How prior authorization works, start to finish

Prior authorization does not usually fail at submission. It fails in the space afterwards — the days between sending a request and receiving a decision, when the file belongs to nobody in particular and the date of service keeps getting closer.

That gap is where the expensive outcomes live. A service delivered without a decision. An approval that arrived but was never recorded, so the claim went out without the number. An approval that was recorded and then expired, because the date range ended two weeks before the procedure was finally scheduled. A peer-to-peer review requested on a Tuesday and noticed on the following Monday, after the window closed.

None of these are difficult problems. They are attention problems, and attention is what a shared front-desk role does not have.

What the work involves, in order

1. Determine whether authorization is required for this payer, this plan, this code and this place of service. The requirement is plan-specific and changes; it is not a property of the procedure.
2. Identify the correct route. Some benefits are carved out to a separate manager, and a request sent to the health plan for a carved-out benefit is not slow, it is in the wrong place entirely.
3. Gather the documentation the payer asks for before submitting. A request submitted incomplete usually returns as a request for information, which costs the whole decision window again.
4. Submit through the payer’s preferred channel, and record the confirmation.
5. Chase at a defined interval until a decision exists. Not once. At an interval.
6. Record the decision in full: authorization number, approved units or visits, the date range, the approved codes, the place of service and the rendering provider.
7. Track the expiry and flag it before it lapses.
8. Escalate a denial or a peer-to-peer request the same day it appears.

Steps five and seven are the two that get dropped when the work sits with somebody who also answers the phone, and they are the two that cost the most.

An authorization is a record, not a status

A note that says “auth approved” is nearly useless at claim time. The biller needs the number, and frequently also needs to know which codes were approved, how many units or visits, between which dates, at which place of service and under which rendering provider — because a claim that varies from the approval on any of those can deny even though an approval exists.

So the standard should be that an authorization is entered as a complete record in the practice system, in agreed fields, with the payer reference and the date of the decision. That record is what protects the claim, and it is also what makes the appeal possible if the payer later disputes it.

The chase is the job

Payers publish decision windows. Requests do not always move within them, and a request that has stalled looks identical to one that is progressing normally until somebody asks. A chase schedule — check at day three, again at day seven, then daily as the date of service approaches — turns an invisible stall into a visible one while there is still room to act.

Every chase should leave a trace: who was spoken to or which portal was checked, on what date, what they said, and what the next review date is. That trace is the difference between a queue somebody else can pick up and a queue only one person understands.

Expiry, the quiet failure

An approved authorization with a date range is a permission with a deadline. Procedures get rescheduled; patients cancel; a surgery booked for March happens in June. If nothing watches the range, the claim goes out against an authorization that had lapsed, and the denial arrives after the service has been delivered — the most expensive point at which to discover a problem.

The fix is mechanical: a review point inside the range, a flag before the end of it, and a rule about who decides whether to extend, re-request or reschedule. It costs a few minutes per authorization and prevents a category of write-off entirely.

Peer-to-peer reviews

When a payer requests a peer-to-peer, the window is usually short and the clinician’s calendar is usually full. The authorization seat cannot conduct the review, but it can do everything around it: notice the request the day it arrives, get it to the right clinician immediately, offer the payer’s available slots, book it, assemble the documentation, and record the outcome.

This is the clearest example of why the escalation path has to be written down before the queue starts. A peer-to-peer request that sits in a general inbox for four days has already failed.

What to define before handing over the queue

The requirement list

Which service types, for which payers, generally require authorization in your mix. It will not be complete and it does not need to be — it needs to be written, so the seat knows what to check rather than guessing.

What may be submitted without clinician review

Some practices are comfortable with a seat assembling and submitting from the chart; others want every request reviewed first. Both work. An unstated rule means either delay or overreach, and both are worse than either choice.

The chase interval

State it. Day three, day seven, then daily inside the last week. An interval that exists on paper is followed; an instruction to “follow up regularly” is not.

The escalation routes

Denial, peer-to-peer request, request for additional information, and approaching date of service with no decision. Four routes, four named roles, four response expectations.

The record format

Which fields hold the number, the range, the units and the approved codes. Every biller downstream depends on knowing where to look.

How to assess the queue

• What proportion of scheduled services requiring authorization had a decision before the date of service?
• How many services were delivered without a decision, and why?
• How many authorizations expired before the service was rendered?
• How quickly were peer-to-peer requests routed to a clinician?
• Which payers or benefit managers are consistently slow, and has the chase interval been adjusted for them?
• Are authorization-related denials falling for the services that went through the queue?

Submissions sent is not on that list on purpose. It is the number that always looks healthy.

When to consider outsourcing it

A dedicated seat is worth considering when authorization work is currently distributed across people who have other jobs, when services are being delivered without decisions, or when nobody can say with confidence how many authorizations are outstanding right now. It is also worth considering when the practice is growing: authorization volume grows with scheduling volume, and it is the first queue to break.

Before choosing a provider, ask:

• Will the seat work US hours, so a payer can be called and a clinician reached the same day?
• Will they work in your system and your payer portals, under your access controls?
• Is it a full-time dedicated seat or shared capacity?
• How will peer-to-peer requests reach your clinicians, and how fast?
• Who tracks expiry, and where is the date range recorded?
• What does the daily handover look like — what will you actually receive each morning?

How Soft Home Global staffs this

Soft Home Global provides trained full-time back-office seats from Rawalpindi, Pakistan, for US medical billing and revenue cycle work. Prior authorization is published at $1,900 per seat per month, billed per seat with a one month minimum.

The seat works US business hours inside your practice management system and your payer portals: determining the requirement, assembling documentation, submitting, chasing on the interval you set, recording the decision in full, and flagging expiry before it lapses. Peer-to-peer requests and denials are escalated the day they appear, on the route you define.

Where an authorization depends on eligibility — a plan that turns out to have terminated, or a benefit carved out to another manager — the account moves to the eligibility queue rather than being called on repeatedly. Soft Home Global staffs that seat as well, so the handoff is internal rather than an email into a general inbox.

The first thing to measure

Pull every service delivered last month that required authorization, and check how many had a complete authorization record on file before the date of service — number, range, units, codes. The gap between that count and the total is the size of the problem.

That number is also the honest baseline to judge a seat against three months later.