A patient access prior authorization handoff is not one workflow but a sequence of distinct API exchanges with different owners, inputs, and timing rules. CMS describes the Prior Authorization API as letting providers determine whether authorization is required, identify documentation requirements, exchange requests, and receive decisions from within an EHR or other health IT system [1]. The handoff begins when a provider's treatment recommendation triggers a coverage requirements check, continues through documentation assembly, and ends with a submitted request that a payer adjudicates. The critical distinction is that a Coverage Requirements Discovery response does not start the payer's decision timeframe clock, while a submitted Prior Authorization Support request does [3]. Understanding where each step begins and ends matters for teams designing patient access workflows, because the same patient encounter can produce multiple API calls with different response obligations.
The Three API Handoffs in One Encounter
CMS documentation separates the prior authorization workflow into three implementation guides that handle different handoff points. The Coverage Requirements Discovery (CRD) IG defines a workflow for payers to provide coverage requirement information through the EHR, informing the provider whether prior authorization is required and what documentation rules apply [1]. The Documentation Templates and Rules (DTR) IG can write rules that support automatic extraction of patient information from the EHR for provider review and confirmation before submission [1]. The Prior Authorization Support (PAS) IG then enables requests and decisions to be transmitted via FHIR API from within the provider's EHR or clinical system [1].
Each handoff carries a different record ownership model. The CRD exchange is a query and response pair: the provider system asks about coverage requirements, and the payer returns rules without opening an adjudication case. The DTR step operates on the provider side, pulling structured data from the EHR into a documentation package that a clinician reviews. The PAS submission is the point where the provider transfers a request record to the payer or an intermediary endpoint, and the payer assumes responsibility for adjudication within the applicable timeframe [3]. Teams that treat these as one continuous transaction risk misreading which system controls the record at each stage.
Where the Decision Clock Starts and Stops
CMS is explicit that a CRD check and response, by itself, does not start the decision timeframe clock for a payer [3]. A provider can receive a CRD response indicating that prior authorization is required and what documentation is needed, yet the payer's 72-hour expedited or seven-calendar-day standard response window has not begun. The clock starts when the payer receives an actual prior authorization request, which in the PAS workflow is the submitted request assembled from the documentation step [3]. This distinction has practical consequences for patient access teams that measure turnaround time from the moment a coverage check is initiated.
The response deadlines apply regardless of channel. CMS states that the 72-hour and seven-calendar-day requirements apply whether the request arrives through the API, portal, fax, phone, or another channel [3]. This means the handoff workflow does not create a separate timing regime for API submissions; it changes the transport and data structure, not the adjudication obligation. The final rule did not require impacted payers to use the Prior Authorization API for real-time decisions, though CMS anticipates that some responses may be made in real time while others continue to require clinical reviewer evaluation [3].
A Suggested Validation Scenario for a Handoff Workflow
A useful hypothetical validation scenario would simulate a specialty referral where the ordering provider and the rendering provider use different EHRs. The ordering provider's system sends a CRD query for an imaging service and receives a response indicating prior authorization is required with specific documentation rules [1]. The system then uses DTR to extract structured patient data into a draft documentation package for clinician confirmation [1]. The clinician confirms the package, and the system submits a PAS request to an intermediary endpoint that converts it to X12 278 [4]. The rendering provider's imaging center system subscribes to PAS notifications for the same request [4]. The test would verify that the CRD response timestamp is not recorded as the start of the payer's decision clock, that the intermediary returns a FHIR response to the submitting EHR, and that both provider systems receive status updates until the request is finalized [3][4]. This scenario is a suggestion for validation design, not a description of any documented test result.
The Intermediary Endpoint and the X12 278 Boundary
The Da Vinci PAS implementation guide describes a handoff architecture where the provider's EHR submits a FHIR-based prior authorization request to an intermediary endpoint, not directly to the payer's X12 transaction system [4]. That endpoint is responsible for ensuring that HIPAA Transaction and Code Set requirements are met, and the payer's response is intended to return to the intermediary endpoint and be available to the provider's EHR using the FHIR standard [4]. The X12 portion of the exchange happens in what the guide calls a black box that the submitting system needs no visibility into [4].
This design creates a record ownership boundary that patient access teams should document. The provider system owns the FHIR request record up to the intermediary endpoint. The intermediary owns the conversion to the HIPAA X12 278 transaction, which CMS identifies as the currently adopted standard for prior authorization transactions [2]. The payer owns the adjudication record once the request is received. The PAS guide states that the goal is to provide real-time prior authorization, where possible, in the provider's clinical workflow [4]. However, the guide does not specify which intermediary implementations achieve real-time responses, and the source does not document independent testing of that goal.
Monitoring Requests Across Provider Systems
A submitted prior authorization request may need monitoring by more than one provider system. The PAS guide notes that the EHR must monitor for changes to the request until results are finalized, and this monitoring might be done by the system that submitted the request, the system of the provider expected to perform the authorized procedure, or both [4]. The guide gives the example of a family physician system and an imaging center system both tracking the same authorization [4]. A PAS Subscription has been created to allow systems to monitor for specific requests, with an inquiry operation available in response to a subscription notification indicating that the prior authorization has changed [4].
This monitoring requirement introduces a handoff that is often overlooked in patient access design: the transition from the ordering provider's workflow to the rendering provider's workflow. The PAS guide describes the prior authorization process from the EHR side as five steps: determine whether authorization is required and what documentation is needed, gather information to support the request, submit the request, and monitor the request for resolution [4]. The fifth step is implied by the monitoring discussion but not enumerated in the same anchor. Teams should treat the monitoring subscription as a separate integration point with its own authentication and notification handling.
What the Patient Access API Adds to the Handoff
The Patient Access API operates on a different timeline and serves a different record consumer than the Prior Authorization API. CMS states that through the already established Patient Access API, impacted payers are required to make information available to patients about prior authorization requests and decisions, excluding those for drugs, by January 1, 2027 [2]. This means the patient-facing view of a prior authorization is a downstream read of the same adjudication record that the provider submitted through the PAS workflow. The patient access handoff is therefore a payer-to-patient data exposure, not a provider-to-payer transaction.
The 2024 CMS final rule builds on the 2020 rule by outlining requirements for additional information that certain payers must provide via the Patient Access API and new requirements for three additional APIs: Provider Access API, Payer-to-Payer API, and Prior Authorization API [2]. The Patient Access API requirement for prior authorization information is dated January 1, 2027, while the Prior Authorization API implementation requirement for certain CMS-regulated health plans also starts January 1, 2027 [2][5]. Patient access teams should not assume that the patient-facing API and the provider-facing API share the same data model or update frequency; the sources do not document that synchronization behavior.
Sources & references
- Payer-To-Payer Application Programming Interface Workflow · www.cms.gov
- Application Programming Interfaces (APIs) and Relevant Standards and Implementation Guides (IGs) | CMS · www.cms.gov
- Improving Prior Authorization Processes | CMS · www.cms.gov
- Use Cases and Overview - Da Vinci Prior Authorization Support (PAS) FHIR IG v2.2.1 · www.hl7.org
- Electronic Prior Authorization | CMS · www.cms.gov
- Improving Prior Authorization Processes | CMS · www.cms.gov
- Application Programming Interfaces (APIs) and Relevant Standards and Implementation Guides (IGs) | CMS · www.cms.gov