The Communications page in Rad AI Continuity centralizes automated messages sent to patients and providers about imaging follow‑up. You can see what’s been sent, what’s scheduled, and what needs reviewing—all in one place.
In this article:
- Overview
- Prerequisites
- How it works
- Viewing and understanding the Communications page
- Using the Communications page
- Frequently asked questions
Overview
The Rad AI Continuity workflow starts when a finalized imaging report is received. Continuity analyzes the report for follow‑up findings, creates an entry for navigators to review, then tracks the resulting order, scheduling, and completion steps over time.
At several of these stages, Continuity can send automated notifications and reminders to patients and providers to help close the follow‑up loop.
In Communications, you can:
- View all messages in one place—see those that are pending review, approved, sent, failed, and rejected for both patients and providers.
- Review and approve messages before they’re sent (if manual review is enabled for your organization). Learn more about manual review versus auto-approval.
- Monitor message status and timing, including when a message is expected to send. Learn more about message statuses and timings by message type.
- Handle failures and manual workflows by copying message content, contacting patients and providers outside of Continuity (via fax or phone, for example), and logging what was done for audit purposes. Learn more about manually sending failed or rejected messages.
Prerequisites
Before you can manage patient and provider communications, make sure:
-
Communications has been implemented for Rad AI Continuity at your site.
The Rad AI Implementation team will work with your organization to configure how patient and provider communications are triggered and delivered based on your EHR setup.
-
You’re logged in to Rad AI Continuity. Learn how to log in.
- You have permission to view communications. Access to Communications is limited to specific roles, such as navigators and admins.
For help with enabling, configuring, or getting permission to view communications, contact your Customer Success Manager or support-continuity@radai.com.
How it works
The Rad AI Continuity workflow
Here’s how communications fit into the Rad AI Continuity workflow at a high level:
- Rad AI Continuity receives an ORU message from your EHR/RIS system.
- Rad AI Continuity analyzes the ORU and—if findings are detected—creates an entry in our system.
- A navigator at your organization reviews the follow-up recommendations, then approves or rejects the entry. Learn more about the recommended navigator workflow.
- Rad AI Continuity determines whether patient or provider communication is required, based on rules established with your organization during implementation. If communication is required, a message gets added to the communications table. Examples of supported message types include:
- Notifying a provider that a new follow‑up recommendation was found.
- Reminding a provider to approve an outstanding follow‑up order.
- Notifying a patient that they need to schedule imaging.
- A navigator at your organization reviews pending communications, then approves or rejects the message. Alternatively, you may choose to enable auto-approval for your organization, in which case manual review is not required. Learn more.
Manual review versus auto-approval
Your organization can choose to handle patient and provider communications in two distinct ways:
Manual review
- Messages are created in the communications table with the status "Pending review".
- A user must open each message, then "Approve" (send on schedule) or "Reject" (do not send).
- Best for onboarding, testing, or higher‑risk communication types where extra oversight is desired.
Auto‑approval
- Messages that meet your rules are automatically marked "Approved" and sent on their scheduled date.
- Users check the communications table to monitor, handle failures, and manage exceptions. Message approval can be undone if it’s before the scheduled send date. Learn more.
- Best for ongoing, high‑volume programs where you want to minimize manual work.
Configuration is set with Rad AI during implementation and can be updated later via your Customer Success Manager or support-continuity@radai.com.
Viewing and understanding the Communications page
To view the Communications page, select Communications from the main navigation menu at the top.
Each row in the table represents a single message to a patient or provider, and shows the message type, intended recipient, delivery channel, and the current status: pending review, approved, sent, failed, or rejected. Learn more about message statuses.
Recipient types
There are two main types of communication recipients in Rad AI Continuity: patients and providers. Each type has several communication channels that can be enabled or disabled according to your organization's preferences.
Patient communications
Patient communications are delivered via mail or EHR, the latter limited to the patient’s Epic MyChart. See communication channels for more details.
Provider communications
Provider communications are delivered via mail or Epic EHR, the latter limited to the provider’s Epic InBasket. See communication channels for more details.
Rad AI Continuity determines which provider to contact and how to contact them based on patient class:
| Patient class | Contacted provider |
|---|---|
| Outpatient | Ordering |
| Inpatient/ED | PCP (Primary Care Physician) |
| Unknown | Ordering |
Rad AI Continuity then determines how to contact the provider, based on whether they’re integrated with your Epic system or not:
| Epic integrated? | Contact method |
|---|---|
| Yes | Automatic (Epic InBasket) |
| No | Manual (e.g., phone call) |
Message types
Patient message types
The following message types and timings are defaults that can be adjusted to suit your organization’s needs.
| Message type | Description | Default timing |
|---|---|---|
| Finding detected | Notifies the patient that they have a follow-up recommendation from their recent exam |
14 days after the report with a detected finding was signed |
| Exam scheduling | Notifies the patient that they need to schedule a follow‑up exam once an order has been approved |
3 days after the follow‑up order has been approved |
| Exam scheduling reminder | Reminds the patient that they need to schedule a follow-up exam |
30 days before the exam due date |
| Exam scheduling overdue | Notifies the patient that they are overdue for their follow‑up exam | 30 days after the last allowable due date within the recommended range |
| Missed exam | Notifies the patient that they have missed a scheduled follow-up exam. At this point a navigator may contact the patient manually to reschedule their exam. |
7 days after the scheduled exam date |
Provider message types
The following message types and timings are defaults that can be adjusted to suit your organization’s needs.
| Message type | Description | Default timing |
|---|---|---|
| New order | Notifies the provider of a detected finding and the need for a follow-up order |
7 days after the report with a detected finding was signed |
| Order reminder | Reminds the provider that they need to approve a follow-up order |
14 days before the exam due date |
Message statuses
The message status shown in the Communication table helps you track messages and determine what action needs to be taken (if any).
| Message status | Description |
|---|---|
| Pending review | The message is created and scheduled, but still waiting for a user to review and approve/reject. |
| Approved | The message has been reviewed and approved, and will be sent automatically on its scheduled date and time. |
| Sent | The message was successfully sent via the configured communication channel (e.g., EHR, mail). |
| Action required | The message could not be sent due to an issue that needs your attention. The navigator should review the message and complete any missing information. |
| Failed | The system tried to send the message but did not succeed (e.g., due to an EHR error). The navigator needs to follow up manually. |
| Rejected | The message was reviewed by a user and explicitly declined; it will not be sent automatically. |
Communication channels
The communication channel is the method by which the message is sent. It depends on the recipient type, EHR, and how Rad AI Continuity is implemented at your organization.
| Recipient type | EHR type | Communication channel |
|---|---|---|
| Patient | Epic | EHR (Epic MyChart only) |
| Any EHR | Print mail | |
| Provider | Epic | EHR (Epic InBasket only) |
| Any EHR | Print mail |
Using the communications page
Filtering by message status
You can filter the communications table by message status: pending review, approved, sent, failed, or rejected. Learn more about message statuses.
- Click the drop-down next to Filter by message status at the top of the table.
-
Select the statuses you want to show and deselect the statuses you want to hide. The table will update automatically with each selection/deselection.
⚠️ Filters reset to default when you navigate away from the Communications area. “Pending review” is the default filter.
Sorting by property
You can sort the communications table by a specific property, in ascending or descending order.
In the communications table, click a column header to sort by that property. Each click toggles between ascending and descending order (the initial direction may vary by property).
Reviewing pending messages
If your site uses manual review, messages will first appear in the communications table with the status “Pending review”. Messages will not be sent until you review and approve them. Learn more about manual review versus auto-approval.
To review messages:
-
Filter the table by message status to show only messages that are “Pending review”.
- Click a message to open the “Preview” dialog, which shows the following information:
-
A preview, based on a template, that shows the message content as presented to its recipient.
⚠️ If the message recipient is a provider, the preview will only appear once the provider is selected (see step 3 below).💡 To edit the message template, contact your Customer Success Manager or support-continuity@radai.com. - A message overview, including information such as the communication channel, corresponding entry status, category (e.g., Fleischner), subcategory (e.g., < 6mm single part-solid), and message send date.
-
Patient information, including the patient’s name and Medical Record Number (MRN).
💡 Click View patient details to open the Patient View in a new tab. Learn more about the Patient View. -
Recipient information, including the recipient type and recipient name.
💡 (If enabled for your organization) Click View provider details to open the Providers page in a new tab, where you can edit provider contact information. Learn more about editing provider contact information.
-
-
(If message recipient is a provider) Under Recipient information, click the Recipient drop-down to select a provider. Rad AI Continuity will recommend a provider based on patient class. Learn more.
Alternatively, you can type into the field to search for a provider by name or National Provider Identifier (NPI). Once found, click their name to add them as the recipient.⚠️ The selected recipient will be applied by default to all future communications related to this entry. - Review all the information and do one of the following:
-
If the information looks correct, click Approve. The message status will update to “Approved”, and Rad AI Continuity will attempt to send the message on the scheduled send date and time.
⚠️ If the scheduled send date/time is in the past, Rad AI Continuity will attempt to send the message as soon as possible. -
If the message should not be sent, click Reject. The message status will update to “Rejected” and will not be sent.
💡 Messages to providers who are not integrated with your EHR will fail to send. In these cases, you may wish to pre-emptively reject the message, then follow-up by sending a message manually.
-
Undoing message approval/rejection
If a message was approved or rejected in error, you can revert it back to “Pending review” status.
-
Filter the table by message status to show only messages that are “Approved” or “Rejected”.
- Find the message you want to undo and click it to open the “Preview” dialog.
- Click the Undo Approval or Undo Rejection button. The message status will revert to “Pending review”.
Manually sending failed or rejected messages
Once a message is approved (or auto‑approved), Rad AI Continuity attempts to send it on the scheduled date and time, based on the message type’s default timing. It’s recommended that navigators regularly review outbound communications to catch any delivery failures.
Messages that cannot be sent automatically are marked “Failed” and require manual action. Examples include:
- Providers who are not integrated with your EHR.
- Providers who exist in the EHR, but do not have Epic InBasket configured.
- The EHR cannot accept or deliver the message due to configuration or data issues.
To monitor and manually follow up on failed or rejected messages:
-
Filter the table by message status to show only messages that are “Failed” and “Rejected”.
- Click on any message marked “Failed” or “Rejected” to view details in the “Preview” dialog.
- Review the previewed message content on the left. Then either:
- Call the patient/provider and deliver the message over the phone.
- Highlight and copy the message content, then paste it into your preferred communication channel (e.g., email, fax) and send.
- Call the patient/provider and deliver the message over the phone.
- Log the communication:
- In the Preview dialog, click View full details next to the Patient information section on the right. This will open the Patient View.
- In the Patient View, click the entry’s three-dot icon under the Actions column then click either Log patient contact or Log provider contact.
- Select the type of contact you had with the patient/provider, enter a comment, then click Save. The log will appear under the History tab in the Patient View.
- In the Preview dialog, click View full details next to the Patient information section on the right. This will open the Patient View.
Frequently asked questions
Is fax an option for provider communications?
Not currently, but we’re working on adding fax as an option in 2026.
Do we have to manually approve every message?
No, manual approval is optional. During implementation, you can choose to:
- Require navigators to manually review messages before they’re sent, or
- Enable auto‑approval, which allows Rad AI Continuity to send messages automatically based on your configured rules.
Manual review is often useful during onboarding and testing. After go‑live, some customers switch to auto‑approval to reduce manual workload—while still allowing you to monitor communications.
How many times does Rad AI Continuity attempt to send messages?
Rad AI Continuity will attempt to send messages up to three times (once per day for three days) before marking it “Failed”, at which points it requires manual follow-up by a navigator.
What’s the difference between “action required” and “error” statuses?
Action required means the message could not be sent, but the issue can likely be fixed through user intervention. Review the message to complete any missing information, or contact Rad AI for help so the message can be sent.
Error means the message failed to send after multiple attempts. In this case, the system cannot retry automatically and the message should be sent through a manual channel. Learn more.
Can we change the timing or templates for messages?
Yes. The types of messages, whether they’re enabled, and their send timings are chosen during implementation and can be configured to match your organization’s needs. Templates for patient and provider messages are also configurable so they align with your organization’s clinical and communication policies.
To request changes, contact your Customer Success Manager or support-continuity@radai.com.
Can I filter the Communications table by message type, due status, or channel?
Not currently, but you can sort the table by property, in ascending or descending order.