Scheduling is where patients, revenue and capacity meet, and in Epic organizations that means Cadence. Connecting outside apps, call centers, AI agents or analytics tools to Cadence can unlock self-scheduling, fewer no-shows and better capacity use. It can also create double bookings and frustrated staff if done carelessly. This guide covers the APIs, HL7 feeds and automation options available, and how to choose between them. For broader Epic connectivity, see our Epic integration services. Planning a scheduling integration? Book a free architecture review.
How Cadence Scheduling Data Flows
Cadence manages provider schedules, visit types, templates, slots and appointments across an Epic organization. Outside systems can read and sometimes write this data through FHIR APIs, receive it through HL7 feeds, or interact through the user interface. Each route has different capabilities, approvals and risks.
What Cadence controls
Cadence holds provider templates, visit types, durations, scheduling rules, locations and appointment statuses. These rules are built by each health system’s Epic team. Any integration must respect them, because booking outside the rules can create visits that clinics cannot actually staff or bill correctly.
FHIR read access to appointments
Epic exposes appointment data through FHIR APIs, allowing approved apps to read upcoming and past visits for a patient. This suits reminder tools, patient apps and care coordination dashboards. Read access is usually easier to approve than booking access, so many projects start here.
Find and book operations
For booking, Epic supports Appointment find and book operations based on the Argonaut Scheduling implementation guide, which carry STU3 roots. These operations only work after the health system configures which visit types, providers and departments are open to outside booking in Cadence.
HL7 SIU scheduling feeds
Many organizations send scheduling events from Cadence as HL7 SIU messages, covering new bookings, reschedules, cancellations and no-shows. Downstream systems such as reminder platforms, transport services and analytics tools subscribe to these feeds through an interface engine, staying synchronized without polling APIs.
Why customer build matters
Two Epic organizations can expose very different scheduling capabilities. One may open dozens of visit types for online booking, while another allows only reads. Always confirm what your target health system has configured before designing features, because the build decides what an integration can do.
Common Cadence Integration Use Cases
Most Cadence integrations aim to reduce manual scheduling work, improve access for patients or recover lost capacity. Choosing a clear use case first keeps scope realistic and helps health system Epic teams prioritize approvals, which is often the slowest part of any scheduling project.
Partner and patient self-scheduling
Digital health companies and patient apps often want to book visits directly into Epic. Using find and book operations, an app can show real open slots and confirm appointments instantly, provided the health system has enabled those visit types for outside scheduling.
Referral and intake workflows
Referral management tools can read appointment status to confirm when referred patients are scheduled and seen. This closes the loop between referring providers and specialists. Combined with document exchange, it reduces phone calls and helps practices track referral leakage across their network.
Reminders and no-show reduction
Reminder platforms use SIU feeds or FHIR reads to send texts, calls and emails before visits, and to capture confirmations or cancellations. Our patient no-show reduction software work shows how these tools recover capacity when integrated correctly with scheduling data.
Voice and AI scheduling agents
AI agents can answer calls, verify patients and book or reschedule visits through Cadence APIs. They need strict guardrails to respect scheduling rules. Our AI appointment confirmation agents show how automation can reduce front-desk workload without bypassing clinic policies. Every agent action should be logged for review.
Capacity and access analytics
Analytics teams combine Cadence data with visit outcomes to measure access, template utilization and lag times. Scheduling feeds or reporting extracts power these dashboards. Better visibility helps operations leaders adjust templates, add capacity where demand is high and reduce wasted slots.
APIs, HL7 or RPA: Choosing the Right Approach
There are three ways to connect to Cadence: FHIR APIs, HL7 feeds and robotic process automation working through the Epic interface. Each suits different needs. Choosing the wrong method leads to fragile integrations, extra costs or approval delays that stall the project for months.
When to use FHIR APIs
Use FHIR APIs when an app needs real-time reads or bookings for individual patients, such as self-scheduling or patient-facing tools. APIs are structured, documented and supported by Epic. Expect approval steps, scope limits and customer build requirements before booking features go live.
When to use HL7 SIU feeds
Use SIU feeds when systems need to stay synchronized with all scheduling events across departments, such as reminders, transport or analytics. Feeds push changes as they happen. They require an interface engine and coordination with the health system’s integration team to configure.
When RPA still appears
Some organizations use RPA bots to automate scheduling tasks inside Hyperspace when APIs or feeds are unavailable, such as bulk rescheduling or work queue processing. Bots are faster to start but fragile across upgrades. Treat them as a bridge, not a permanent architecture.
Combining approaches
Many mature setups combine methods: SIU feeds keep systems synchronized, APIs handle patient-level booking, and limited automation covers edge tasks. Designing this mix intentionally avoids duplicate data and conflicting updates. Our healthcare RPA services help define where automation fits. Document which system owns each update.
Steps to Implement a Cadence Integration
Scheduling integrations touch patients directly, so mistakes are visible immediately. The steps below reflect how we deliver Cadence projects, putting health system alignment and testing ahead of development. Each step produces a reviewable result that clinic operations, IT and compliance teams can approve.
Step 1: Define use case and scope
Agree on the exact workflow, such as online booking for new primary care patients at three clinics. List visit types, departments, providers and patient groups involved. Clear scope speeds up health system approvals and keeps Cadence build changes small and manageable.
Step 2: Confirm build and access
Work with the health system’s Epic team to confirm which APIs, scopes and visit types are available, and what Cadence configuration is needed. Request sandbox or test environment access early, because approval and build timelines usually drive the overall project schedule.
Step 3: Design rules and edge cases
Map how your app handles new versus existing patients, insurance requirements, visit type selection, cancellations and time zones. Document how conflicts are resolved. Edge cases cause most scheduling failures, so designing them upfront prevents double bookings and incorrect visit types after launch.
Step 4: Build and test thoroughly
Develop against the test environment, then run scenarios covering booking, rescheduling, cancellations, full schedules and network failures. Involve schedulers in testing, because they know where real-world workflows differ from documentation. Fix every issue before any patient-facing release. Record results so approvers can review them.
Step 5: Launch in phases and monitor
Start with one clinic or visit type, monitor booking accuracy, errors and patient feedback, then expand gradually. Track key measures such as booking completion, no-show rates and staff workload, so the business value of the integration is clear to leadership.
Common Pitfalls in Cadence Integrations
Even well-planned scheduling projects run into recurring problems. Knowing them in advance lets teams design around them rather than discovering them after patients are affected. Most issues come from mismatches between real clinic workflows and what integrations assume about scheduling rules.
Visit types and decision trees
Patients rarely know which visit type they need. If your app asks them to choose, they will often pick wrong. Build guided questions that map answers to approved visit types, and agree on this logic with clinic leadership before launch.
Race conditions and double booking
Between finding a slot and booking it, another scheduler may take it. Handle booking failures gracefully by offering alternatives immediately. Never assume a displayed slot is still available, and test concurrent booking scenarios during development to avoid frustrated patients and staff.
Keeping cancellations in sync
If patients cancel in your app but Epic is not updated, slots stay blocked and capacity is lost. If staff cancel in Epic but your app is not informed, patients arrive for visits that no longer exist. Use SIU feeds or regular checks to stay synchronized.
Approvals and timelines
Health system approvals, security reviews and Cadence build changes often take longer than development. Start these conversations early, provide clear documentation and assign a single owner to manage them. Our Epic developers help navigate these steps efficiently. Track every request in one shared plan.
How Taction Helps With Cadence Integrations
Taction has delivered 200+ healthcare projects, including Epic integrations, scheduling platforms and patient engagement tools. We help digital health companies and health systems connect to Cadence reliably, choosing the right mix of APIs, feeds and automation for each workflow. Start with a free architecture review.
Free scheduling integration review
Share your scheduling use case, target health systems and current tools. We assess feasibility, recommend APIs, feeds or automation, outline required Cadence build, and flag risks early. You receive a clear plan for approvals and delivery. Book your free review.
API and feed development
We build Epic FHIR scheduling integrations and HL7 SIU interfaces through engines such as Mirth Connect, with monitoring and alerting included. Our HL7 integration services team ensures every scheduling event reaches the systems that depend on it reliably. Failed messages trigger alerts, not silent gaps.
Scheduling platforms and patient apps
We design and build scheduling platforms and patient-facing booking experiences that work with Epic. Explore our appointment scheduling platform solutions to see how we combine usability, scheduling rules and EHR integration into one experience. Each build is tailored to your visit types, clinics and patient groups.
Automation with guardrails
Where APIs fall short, we design limited, well-governed automation and AI agents that respect scheduling rules and audit requirements. We document every bot and agent, monitor performance and plan migration to APIs as health systems expand their scheduling capabilities. Clinic leaders approve every rule first.
Frequently Asked Questions
These are the questions we hear most about Epic Cadence integration from digital health companies and health systems. Because capabilities depend on each organization’s build, a review of your specific situation gives the most reliable answers.
Can third-party apps book appointments in Epic Cadence?
Yes, through Epic’s Appointment find and book operations, but only when the health system has configured which visit types, providers and departments allow outside booking. Many organizations start with read-only access and expand booking capabilities after confirming workflows and approvals.
Which HL7 messages does Cadence send?
Cadence scheduling events are commonly sent as HL7 SIU messages, covering new appointments, reschedules, modifications, cancellations and no-shows. Downstream systems receive these through an interface engine. Exact message types and fields depend on each health system’s interface configuration. Confirm the list with the interface team.
Should we use APIs or HL7 feeds for scheduling?
Use FHIR APIs for real-time, patient-level reads and bookings. Use HL7 SIU feeds when systems must stay synchronized with all scheduling events. Many organizations use both, with feeds for synchronization and APIs for patient-facing booking features. The right mix depends on your workflow and volume.
Is RPA a good option for Cadence automation?
RPA can automate specific tasks inside Hyperspace when APIs or feeds are unavailable, but bots are fragile across Epic upgrades and interface changes. Use RPA as a short-term bridge with strong governance, and move to APIs or feeds when they become available.
How long does a Cadence integration take?
Timelines depend mostly on health system approvals and Cadence build, not development. Read-only integrations can move faster, while booking integrations need more configuration and testing. Starting approval conversations early and keeping scope focused shortens the overall timeline significantly. Phased launches reduce risk further.