US Medical Billing
A/R & Follow-Up

Documenting a Follow-Up Call

Twenty minutes on hold produces one durable asset, and it is not the answer. A verbal answer is not a decision and binds nobody — so what the practice actually acquired is a record of what was said, by whom, and when. A call whose record cannot be used later is a call that, for every practical purpose, did not happen.

Updated 10 min read

On this page

Key takeaways

What a call actually produces

A call is a way to get an answer that no automated channel will give — usually because the question is about something specific enough that no standard transaction carries it. That is genuinely valuable. What it is not is a decision. A representative saying a claim “is being reprocessed” has not reprocessed it, cannot be held to it, and in most cases has no authority to commit the payer to anything.

The corpus states that rule for the appeal case — a phone assurance that an appeal is pending or even approved is not the decision — and it generalizes. A decision that moves money exists when it is on a remittance or in a written notice. Everything before that is information, and information decays unless it is written down.

The three jobs the note will be asked to do

The fields, and why each one is there

Date and time
Not for tidiness. Filing and appeal windows are counted, so a call near a boundary is either evidence or nothing depending on whether the date is recorded. Record the time as well where a payer's own day-boundary matters.
The payer's call reference number
The most valuable field on the list, because it is the only one that gives the practice leverage over the payer's own record. Ask for it before the call ends. If none is offered, ask again — and if there genuinely is none, write down that none was given, which is itself a fact about the payer worth having.
Who you spoke to
A first name and whatever identifier the representative offers. A call attributed to nobody is much harder to reconstruct, and the difference in effort at the time is a few seconds.
What you asked, as they heard it
The claim as identified to them — which claim, which dates, which member. Half of all disagreements about what a payer said are really disagreements about which claim was being discussed.
What they said, in their words
Recorded separately from what you concluded from it. This is the discipline that distinguishes a usable note from a useless one: “rep said the claim is in review, no additional documentation needed” can be re-examined; “claim is fine” cannot.
What was promised, and by when
Any commitment, with the date it should land — and the consequence if it does not. A promise with no date attached is not something anyone can follow up on.
The next action and its date
The field that stops the call from being a way of making the account look worked. A contact that ends without one has advanced the claim's age and nothing else.

The artifact already exists

Is this a call that can fix something, or only ask?

Before dialing, it is worth knowing which kind of call this is, because the answer changes what should be recorded and what should be protected in the meantime.

Medicare draws a distinction that is useful well beyond Medicare. A reopening is, in the words of 42 CFR 405.980(a)(1) (opens in a new tab), “a remedial action taken to change a binding determination or decision that resulted in either an overpayment or underpayment.” It is a route for correcting something — including, expressly, a clerical error — and it is not an appeal. The two run on different tracks with different timeframes.

The asymmetry that decides how to use it

The reopening timeframes are not reproduced here. They are Medicare's, they differ by the basis being relied on — routine, good cause, clerical error, fraud — and a reader applying them to a commercial payer would be applying somebody else's rule. What travels is the shape: some things are corrected and some are appealed, the routes are different, and using the wrong one spends a window on a process that was never going to produce the outcome.

  • A call that can fix. Where the issue is a demonstrable error rather than a disagreement about coverage or judgment. Record what correction was requested, on what basis, and what reference the request was given.
  • A call that can only ask. Where the issue is what the payer decided. The call gathers information; the remedy is elsewhere and is on a clock. Record the information, and make the next action the one that protects the clock.

Verification runs both ways

A conversation about a claim is a conversation about a patient, and practices tend to think about identity verification only for calls with patients or their representatives. The obligation is stated more broadly than that.

Under 45 CFR 164.514(h) (opens in a new tab), prior to a disclosure permitted by the rule, a covered entity must verify the identity of a person requesting protected health information and that person's authority to have access to it, where the identity or authority is not already known — and must obtain any documentation, statements or representations, oral or written, that the rule makes a condition of the disclosure.

Two practical consequences for a follow-up call

Where the record has to live

A note that exists somewhere other than the account is a note that will not be found by the person who needs it. That sounds obvious and is routinely violated, usually by a spreadsheet that one team keeps because the account note field is awkward.

  • On the account, not beside it. The test is whether someone opening the claim cold sees the call without knowing to look elsewhere.
  • Structured enough to be counted. Free text is fine for what was said; the date, the reference and the next action should be fields, or the practice can never answer how much of the book has actually been contacted.
  • Attached to the claim the call was about. A note on the patient rather than the claim is unfindable at the moment it matters, particularly where a patient has several open claims.
  • Durable beyond the person. The whole point is that the next person does not have to call again. A record only its author can interpret has failed at its third job.

And the best call is the one not made

How call records fail

  • “Called payer, pending.” The commonest note in any billing system, and it contains nothing: no reference, no name, no statement, no next action. It records that time was spent.
  • No reference number. Asked for at the end of the call or not at all. Without it the practice has an account of a conversation the payer cannot locate.
  • Conclusion recorded instead of statement. “They said it was fine” cannot be re-examined when it turns out not to have been.
  • A promise with no date. Nothing to follow up on, and no way to tell whether the payer did what it said.
  • The correction route mistaken for the remedy. A request to fix something was made, felt like progress, and the appeal window kept running — and where the request is refused, that refusal may not itself be appealable.

The one-line test

Common questions

Is a call reference number really that important?

It is the field that gives a practice leverage over the payer's own record, which nothing else on the note does. Six months later, a disagreement about what was said is resolved by the payer looking at its side of the conversation, and the reference is how it finds it. Ask for it before the call ends rather than after — and where none is given, record that none was given, because a payer that does not issue references is a fact worth knowing when deciding how much weight its verbal assurances can carry.

Can we rely on what a representative tells us?

For deciding what to do next, often yes. For treating a matter as resolved, no. A verbal answer is not a decision: a decision that moves money exists on a remittance or in a written notice, and a representative generally has no authority to commit the payer. The practical rule is to act on the information and protect the position anyway — keep the follow-up date, keep the window in view, and let the written outcome be what closes the item.

What is the difference between asking for a correction and filing an appeal?

They are different routes with different properties, and Medicare states the distinction clearly. A reopening is a remedial action to change a binding determination that produced an overpayment or underpayment — a correction route, reaching clerical errors expressly. An appeal challenges the determination. The asymmetry that matters operationally is that a decision on whether to reopen is binding and not subject to appeal, so asking costs little and refusal is final. That makes a correction request something to pursue alongside protecting the appeal window, never as a substitute for it. Timeframes for each differ and are not reproduced here, because they are Medicare's and a commercial payer will have its own.

Someone called claiming to be from a payer and asked about a patient's claim. What now?

Treat it as a disclosure question before treating it as a service question. Prior to a disclosure, a covered entity must verify the identity of the person requesting protected health information and their authority to have access, where those are not already known to it. An inbound call is a materially different situation from one the practice placed to a published number, and the safe route is to take the details and call back through a number the practice already holds. It costs one call and removes the entire question.

Authoritative sources

Ready to improve your revenue cycle?

Tell us about your practice and we’ll tell you where we would start.