you to fill in the Retirement Reason field.
To create lists of values (LOVs) of Question Group types, you (or someone who has the correct privileges) must modify the installation reference codelist DCM DCI QG TYPE CODE. Note that Question Groups have the same type codes as DCMs and DCIs.
Examples of Question Group types are: DEMOGRAPHY, MEDICAL HISTORY, PHYSICAL EXAM, ADVERSE EVENTS, MEDICATION, and COMPLETION. See the Reference Codelist chapter of the Oracle Clinical Administrator's Guide for instructions on maintaining reference codelists.
This section includes the following topics:
See also "Assigning DVGs to a Question".
A Discrete Value Group or DVG is a list of the values that constitute acceptable responses to a Question. You can assign a DVG or DVG subset to a Question to limit the allowable responses to the Question. If a data entry operator enters a different value as a response to the Question, the system creates a univariate discrepancy of type DVG and alerts the data entry operator (see the "Using the discrepancy database" chapter in the Oracle Clinical Conducting a Study guide for information on discrepancies).
For example, if a Question is associated with the DVG Yes_No, whose only acceptable values are Yes and No, and a data entry operator mistakenly enters "Male" as a response to the Question, the system alerts the operator, who can then check his or her work. If the operator does not correct the response, the discrepancy remains in the system, and you can use the discrepancy management system to clean the data.
DVGs are reusable; you can assign the same DVG to multiple Questions. For example, a DVG with the values Yes and No is appropriate for many Questions.
Qualifying Questions When you create a DCM with a Qualifying Question you must associate a DVG with the DCM Qualifying Question. The system uses the Question response—which must be one of the values in the DVG—to distinguish one received DCM from another collected at the same visit. See "Configuring Qualifying Questions".
DVG Subsets For any DVG, you can create one or more subsets that include only a subset of the values in the base DVG as allowable values. When you assign a DVG to a Question, you must specify the subset number. The system allows only the active values in that subset as responses to that Question. The base DVG contains the complete set of responses that could logically be accepted as responses to a type of Question, and its subsets contain only those responses that are appropriate in certain circumstances.
For example, you could create a DVG called AE_SEV to describe the severity of an adverse event, with the values Mild, Moderate, Severe, and Life-Threatening. You could assign this DVG to a Question to be collected during the active part of the study. However, during the screening phase of the study, you could use a similar Question assigned to a subset of the DVG that did not include the value Life-Threatening.
You can also use subsets to store different layouts for DCI Forms-based layouts used in the Oracle Clinical Remote Data Capture PDF option (see "Layout Options").
See "Creating a DVG Subset" for further information.
Kinds of DVGs There are three kinds of DVGs. It is possible to assign both an alpha DVG and either an internal or thesaurus DVG to a single Question.
Internal (named to distinguish them from thesaurus DVGs, whose values originate in a table external to Oracle Clinical): This is the most commonly used type of DVG. You define the list of values in Oracle Clinical, in the Discrete Value Groups window.
Alpha> Alpha DVGs allow a data entry operator or batch data load to enter information about a response when it is not possible to enter an actual response. For example, an alpha DVG can contain values like Not Applicable, Unknown, or Not Done, which are more informative than a null response value. In addition, you can use alpha DVGs to enter alphanumeric values (such as TRACE) for Questions with data types of Date or Number, if it is not possible to enter a normal response of the correct data type. Alpha DVG values are stored in the Exception Value Text field. For each alpha DVG value, you can choose whether or not to raise a discrepancy when an alpha DVG value is entered. See "Alpha DVGs".
Thesaurus: Thesaurus DVG values are contained in external tables that you create in the Oracle Clinical database for this purpose, or that are part of a Thesaurus Management System instance in the same database. See "Thesaurus DVGs" for more information.
Alpha DVGs are designed to allow you to capture information when it is not possible to collect the expected type of response. They are most commonly used for two purposes: entering information on missing data and entering alphabetical characters for numeric lab test results. See "Missing Data" and "Alphanumeric Lab Values".
Alpha DVG values are stored in the Exception Value Text field, not the Value Text field used to store real responses.
For each alpha DVG value, you can choose whether or not to raise a discrepancy when an alpha DVG value is entered. See Create Mand Disc? in "Adding Values to a DVG".
You cannot use alpha DVG values for the Default Value field in the Question form.
Missing Data You can use alpha DVGs to allow a data entry operator to enter information about a response when it is not possible to enter an actual response value. For example, an alpha DVG can contain values like Not Applicable, Unknown, or Not Done, indicating that there is no data to enter, and the reason why.
If the Question is of data type Date or Number, or if the Question has an internal or thesaurus DVG, without an alpha DVG the data entry operator would have to leave the response blank and this information would be lost.
Even for Questions with a character data type, using an alpha DVG standardizes missing data information, making it easier to follow up.
Note:
You can specify whether or not the system should create a discrepancy whenever a data entry operator enters a particular value in an alpha DVG. This is especially useful for missing data. For example, you could choose to create a discrepancy when the value Not Done is entered, but not when the value Not Applicable is entered. See "Create Mand Disc?".Alphanumeric Lab Values Alpha DVG values, which may include alphabetical characters, can be entered as responses to Questions with data types of Date or Number. This may be useful for lab test Questions where a numeric response is expected, but the lab returns a result such as TRACE, a standard industry term that does not fit into a numeric data type field. Using alpha DVGs, you can capture the actual lab result and perform statistical analysis on the data.
In this case, only an alpha DVG is associated with the lab test Question. The normal expected response is one of a range of numeric values, not a discrete value from an internal DVG.
Alpha DVGs impact other Oracle Clinical systems as follows:
In Data Extract, you can include alpha DVG values for missing data in views, either in a separate column or in the same column as the data field (using FULL_VALUE_TEXT). For Questions based on a regular DVG, the sequence number of the alpha data value is available in data extract when the sequence number of the DVG value is specified.
In Procedures, the special processing of the Perform Detail Only If Not Null box (in the Detail Variables window) treats responses as null for the purpose of exclusion from testing, even if they have an alpha DVG value.
Adding/removing an alpha DVG to/from a Question can trigger the creation or obsolescence of mandatory discrepancies, or cause a different type of discrepancy to become obsolete or to be created. See "Applying New DVGs to Existing Data".
This section contains the following topics:
A thesaurus DVG serves the same general purpose as an internal DVG, but its values are contained in a decode table external to Oracle Clinical. Thesaurus DVGs are useful when one or more Questions have a large but fixed number of acceptable responses. You can use thesaurus DVGs with look-up tables of any kind.
You must create or import the decode table or view, or, if you have Oracle Thesaurus Management System (TMS) installed in the same instance as Oracle Clinical, you can create a view of a TMS dictionary to use as the value source (see "Using Thesaurus DVGs with TMS").
You must create a public synonym for each table and view, and grant access to each table or view to the database role RXCLIN_READ.
Note:
Oracle Clinical has no control over the decode table or view that you use as a value source for thesaurus DVGs. You can add, update, and delete values at any time. You must maintain control over the table and its data.When you define a thesaurus DVG, you specify the name of the table (or view) and column whose values are acceptable responses to Questions associated with the DVG. You can also specify up to three additional columns in the same table or view whose value the system displays during data entry for each row in the Question's list of values. The system also creates variables for these values for use in Validation and Derivation Procedures (see "DVG Variables").
Restricting Values by Study and Site
If you are using RDC Onsite for data entry, you can restrict the thesaurus DVG values displayed based on the user's current study and site context. In the decode table or view, in addition to a column for DVG values you must define one column for the Clinical Study ID and another for the Site ID, and add a WHERE clause to fetch DVG values only in rows whose Clinical Study ID and Site ID match the current user context passed by RDC Onsite.
RDC Onsite populates two variables, one for Clinical Study ID and the other for the Site ID. Two predefined functions in the rdc_thes_dvg_client package, getClinicalStudyId and getSiteId, read these variables. You can call these functions from your WHERE clause.
The WHERE clause should also include an NVL to '%' to avoid an error in Oracle Clinical, which does not populate the variable. Then when a user invokes the same thesaurus DVG from Oracle Clinical, the system returns the entire list of values.
In RDC Onsite, if no rows match the user's current study and site, the system displays the message "No values are available. Please enter your response manually." The field then accepts a manually entered response.
For example:
where to_char(clinical_study_id) like
nvl(to_char(rdc_thes_dvg_client.getClinicalStudyId), '%')
and to_char(site_id) like
nvl(to_char(rdc_thes_dvg_client.getSiteId), '%')
Thesaurus DVG Limitations Thesaurus DVGs have the following limitations, compared to internal and alpha DVGs:
You cannot define subsets of thesaurus DVGs.
Thesaurus DVGs cannot be expanded through the Oracle Clinical user interface. To add values, add rows to the table in the database.
You cannot use Enter by Sequence or Resequence.
The only possible Display Type is LOV, regardless of the layout system you are using.
The system disregards Max Value Length. Define the length in the Thesaurus DVG window, using the actual length of the table or view column.
To create a Thesaurus DVG, do the following:
In the Discrete Value Groups window, define the first fields as usual (see "Creating a DVG").
Select Thesaurus from the list of values in the field: Thesaurus or Alpha or Internal. The Thesaurus button becomes active.
Accept the default values for other fields. Do not enter any values in the lower section of the window. You can update the fields Status, Status Comment, and Retirement Reason when appropriate.
Click the Thesaurus button. The Thesaurus Values window opens.
In the Source Table/View field, enter the name of the look-up table or view in the database. The system does not validate that the table or view exists, so you can define it here before creating it in the database.
In the Term Column field, enter the name of the column whose values you want to use as DVG values. In the Length field for this and every other column, enter the length of the actual table or view column.
The Decode Table/View field defaults to the value you entered for the Source Table/View field. Accept the default.
The Term Column field under the Decode Table/View defaults to the value you entered for the Source Table/View Term column. Accept the default.
The Label 1...3 Col fields are optional. Use them to specify another column in the Decode table whose values you want the system to display in the data entry list of values to help the data entry operator decide which value to choose.
The system also populates variables with these values for use in Validation and Derivation Procedures. The system uses the Short Label as the suffix for the thesaurus DVG variable it creates for the value of this column in a Validation and Derivation Procedure (see "DVG Variables").
The system saves your definition regardless of whether an actual table or view exists that matches the definition you enter here. You can either create (or import) the table first and then define a DVG based on it, or define the DVG first and then create the table. However, if you activate the DVG, associate it with a Question in a study DCM and try to enter data for the Question before creating the actual table or view, the data entry operator receives an error message and cannot use the list of values for the Question.
If you have the Oracle Thesaurus Management System or TMS installed in the same instance as Oracle Clinical, you can use the terms contained in a single dictionary level as the values in a thesaurus DVG. All levels of all TMS dictionaries are contained in the table TMS_DICT_CONTENTS (see the TMS Technical Reference Manual for further information). To limit the values to a single level of a single dictionary—normally either the verbatim term level or the classification level—you must create a view of the table to use as the source.
You can specify up to three columns for display in the list of values to provide additional information about the term. For example, for each term in the list of values, you could specify that the system display the values in its APPROVED_FLAG column, COMMENT_TEXT column, and the user-defined VALUE_1 column.
Example Create a view of the classification level of WHO-Drug in the same instance as Oracle Clinical and TMS. Create a thesaurus DVG based on this view and specify the columns APPROVED_FLAG, COMMENT_TEXT, VALUE_1 for display, and assign it to the Question CONCOMMITANT_MEDICATION. During data entry for the Question CONCOMMITANT_MEDICATION, the operator looks for terms in the list of values that match the response on the CRF and have an approval status of Yes and appropriate comments and company-specific values.
If it is clear to the operator which term is the correct one, given the response on the CRF, he or she enters the term as a response in Oracle Clinical. When the response (verbatim term) is processed by TMS, it is autoclassified and no further action is necessary.
If it is not clear to the operator which term is correct, given the response on the CRF, he or she enters the term as it appears on the CRF. The system creates a univariate discrepancy of type Thesaurus DVG. TMS processes the response (verbatim term), and either autoclassifies it or creates an omission for manual classification. After the verbatim term is classified in TMS, the Oracle Clinical discrepancy is automatically closed during the next batch validation.
You can create a new DVG, copy an existing DVG from the Global Library, and modify DVGs.
From the Glib menu, select Discrete Value Grps, then choose (Prov) Discrete Value Grps.
To copy an existing DVG, from the Special menu, select Copy DVG. Rename and modify as necessary. See descriptions of each field below.
To create an entirely new DVG, you must:
Enter the required information the upper half of the Maintain Discrete Value Groups window or the Maintain Provisional Discrete Value Groups window (depending on your access privileges). The fields are described in this section.
For internal and alpha DVGs, enter values in the lower half of the window (see "Adding Values to a DVG"). For thesaurus DVGs, click the Thesaurus button and continue (see "Thesaurus DVGs").
You must set the DVG to Active status before you can assign it to any Questions. You must have access to the Maintain (non-provisional) Discrete Value Groups window.
The fields in the upper portion of the Maintain Discrete Value Groups window include:
Subset Number The system generates the subset number, using 0 for the base DVG, 1 for the first subset, 2 for the second, and so on.
DVG Type Select one DVG type from a list of values. These values are configurable; the Oracle Clinical system administrator can maintain the list in the installation reference codelist DISCRETE VAL GRP TYPE CODE. Use this setting to categorize the DVG for querying purposes. This setting does not trigger any enforced behavior.
Thesaurus or Internal or Alpha Select the kind of DVG from the list of values. If you select Thesaurus, skip the rest of this section and follow directions under "Thesaurus DVGs". For information on different kinds of DVGs, see "Kinds of DVGs". See also "Alpha DVGs" and "Thesaurus DVGs".
Status You must have access to the Maintain (non-provisional) Discrete Value Groups window to change the status of a DVG. There are three possible statuses:
Provisional. The default status for newly created DVGs. When a DVG is Provisional you can change any of its attributes and assign it to Provisional Questions.
When you are ready to set the Question and DVG to Active, you must set the DVG to Active before you can set the Question to Active.
You cannot assign a provisional DVG to an active Question.
Active. Change the DVG's status to Active when you are satisfied with its attributes and values and want to assign it to Questions for use in Oracle Clinical. If the DVG's Subsettable, Expandable, and Resequence attributes are selected, you can change related characteristics (see "Subsettable?", "Expandable?", and "Resequence?" below).
Note:
When you change a provisional DVG to Active, the system automatically generates Subset 1 (one) at the same time. It is a copy of the base DVG, which is Subset 0. To see Subset 1, query for it.Retired. Retired DVGs can no longer be assigned to Questions. However, Questions that already use the Retired DVG can continue to use it.
Note:
You can change a DVG's status from Active to Retired at any time.You cannot change a DVG's status from Active to Provisional if the environment is replicated; that is, if there are more than one source locations set to Active in the installation reference codelist SOURCE LOCATION CODE; or if the DVG is used by any other object:
active subsets of the DVG exist
the DVG is assigned to a question in an active question group
the DVG is assigned to an active question
the DVG is assigned to a DCM Question
the DVG is assigned to a DCM Qualifying Question
Subsettable? If selected, it will be possible to create subsets of this DVG even after the DVG is set to Active.
Expandable? If selected, it will be possible to add new values to this DVG even after the DVG is set to Active.
Enter by Seq? (Enter by Sequence Number?) If selected, the data entry operator can enter the sequence number only and the system will enter the full value.
Note:
Do not select Enter by Seq? for DVGs with numeric values; the system cannot distinguish whether the number entered is intended as a numeric value or as a sequence number.Note:
If you have both an internal and an alpha DVG assigned to a single Question, the Enter by Seq? setting for the internal DVG always overrides that of the alpha DVG:If the internal DVG's Enter by Seq? box is deselected, then data entry by sequence number is not allowed for either the internal or the alpha DVG values, regardless of the alpha DVG's setting.
If the internal DVG's Enter by Seq? box is selected, then data entry by sequence number is allowed for both the internal or the alpha DVG values, regardless of the alpha DVG's setting.
If entry by sequence number is allowed, and each of the two DVGs has a value with the same sequence number defined, the system uses the setting of the DVGSEQOVERALPHA value in the local reference codelist OCL_DE_CONFIG to determine which value to take. To avoid this situation, follow a convention of giving alpha DVG values high sequence numbers such as 90, 91, and 92.
Resequence? If selected, it will be possible to change the order of the values in this DVG even after it is set to Active.
The following fields determine the layout of the DVG in data entry layouts based on DCI Forms: Display Type, CB Layout, CB Label Source, Checked Flag Value. You can give these fields different settings in different subsets of the same DVG.
If you are using the traditional, character-based data entry layout, you can leave these fields blank. Proceed to the next field, "Upper Case?".
Display Type Choose the type of DVG to display in the default generation. These are the choices:
CBG: (Check Box Group) The list of values appears as a group of boxes. Each box is labeled with one of the values in the DVG. The boxes behave like radio buttons; the Investigator or data entry operator can select only one value. If you choose CBG, you must also choose a CB Layout value and a CB Label Source value, described below.
Flag: Use the Flag display type for DVGs with only two values. When you use a Flag field, it is important to know that the database interprets an unvisited Flag value as NULL. If you pass over a Flag field when you are entering the data in the rendered form, the database assumes the Question is unanswered and marks it NULL instead of either of the discrete values. If you choose a Flag display type, you also choose a Checked Flag Value, described below.
LOV: This option is the default value for this field. It displays the DVG as a list of values. In character-based data entry layouts, this is the only display type used.
CB Layout (Check Box Layout; available only for DVGs with a display type of CBG.) This setting determines how the boxes are aligned on the page. You can select either a Vertical or Horizontal grouping, with the labels to either the Left or Right of the boxes.
CB Label Source (Check Box Label Source; available only for DVGs with a display type of CBG.) Choose either the Value or Long Value of each value as the default prompt, or label, for each box in the group.
Checked Flag Value This parameter is active only for DVGs with a display type of Flag. The default is deselected. You can specify the default label value in this field.
Upper Case? If selected, the system forces the value to uppercase, regardless of how it is entered.
Max Value Length The default value is 15 characters. If you need longer values, you can enter a higher number, up to 80 characters.
Retirement Reason When you change an Active DVG's status to Retired, the system prompts you for a reason. Select one from a list that includes: DCI/DCM Enhancement, DVG Obsolete, Improved Version, Inaccurate Logic, New Questions Defined, New Standard Established for DCI, New Subset Created, Not Replaced, Poor Design for DCI, Redundant, Relevant Questions Retired, Replaced, Replaced—Enhancements, Replaced—Errors.
Status Comment When you change the status of the DVG, you can enter a free form text comment.
For internal and alpha DVGs, after you save the basic definition you must enter the values you want to include in the DVG.
For thesaurus DVGs, do not add values here. Thesaurus DVG values are contained in an external table. Click the Thesaurus button to continue defining the DVG. See "Thesaurus DVGs" for further information.
To assign values to an internal or alpha DVG, move the cursor to the Discrete Values section of the window. For each value, enter the following and save your work:
Seq# (Sequence Number) Determines the order in which the values appear in the list of values during data entry. See "Enter by Seq?".
Value The actual value (such as Yes, No; Male, Female; Severe, Moderate, Mild) that you are defining as an acceptable response to a Question associated with this DVG (or DVG subset). This value appears in the list of values for the Question during data entry.
Note:
If you use special characters (such as > < [ ] & -) anywhere in the DVG value, auto fill does not work during data entry for that value. That is, even if auto fill is turned on in a user's data entry configuration preferences, if that user enters ">" as the beginning of a DVG value of ">10" for example, the system does not automatically enter the rest of the value, as it should with auto fill turned on. The user must type the full value or select it from the list of values.Long Value A description that also appears in the list of values during data entry.
Active? (For use in subsets.) All subsets are created containing all the values included in the base DVG at the time they are created. To exclude a value from the subset, clear this box.
Create Mand Disc? (Create Mandatory Discrepancy?) For alpha DVG values only. If selected, the system creates a univariate discrepancy when this value is entered.
For internal and thesaurus DVGs, a univariate discrepancy is always created when a Question response is not included in the DVG's list of values. If you are using an alpha DVG as a way to collect information when an appropriate response is missing, you may want to create a discrepancy if any alpha DVG value is entered. If your alpha DVG is for a lab test Question, and the alpha DVG value (such as "trace") contains medically valid information, you may prefer not to raise a discrepancy.
When you define a new DVG, the system gives it a subset number of zero (0). Subset 0 serves as a base DVG. When you activate the DVG, the system automatically creates Subset 1, which is identical to Subset 0. In Subsets 1 and higher, you can choose to make one or more values Inactive, effectively creating a subset of the values in the base DVG. If the DVG Subsettable? box is selected, you can create additional subsets at any time.
When you assign a DVG to a Question, you must assign a specific subset of the DVG. The system then allows only the active values in that subset as responses to that Question.
Subset 0 serves as the base DVG on which all others are based, and its functionality is different from that of all other subsets:
When you add values to a DVG, you must add them first to Subset 0.
You cannot assign Subset 0 to a Question.
When you create a subset, it contains all the values included in the base DVG (Subset 0) at the time of the subset's creation. If the base DVG (Subset 0)'s Expandable? box is selected, you can add values to it. If any other subset's Expandable? box is selected, you can add values that were added to the base DVG after the creation of that subset.
You can also use subsets to store different layouts for DCI Forms-based layouts used in the Oracle Clinical Remote Data Capture PDF option (see "Layout Options").
From the Glib menu, select Discrete Value Grps, then choose Discrete Value Grps. Information about the DVG is displayed in the upper half of the window, and its discrete values, in display sequence, in the lower half.
Query and select the base DVG (Subset 0).
From the Special menu, select Create DVG Sub. The system copies the entire base DVG and assigns it the next available subset number.
Re-query to display the new DVG subset.
Clear the Active? box of each value you do not want in the subset.
For an alpha DVG, you can also change the setting of the Create Mand Discrep.? box. See Create Mand Disc? in "Adding Values to a DVG".
Save.
Create a DVG called AE_SEV (Severity of Adverse Event {non-cancer}) with the values MILD, MODERATE, SEVERE, and LIFE THREATENING. Assigning this DVG to a Question limits responses to one of those four values. The response INTERMITTENT creates a discrepancy.
| Name: AE_SEV Domain: STANDARD Subset Number: 0 | ||||
| Seq # | Value | Long Value | Active? | |
| 1 | MILD | Mild | ✸ | |
| 2 | MODERATE | Moderate | X | |
| 3 | SEVERE | Severe | X | |
| 4 | LIFE THREATENING | Life threatening | X | |
However, a particular Question might not require all four values, and making inappropriate responses available to data entry could lead to discrepancies. For example, in the Screening phase of a study, the response LIFE THREATENING should not be allowed. For this case, you could create a subset with only three Active values:
| Name: AE_SEV Domain: STANDARD Subset Number: 1 | ||||
| Seq # | Value | Long Value | Active? | |
| 1 | MILD | Mild | X | |
| 2 | MODERATE | Moderate | X | |
| 3 | SEVERE | Severe | X | |
| 4 | LIFE THREATENING | Life threatening | ||
You can assign different subsets to different Questions for use during different phases of the trial.
See "Alpha DVGs" for information about alpha DVGs. You create alpha DVGs the same way you create internal DVGs.
Date A base alpha DVG for Date Questions might have three values: Not Applicable, Unknown, and Continuing (see table below). All three values are suitable for End dates. A subset for use in Start dates might include Not Applicable and Unknown but not Continuing.
Numeric An alpha DVG for numeric fields might have five values: Unknown, Not Applicable, Not Done, and Trace. You could create a subset with only the values Unknown, Not Applicable, and Not Done for use with general Questions of data type Number. For lab Questions of data type Number you could use Subset 1, which includes all five values.
The Global Librarian or other person with access to the full-privilege window, Maintain Discrete Value Groups, is responsible for modifying and changing the status of DVGs. Anyone with access to Maintain Provisional Discrete Value Groups can modify and delete DVGs with a status of Provisional, and copy Active or Provisional DVGs.
For access to the Maintain Discrete Value Groups window, you must have one of the following roles: RXC_GL_FULL, RXC_GLA, RXC_PA, RXC_QADM, RXC_SDEF, RXC_TA, RXC_SUPER.
You can modify any attribute of a Provisional DVG.
If you have access to the Maintain Discrete Value Groups window (non-provisional), you can modify attributes of an Active DVG that are defined as modifiable; see "Subsettable?", "Expandable?", and "Resequence?".
You can copy an Active DVG and use it as a basis for a new, Provisional and therefore fully modifiable, DVG.
To copy a DVG, select the DVG and select Copy DVG from the Special menu in the (Provisional or full-access) Maintain Discrete Value Groups window.
Managing DVGs includes setting the status of DVGs, rejecting provisional DVGs, and defining sets of DVGs. Only the Global Librarian or other person with access to the full-privilege Maintain Discrete Value Groups window can change the status of a DVG to Active or Retired. When you retire a DVG, the system prompts you to fill in the Retirement Reason field.
You cannot delete an Active or Retired DVG from the Global Library. To delete a Provisional DVG, from the Glib menu, select Discrete Value Grps, and choose (Prov) Discrete Value Grps. Select the DVG and use the Delete Record and Commit functions.
You cannot delete an Active DVG.
If you have the necessary privileges, you can add a DVG to an existing Question in the Global Library; from the Glib menu, select Questions, then select Questions (see "Assigning DVGs to a Question").
After you assign a DVG to a Question in the Global Library, you can assign it to instances of the same Question contained in Question Groups and DCMs, as follows:
If the DVG Modifiable (or Alpha DVG Modifiable) box is selected in the definition of the Question in the Maintain Questions window, you can select the same box for the Question in the context of a Question. You can then add, delete or change the DVG (or alpha DVG) assigned to the same Question within a Question Group; from the Glib menu, select Question Groups, then choose Question Group Questions.
You can change from one subset to another of the same DVG regardless of the setting of the box in the parent Global Library Question.
If the DVG Modifiable (or Alpha DVG Modifiable) box is selected in the definition of the Question in the Maintain Questions window, and in the Question Group Question window on which a DCM Question is based, you can select the same box for the Question in the context of a DCM Question Group. You can then add, delete or change the DVG (or alpha DVG) assigned to the same Question within a DCM; from the Definitions menu, select DCMs, then choose DCMs, select Question Groups, and then Questions.
You can change from one subset to another of the same DVG regardless of the setting of the box in the parent Glib Question.
When you change the DVG or alpha DVG assignment to a DCM Question, the system automatically cascades the change to all subsets and layouts of the DCM.
When you change a DVG or alpha DVG subset assignment to a DCM Question, the system automatically cascades the change to all layouts of the DCM, but not to subsets of the DCM.
When you add DVGs to existing DCM Questions currently in use in a study, the next batch validation checks existing responses to the Questions with newly assigned DVGs, and creates and obsoletes discrepancies as a result, as follows:
If a response is different from any of the DVG values, the system creates a discrepancy.
If a response that does correspond to a DVG value previously raised a discrepancy, that discrepancy becomes obsolete.
If a response was discrepant before the DVG was applied, and is still discrepant afterward, the system closes the original discrepancy and opens a new univariate discrepancy of type DVG.
If you have updated one or more DVGs or DVG subsets used by Questions in a DCI DCM, you can update an existing DCI Layout to reflect the changes without generating another default layout. Run Update from the DVG Definition utility; from the DCMs window, select the Graphic Layout tab, then from the Special menu, select Graphic Layout, and choose Update from DVG Definition. The utility propagates changes to the following properties:
Active Values
Display Type Code
Checked Flag Value
Note:
The utility does not propagate any changes related to CB (Check Box) Layout. To propagate changes to CB properties, you must generate a new default layout.
|
 Copyright © 1996, 2014, Oracle and/or its affiliates. All rights reserved. Legal Notices |
|