# Generate a Case Narrative Generate case narratives based on Admin-defined templates and edit them on your local machine or directly in your browser with Microsoft 365. Vault may automatically generate narratives when you create a _Case_ or move a _Case_ to a specific lifecycle state. If configured by your Admin, you may also be able to manually generate and update a narrative. You can review all versions of the narrative document and, if applicable, [edit][5] the document in your _Library_. In Vaults using Veeva AI for Safety, you can use the [Narrative Agent][10] to proofread and rewrite the narrative content. Depending on your Admin's configuration, Vault may generate narrative documents in the following ways: * [Rules based narratives][2] * [Narrative templates][3] ## Prerequisites Depending on your business needs, your Admin may need to complete the following: * Configure Rules Based Narratives * Configure Narrative Templates * Enable the Narrative Agent ## About Narrative Generation The content included in generated narrative documents depends on your Admin's configuration. Upon generation, Vault populates certain fields on the [_Case_][11] and [document][1]. You can access the generated DOCX narrative file in your _Library_ and directly from the _Narrative_ section of the _Case_ or _Localized Case_. Whenever you bulk unblind all _Cases_ in a _Study_, Vault generates two minor document versions. The first copies the previous _Case's_ narrative. The second appends the _Narrative Text_ value to the end of the document. ### Case Fields {#case-fields} Upon narrative generation, Vault populates fields in the _Narrative_ section of the _Case_. Vault populates the _Narrative Preview_ field on the _Case_ with a plain text preview of the narrative document. This preview shows how the narrative will be exported in generated regulatory reports, such as CIOMS I, FDA 3500A, and E2B. This field can contain up to 32,000 characters. If the content includes a table, Vault replaces the table dividers with commas in the preview. If your Admin has configured blind protection on narrative documents, does not display any sensitive information.

Note: Vault does not generate a value for the Narrative Preview field if the narrative document contains invalid characters that do not render. If applicable, we recommend updating or removing invalid characters in your narrative document so Vault populates this field.

### Document Fields {#doc-fields} Upon narrative generation, Vault populates the following fields on the generated document, which you can review in the the _Case_ section of the _Document Information_ panel on the Doc Info pane: * **Narrative Content Source**: This field specifies the content source. Vault updates this field with every version check-in: * **Imported from Source**: Content copied from the source by importing or promoting an _Inbox Item_ or creating a follow-up _Case_. * **Rules Based Narrative**: Content generated by running the [_Generate Narrative from Outline_ action][2] on a _Case_. * **Narrative Agent**: Content generated by Veeva AI by running the [_Generate Text with Narrative Agent_ action][10] on a narrative document. * **Blank**: Content generated from a narrative template or manually added by a Vault user from uploading or checking in a new document version. * **Case Access Group**: If your Admin has configured Case Access Group Security, Vault populates this field based on the corresponding value of the associated _Case_. You cannot modify this field value. * **Blind Protection**: If your Admin has configured blind protection on narrative documents, Vault populates this field with _On_ and unauthorized users cannot view the narrative. When Vault generates a narrative for a follow-up _Case_, the new narrative document inherits the _Blind Protection_ value from the narrative of the previous _Case_ version unless the follow-up _Case_ was unblinded during processing. In this scenario, Vault always sets the _Blind Protection_ field to _On_. You can adjust blind protection for the narrative by updating this field value. ## Rules Based Narratives {#rules} When you run the _Generate Narrative from Outline_ action on a _Case_, Vault generates a narrative based on Admin-defined outlines that include specific elements of a _Case_. These narratives are grammatically correct, logical, orderly, and complete, while requiring minimal editing. Your Admin can configure outline templates for different [types of _Cases_][6] and optionally for different locations and organizations. Depending on your Admin's configuration, this action may run automatically when you create a _Case_ or move a _Case_ to a specific lifecycle state. If configured, you can manually run this action by selecting it from the **All Actions** menu of a _Case_. In Vaults using Veeva AI for Safety, the [Narrative Agent][10] may run automatically, or you can run it manually, depending on your Admin's configuration. Whenever you run the action on a _Case_ that already has an associated narrative document, Vault re-renders and upversions the existing narrative document based on the applicable outline template, including a [follow-up summary][12] of changes since the generation of the previous narrative. You can review any changes to follow-up statements on the _Case Version Compare_ page. The following behavior applies to generating rules based narratives: * For _Cases_ promoted from imported _Inbox Items_ with E2B narrative content, you can review the source E2B narrative in the previous version. * When you create a follow-up _Case_, Vault generates a new minor narrative document version by copying the previous _Case's_ narrative. If the previous _Case_ does not have a narrative, Vault generates a new narrative document for the follow-up _Case_. When configured by your Admin, Vault also generates a new minor document version of the narrative for the follow-up _Case_ based on the applicable outline template. ### Types of Cases {#types} Vault matches the appropriate Admin-defined outline template to _Case_ data for the following types: * **Post-Market Non-Serious**: Applies to the following _Cases_: * Postmarket _Cases_ (do not have _Studies_) * Non-clinical trial study _Cases_ (_Study Type_ is _Other_) with a _Serious_ value of _No_ or blank * **Post-Market Serious**: Applies to the following types of _Cases_: * Postmarket _Cases_ (do not have _Studies_) * Non-clinical trial study _Cases_ (_Study Type_ is _Other_) with a _Serious_ value of _Yes_ * **Clinical**: Applies to clinical trial study _Cases_ (_Study Type_ is not _Other_) ### Follow-Up Summary {#summary} Upversioned narratives include a follow-up summary of changes to the _Case_ between major versions, including updates to the following: * Patient details * Added studies * Adverse events and changes to seriousness * Suspect product details * Assessment results * Drug histories * Treatment drugs * Medical histories * Narrative text for unblinded studies Vault extracts summary statements based on specific [change conditions][13], each listed with the _New Info Data_ (`DD-Mon-YYYY`) value of the current _Case_ version to indicate when the change occurred. For example, the follow-up summary would include the following statement when a _Case Product_ with a _Drug Role_ of _Suspect_ has been deleted: `27-Oct-2025: The suspect product Cholecap was removed from the case.`

Note: Vault does not include a follow-up summary of changes in narrative documents for Cases generated using the override merge method for automated Case promotion.

#### Change Conditions {#conditions} Vault extracts follow-up summary statement outputs based on the following change conditions:

Object	Change Condition	Summary Statement Output
Case	Any of the following field values change: Age Group Age Value Age (Normalized) Date of Birth (Normalized) Sex Pregnancy Outcome Pregnancy Due Date Last Menstrual (normalized)	`The patient details were updated, including [fields].`
Case	The Study field value changes	`The study [Study] was added.`
Case Adverse Event	New record added and Seriousness value of the Case does not change	`The [Event Reported] was added.`
Case Adverse Event	Existing record deleted	`The [Event Reported] was removed.`
Case Adverse Event	The Seriousness value of the Case does not change and any of the following field values change: Cessation (normalized) Duration Event (LLT) Event (Reported) Event Country Onset (normalized) Outcome	`Details of adverse event [Case Adverse Event], [Event Reported] were updated, including [fields].`
Case Adverse Event	The populated Seriousness value of the Case changes and none of the following field values change: Cessation (normalized) Duration Event (LLT) Event (Reported) Event Country Onset (normalized) Outcome	`Details of adverse event [Event Reported] were updated with seriousness of [Seriousness].`
Case Adverse Event	The populated Seriousness value of the Case changes and any of the following field values change: Cessation (normalized) Duration Event (LLT) Event (Reported) Event Country Onset (normalized) Outcome	`Details of adverse event [Event Reported] were updated with seriousness of [Seriousness], including [fields].`
Case Adverse Event	The Seriousness value of the Case changes from blank to populated, the Seriousness value of the Case Adverse Event changes or new Case Adverse Event with populated Seriousness value added, and any of the following field values change: Cessation (normalized) Duration Event (LLT) Event (Reported) Event Country Onset (normalized) Outcome	`Details of adverse event [Event Reported] were updated with seriousness of [Seriousness], including [fields]. As a result, the case was upgraded to a Serious case.`
Case Adverse Event	The Seriousness value of the Case changes from blank to populated, the Seriousness value of the Case Adverse Event changes or new Case Adverse Event with populated Seriousness value added, and none of the following field values change: Cessation (normalized) Duration Event (LLT) Event (Reported) Event Country Onset (normalized) Outcome	`Details of adverse event [Event Reported] were updated with seriousness of [Seriousness]. As a result, the case was upgraded to a Serious case.`
Case Adverse Event	The Seriousness value of the Case changes from populated to blank, the Seriousness value of the Case Adverse Event changes, and any of the following field values change: Cessation (normalized) Duration Event (LLT) Event (Reported) Event Country Onset (normalized) Outcome	`Details of adverse event [Event Reported] were updated, including [fields]. As a result, the case was downgraded to a Non-serious case.`
Case Adverse Event	The Seriousness value of the Case changes from populated to blank, the Seriousness value of the Case Adverse Event changes, and none of the following field values change: Cessation (normalized) Duration Event (LLT) Event (Reported) Event Country Onset (normalized) Outcome	`Details of adverse event [Event Reported] were updated. As a result, the case was downgraded to a Non-serious case.`
Case Product¹	Drug Role value changes	`The reported product [Product] updated from a [Drug Role] to a [Drug Role] product.`
Case Product¹	Existing record with a Drug Role of Suspect deleted	`The suspect product [Name] was removed from the case.`
Case Product¹	New record added with a Drug Role of Suspect, or the Drug Role of an existing record is Suspect and any of the following field values change: Case Product: Action Taken Product Product (Coded) Product (Reported) Product (Study) Case Product Dosage: First Administration (normalized) Last Administration (normalized) Duration (number) Dose (unit) Patient RoA Patient RoA Text	`Details of suspect product [Name] were updated, including [fields].`
Case Product¹	New record added with a Drug Role of Treatment, or the Drug Role of an existing record is Treatment and any of the following field values change: Case Product: Action Taken Product Product (Coded) Product (Reported) Product (Study) Case Product Dosage: First Administration (normalized) Last Administration (normalized) Duration (number) Dose (unit) Patient RoA Patient RoA Text	`Details of treatment drug [Product] were updated, including [fields].`
Case Drug History	New record added or any of the following field values change: Start Date Normalized End Date Normalized Drug (Reported) Drug (Coded) Indication (LLT) Indication (Reported)	`Details of drug history were updated, including [fields].`
Case Medical History	Any of the following field values change: Start Date Normalized End Date Normalized Condition / Procedure (Reported) Condition / Procedure (PT)	`Details of the medical history [Event (Reported)] were updated, including [fields].`
Case Medical History	New record added	`Medical history of [Event (Reported)] added.`
Case Assessment Result¹	Any of the following field values change: Assessment Result Result	`The [Source Type] assessment [Name] was updated to [Assessment Result].`

Note: 1.In Vaults configured to isolate blinded clinical trial information, follow-up statements for Case Products and Case Assessment Results do not include Case Products with Unblinded as the Blinding Type.

### Limitations Vault can reference up to 15,000 child _Case Products_ and 15,000 _Case Medical Histories_ per Admin-defined _Narrative Statement Option_. If executing a particular record would exceed this amount, Vault ignores that option when generating the narrative and sends you a notification with details about which record would have exceeded this limitation. ## Narrative Templates {#templates} When you promote an _Inbox Item_ to a _Case_, Vault generates a narrative based on Admin-defined controlled document templates using merge fields that generate values based on token resolution. Your Admin can configure custom templates for any combination of _Organization_, _Report Type_, _Study Type_, _Study_, and _Localization_. The following behavior applies to generating narratives from narrative templates: * For _Cases_ promoted from imported _Inbox Items_ with E2B narrative content, the generated narrative includes imported narrative text and translations. * When you create a follow-up _Case_, Vault re-renders the narrative from the initial _Case_, creating a new [follow-up narrative document][7] version for the follow-up _Case_. If the previous _Case_ does not have a narrative, Vault generates a new narrative document for the follow-up _Case_. * For follow-up _Cases_ with imported E2B content, Vault appends the content in the H.1 Case Narrative Including Clinical Course, Therapeutic Measures, Outcome and Additional Relevant Information data element to the end of the narrative. * If you merge a local _Inbox Item_ to a follow-up _Case_ and the language for the _Inbox Item's_ narrative content and initial _Case_ are the same, Vault copies the narrative content from the initial _Case_ to the new narrative document and appends any [localized narrative][8] content to the new document. * In Vaults using Veeva AI for Safety, you can manually run the [Narrative Agent][10] on generated narratives after medical review and QA, if configured by your Admin. ### Follow-Up Narratives {#follow-up-narratives} When you open a follow-up _Case_, Vault copies over the most recent version of the narrative document from the initial _Case_. If the most recent version contains merge fields, Vault updates the field values in the newest version using the information in the follow-up _Case_. To include the initial _Case_ narrative without regenerating merge field values, perform the following steps: 1. Navigate to the initial _Case_ narrative document. 2. Copy the text from the initial narrative. 3. Paste the contents into the narrative document for the follow-up _Case_ using plain text without formatting. You must paste with plain text to ensure Vault does not regenerate merge field values. ### Localized Narratives {#localized-narratives} Vault generates narrative documents for the translation of _Localized Cases_, which includes the English narrative that exists for the related _Case_ when you generate the _Localized Case_. You can review these details and access the narrative document in the _Narrative_ section. You can override a _Localized Case's_ narrative content by selecting the **Sync Global Narrative to Local Narrative** action from the record's **All Actions** menu. This updates the narrative document, preview, and language with the most-recent version from the related global or domestic _Case_. For follow-up domestic _Cases_ generated through the _Inbox Item to Case Compare_ page, if the previous _Case_ version doesn't include a _Localized Case Narrative_ or the _Localization_ of the _Case_ has changed, Vault generates a narrative and attaches it to the _Localized Case_. To generate the narrative, Vault uses the local narrative template for the _Localization_. If no local template exists, Vault generates the narrative from the default narrative template. ### Re-Render the Narrative Document {#re-render} After adding or changing _Case_ details, you can run the _Re-render Case Narrative_ action to update merge fields on the narrative. To do this: 1. Navigate to the _Case_. 2. From the **All Actions** menu, select **Re-render Case Narrative**. 3. In the _Re-render Document_ dialog, select **OK**. When you re-render the narrative document, Vault does not create a new document version. Instead, Vault updates merge fields in the most-current version of the document to reflect the new details. ### Lock Merge Fields {#lock} Before you distribute a _Case_ to health authorities, we recommend moving the narrative document to a steady-state to save the narrative to a specific version in time. This prevents Vault from updating merge field values in the narrative document if any of the associated fields on the _Case_ change, such as non-significant follow-ups or amendments. If you need to update the narrative document after distribution to partners or health authorities, you can create a new version of the document. Use the narrative document's workflow to move it to a steady state, such as _Approved_. ## Narrative Agent In Vaults using Veeva AI for Safety, use the Narrative Agent to refine narrative content. The agent can fix spelling and grammar and improve the narrative's readability by updating the overall flow, style, and voice. The Narrative Agent actions can: * **Proofread**: Corrects spelling and grammar issues while preserving the narrative's original structure and details. * **Rewrite**: Refines flow, style, and voice. This may adjust the document structure but should not modify the clinical course of the narrative. Depending on your Admin's configuration, Vault may run these actions automatically during rules based narrative creation for initial _Cases_, follow-up _Cases_, or both. If configured by your Admin, you can also run these actions from any narrative document's **All Actions** menu.

Note: For content generated by narrative templates with merge fields, narrative documents generated by the Narrative Agent do not maintain any merge field tokens inherited from the template, so you cannot re-render the narrative after running the agent to update the data points. We recommend waiting to run the agent until later in the Case’s lifecycle, after medical review and QA.

To manually run the Narrative Agent on a document: 1. Navigate to the most recent version of the narrative document classified as _Case > Summary > Narrative_. 2. From the **All Actions** menu, select **Generate Text with Narrative Agent**. 3. In the dialog, select **Proofread** (

) or **Rewrite** (

). The dialog indicates that Vault is working on your selection. 4. You can select **Close** or wait for the dialog to close when the action completes. Closing the dialog does not stop the selected action from executing. When the action completes, Vault upversions the document to the next minor version, which contains the AI-generated content. Review the generated content to verify accuracy. If applicable, edit the document as needed. When the action completes, Vault also does the following: * Updates the _Narrative Generation Status_ field on the _Case_ to indicate the current status. * Updates [_Narrative Content Source_][1] document field value to _Narrative Agent_. * Preserves the previous narrative content in the document's previous minor version. * For follow-up _Cases_, includes a change summary in the narrative content for each version. * If Vault encounters any errors, updates the document's lifecycle to _Generation Error_ state.

Note: Always review the generated narrative content after running the Narrative Agent.

## Edit the Narrative Document {#edit} You can edit generated narrative documents using collaborative authoring. ### Recommendations and Best Practices {#best-practices} We recommend the following best practices when editing a narrative document: * Do not add information that should be blinded to the narrative document. * Do not add comments or track changes in the document. If you need to use these review tools for draft purposes, ensure you remove them for the final version. * To copy-paste content with [merge field tokens][3], paste without formatting. If you paste with formatting, the tokens will regenerate if there are changes to the associated field. * Do not insert images into the document. * If you must insert tables, only use single-column tables. Multi-column tables cannot be exported to transmission documents. When tables are exported, they are flattened to comma-separated text values. * Do not archive narrative documents for any _Cases_ that may receive follow-ups. Vault cannot transmit archived documents.

Note: If you see merge field queries in the narrative, contact your Veeva Representative or your Admin. There may be a configuration issue on the narrative template.

## Add a Narrative Translation {#translation} For _Cases_ based on imported E2B files, Vault imports all narrative translations from the E2B file. To manually add a narrative translation: 1. Navigate to the applicable _Case_. 2. In the _Documents_ section, select **Add**. 3. In the _Search: Documents_ window, select **Create**. 4. In the _Create Document_ dialog, select **Upload**, then **Continue**. 5. On the _Upload Files (Step 1)_ page, upload the applicable DOCX file and classify the document as **Case › Summary › Narrative**. 1. Select **Next**. 2. On the _Upload Files (Step 2)_ page, populate the following fields: * **Language (ISO)** * **Organization** 3. Select **Save**. 4. In the _Narrative_ section of the _Case_, select the document to which you want to add this translation to navigate to the document's Doc Info pane. 5. Add the narrative translation you saved to the _Supporting Documents_ section of the _Relationships_ panel. [1]: #doc-fields [2]: #rules [3]: #templates [5]: #edit [6]: #types [7]: #follow-up-narratives [8]: #localized-narratives [10]: #narrative-agent [11]: #case-fields [12]: #summary [13]: #conditions