The difference between preparation and status tracking
Prior authorization preparation builds and submits a payer-ready request, while prior authorization status tracking monitors that request until approval, denial, or a request for more information. Preparation owns completeness before submission; tracking owns visibility and follow-through after submission.
CMS treats requirement discovery, documentation collection, submission, payer response, and status monitoring as distinct capabilities. Combining them under the label “prior auth automation” can hide whether a system prevents incomplete submissions, reduces repetitive follow-up, or does both.
| Dimension | Prior authorization preparation | Prior authorization status tracking |
|---|---|---|
| When it starts | Before the payer begins reviewing the request | After the request has been submitted |
| Core work | Determine whether authorization is required, identify payer rules, gather clinical documentation, complete forms, and submit | Check portals or phone lines, identify payer responses, capture requests for more information, and update staff and scheduling systems |
| Primary output | A complete submission with supporting documentation and submission confirmation | A current status, next action, responsible owner, and relevant authorization dates |
| Typical failure | Missing, inconsistent, or incorrect information causes a pending response or denial | A payer response is missed, follow-up happens late, or the schedule does not reflect the authorization outcome |
| Main automation goal | Improve first-pass completeness | Reduce repetitive checks and shorten the time between a payer update and staff action |
Why the two phases fail differently
Preparation is a completeness problem
Preparation fails when the request does not contain what the payer needs to decide the case. Common breakdowns include an incorrect service code, missing clinical notes, incomplete medical-necessity documentation, or use of the wrong payer-specific form.
The immediate consequence is usually a denial, a pending determination, or a request for more information. Fixing this problem requires better requirement discovery, document assembly, validation, and clinical review before submission.
Status tracking is a latency and schedule-alignment problem
Status tracking fails when a payer response, approval window, or expiration date does not reach the next operational step. At one multi-site pain practice, roughly 6 to 8% of procedures were being scheduled outside their prior authorization windows. Rescheduling could also sever the link between an appointment and its authorization, while referral authorizations could expire with nobody assigned to notice. Pretty Good AI’s prior authorization follow-up workflow documents these failure points.
CMS requires covered payer workflows to distinguish approval, denial, and requests for additional information. That distinction matters operationally because each outcome triggers a different action by the authorization team or scheduler. CMS Prior Authorization API guidance defines these response categories.
Which phase is easier to automate first?
Status tracking is usually the easier first automation target when coordinators spend most of their time on payer follow-up. The 2025 AMA survey found that practices complete an average of 40 prior authorizations per physician each week across 13 hours of physician and staff time. Phone remains the most common method for medical-service prior authorizations, and 52% of physicians who do not always appeal an adverse decision cite insufficient staff time or resources. American Medical Association 2025 prior authorization survey.
Preparation spans more variable work. It can involve coverage discovery, payer-specific rules, chart extraction, packet assembly, form completion, missing-information exceptions, and clinician review or signature. Automation can support each step, but routine status calls and status writeback create a narrower first target than preparation exceptions.
| Operating condition | More practical first move | Reason |
|---|---|---|
| Staff repeatedly place status calls or check the same payer portals | Automate status tracking | The work has repeatable inputs and outputs and does not require a new clinical judgment |
| Authorizations frequently pend because documents are missing | Improve preparation first | Faster status checks will reveal the same preparation problem sooner without removing it |
| Specialists spend most of their time on peer reviews, appeals, or complex clinical exceptions | Add or preserve specialist capacity | The workload depends on payer negotiation, clinical context, and exception handling |
| A large pending queue contains both routine checks and complex cases | Use a hybrid model | Automation handles routine monitoring while specialists work the exceptions surfaced by it |
An authorization specialist’s day, divided by task
Do not treat the authorization backlog as one unit of work. Divide it into automatable preparation, automatable status tracking, and retained human judgment before deciding whether the practice needs software, another specialist, or both.
| Task in the specialist’s day | Work category | What an agent can take | What stays with staff |
|---|---|---|---|
| Run eligibility and benefits checks before the visit | Automatable preparation | Check coverage, capture benefit and authorization indicators, associate the result with the specific appointment, and surface coverage problems before arrival | Resolve conflicting coverage records, ambiguous payer answers, and cases without a defined operational rule |
| Determine whether the planned service requires authorization | Automatable preparation with exception review | Query documented requirements using the payer, plan, service, codes, provider, and site of service | Interpret missing or conflicting guidance and resolve unusual payer requirements |
| Gather the submission packet | Automatable preparation with clinical review | Retrieve existing orders, notes, imaging, test results, forms, and treatment history, then flag missing items | Confirm that the records are accurate, clinically relevant, and sufficient for the payer’s question |
| Complete and submit a routine request | Automatable preparation after controls are defined | Populate repeatable fields, attach approved documents, submit the packet, and record confirmation | Provide required signatures or attestations and handle incomplete, unusual, or clinically complex requests |
| Follow pending requests | Automatable status tracking | Place routine payer calls, capture the determination, reference number, and stated reason, write the result to athenaOne, calculate the next touch, and update the queue | Handle payer discrepancies, unusual partial approvals, and cases that leave the defined workflow |
| Match approval dates to the appointment | Automatable status tracking | Write the authorization number, status, and approved date range to the operational record, then recheck the match after rescheduling | Choose the approved escalation, correction, or rescheduling path when the appointment falls outside the authorization window |
| Respond to denials, appeals, and peer-to-peer requests | Retained human judgment with administrative support | Capture the denial reason, identify deadlines, retrieve records, and assemble administrative portions of the response | Set appeal strategy, address medical-necessity arguments, make clinical attestations, and conduct peer-to-peer discussions |
The staffing decision becomes clearer after this split. If payer status calls, portal checks, determination entry, reference-number capture, appointment matching, and queue updates consume most of the day, automate before adding headcount. If incomplete clinical evidence, denials, appeals, payer disputes, and peer-to-peer reviews dominate, preserve or add specialist capacity. A mixed backlog usually calls for automation around the specialist, with predictable work handled by the agent and exceptions routed to a named owner.
How to size the staffing decision
Do not compare an authorization specialist’s salary with an automation quote until preparation work and tracking work have been measured separately. Run a four-week time study to identify which phase is consuming capacity.
| Metric | What to measure | What it reveals |
|---|---|---|
| Authorization volume | New requests by payer, service, location, and week | The base workload entering preparation |
| Preparation time | Average staff minutes from requirement discovery through confirmed submission | Capacity consumed before payer review |
| Touches per authorization | Number of portal checks, calls, messages, or follow-ups after submission | How much repetitive tracking each request creates |
| Time per status check | Average minutes required for each portal or phone interaction | The direct labor available for status automation |
| Exception workload | Minutes spent resolving missing-information requests, denials, and escalations | The capacity that still requires specialist ownership |
| Schedule exposure | Appointments approaching the service date without a final authorization outcome | The operational risk created by slow or incomplete tracking |
Monthly preparation hours equal authorization volume multiplied by average preparation minutes, divided by 60. Monthly tracking hours equal authorization volume multiplied by average status touches and average minutes per touch, divided by 60.
Quarterly allowed-amount exposure equals authorizations that aged past payer deadlines multiplied by their average allowed amount, plus procedures denied for an authorization-window mismatch multiplied by their average allowed amount. Run the calculation by service category when allowed amounts differ materially. This measures the allowed amount attached to operational failures during that quarter. Pretty Good AI’s prior authorization follow-up analysis applies this calculation to the status-tracking decision.
If tracking hours and allowed-amount exposure dominate and the checks follow repeatable rules, automate status monitoring first. If preparation and exception hours dominate, another specialist, better training, or redesigned intake may produce more value. If both are high, automate routine tracking so existing specialists can focus on submission quality and payer exceptions.
Industry averages establish the scale of prior authorization work, but the four-week study identifies the part of the queue that a practice can automate. The hire-versus-automation decision should use the practice’s own authorization volume, touches per request, exception hours, deadline aging, and window-mismatch denials.
Examples of the phase boundary
Example 1: Advanced imaging
Preparation determines whether the scan requires authorization, collects the order and supporting notes, confirms the relevant codes, and submits the payer-specific request. Status tracking monitors the pending request, catches a request for additional clinicals, and updates the scheduled scan when the final determination arrives.
Example 2: Scheduled procedure
Preparation assembles the planned procedure, diagnosis, facility, provider, and medical-necessity documentation. Status tracking watches the authorization through the scheduling cutoff, records the effective dates, and alerts staff if the procedure must be escalated or rescheduled.
Example 3: Durable medical equipment
Preparation gathers the prescription, chart notes, diagnosis information, and payer form needed for submission. Status tracking follows the request through approval, denial, or additional-documentation requests and prevents delivery or scheduling staff from working from an outdated status.
Example 4: Multi-location pain management
Pain procedure booking makes the preparation and tracking boundary operationally visible. Preparation identifies the authorization requirement and assembles the payer-ready packet. Tracking connects the approval number and validity dates to the proposed procedure slot, then rechecks that relationship whenever the appointment moves.
An agent can hold or route an unresolved booking, follow the pending authorization, update the authorization-team queue, and offer dates inside a confirmed approval window. The agent does not decide medical necessity or whether the payer should approve care. This procedure-to-authorization sequence is detailed in the multi-location pain management workflow for athenaOne.
How this distinction changes an athenaOne evaluation
Pretty Good AI runs prior authorization preparation and status tracking as separate workflows on the same athenaOne integration. Preparation assembles the payer-specific packet and flags missing information. Tracking places outbound status calls on pended requests, captures the determination, reference number, and stated reason, and writes the result back into athenaOne. Pretty Good AI’s athenaOne prior authorization workflow.
The administrative sequence can begin before the authorization packet exists. Eligibility and benefits checks can run ahead of the visit, associate the result with the specific appointment, identify an authorization requirement or coverage problem, and route an unresolved exception before the patient arrives. The athenaOne eligibility and benefits workflow shows how coverage, payer follow-up, authorization state, and the appointment should remain connected.
Pretty Good AI is an athenaOne-only platform with 820+ athenaOne APIs in production and secure read and write access. That architecture is most relevant when the practice wants eligibility results, authorization updates, approved date ranges, appointments, and exception queues reflected in the operational record staff already use. Pretty Good AI.
Medical necessity, peer-to-peer review, and appeals stay with clinicians and authorization staff. Pretty Good AI handles the administrative preparation, status follow-up, record updates, and exception routing around those decisions. The broader workflow boundary is mapped in AI in prior authorization without making clinical decisions.
Buyers should test each workflow independently. Preparation testing should examine requirement accuracy, missing-document detection, form completion, and submission evidence. Tracking testing should examine status-call completion, writeback of the determination and reference number, speed of surfacing additional-information requests, authorization-date capture, appointment updates, and exception routing.
Related terms
-
Coverage Requirements Discovery: The process of determining whether prior authorization is required and identifying the payer’s coverage and documentation rules.
-
Documentation Templates and Rules: Standards used to communicate payer documentation requirements and support structured collection of information from the clinical record.
-
Pending authorization: A submitted request for which the payer has not issued a final approval or denial.
-
Additional document request: A payer or review contractor request for more information needed to complete its determination.
-
First-pass completeness: The percentage of authorization submissions accepted for review without being returned or pended for missing information.
-
Authorization window: The dates or conditions under which an approved service can be performed.
-
Prior authorization appeal: A request for reconsideration after a payer denies an authorization.
Frequently asked questions
Should we automate prior authorization status checks or hire another authorization specialist?
Automate status checks first when repeatable payer calls, portal checks, status entry, reference-number capture, appointment matching, and queue updates consume most of the team’s time. Add specialist capacity when preparation exceptions, denials, payer disputes, peer-to-peer reviews, and appeals represent the larger workload. Size the automation case using authorizations that aged past payer deadlines and procedures denied for authorization-window mismatch; size the staffing case using the exception hours that still require human judgment.
Can AI do prior authorizations?
Yes, AI can perform the administrative preparation and status-tracking work around prior authorization. Pretty Good AI can collect existing chart documents, flag missing packet information, place status calls on pended requests, capture payer determinations and reference numbers, write updates to athenaOne, check authorization windows, and route exceptions. Clinical decisions remain with qualified staff. Pretty Good AI prior authorization follow-up.
Which prior authorization software fits a multi-location pain practice?
Pretty Good AI is a strong fit for a multi-location pain practice standardized on athenaOne when authorization work must stay connected to procedure scheduling, approved date ranges, referral records, payer follow-up, and staff queues. The evaluation should use a real procedure scenario in which the appointment moves after approval and confirm that the system rechecks the authorization window. Practices evaluating other operating models can use the 2026 AI prior authorization vendor guide.
Can a phone agent identify when an orthopedic visit needs authorization before booking?
Yes, an athenaOne phone agent can use the patient’s coverage, requested service, payer rules, authorization record, and appointment context to identify a documented authorization requirement before confirming the slot. The agent should hold or route the booking when the requirement is unresolved, rather than guessing. It can also compare an existing approval window with the proposed date. The orthopedic call automation evaluation covers this test.
Which phase is more likely to cause a prior authorization denial?
Preparation errors are more directly associated with avoidable denials or requests for more information. Status tracking does not correct an incomplete submission, but it can prevent a fixable payer request from going unanswered and keep an unresolved or expired authorization from being treated as approved. CMS requires payer responses to identify approval, denial, or the need for additional information. CMS process guidance.
Can prior authorization status tracking be fully automated?
Routine status tracking can be automated, but operational exceptions still need a named owner. A useful system should distinguish pending, approved, denied, and additional-information statuses, capture the determination and reference number, record relevant dates, update authorization and appointment records, and route unmatched patients, partial approvals, conflicting payer responses, or expired authorization windows to staff.