The five checks that stop a claim from being denied before you send it

Every practice has a denial queue. Very few have a pre-submission routine, and that is the reason the queue never gets shorter. Once a claim is denied, you are paying twice: once for the original work, and again for the follow-up, the rework, and the appeal that may or may not land inside the payer's timely filing window. Prevention is the only part of this cycle with real leverage.

The five checks below are ordered deliberately. Each one catches a category of failure that the later checks cannot see, and running them out of sequence tends to produce a clean-looking claim that still gets rejected.

1. Eligibility on the date of service

Active coverage is not the same as coverage that was active on the day the patient was seen. Plan terminations, employer changes, and Medicaid redeterminations all create windows where a card in hand means nothing. Verify against the date of service, not the date of the visit's scheduling, and re-verify for recurring or series visits rather than trusting the check you ran at intake.

2. Benefits, separately from eligibility

Eligibility answers whether the patient is covered. Benefits answer whether this service is covered, at what level, and what the patient owes. That second question is where deductible status, visit limits, place-of-service restrictions, and carve-outs live — behavioral health and therapy services are especially prone to limits that eligibility alone will never surface.

3. Authorization, tracked to a number

An authorization is only useful if you can produce it later. That means recording the approval number, the approved date span, the approved units or visit count, and the exact procedure codes it covers. A note that says "auth obtained" is not defensible when the payer asks six weeks after the denial.

• Confirm the authorization covers the codes actually billed, not the codes originally requested.
• Check that the date of service falls inside the approved span, including any rescheduling.
• Track remaining units for series care so the last visits of a course don't fall off the end of the approval.

4. Coding linkage and documentation support

Two failures dominate here. The first is a diagnosis that does not support the procedure — medically unlikely or medically unnecessary edits catch this immediately. The second is a modifier that the documentation does not justify, which passes the clearinghouse cleanly and then fails on audit or post-payment review. Both are caught by reading the note against the code set, which is slower than a scrubber and catches what a scrubber cannot.

5. Demographics and payer routing

The least interesting category and one of the largest. Transposed member IDs, a subscriber name that does not match the plan's record, a secondary payer loaded as primary, or the wrong payer ID for a regional plan will all produce a rejection that never reaches adjudication. These are rejections rather than denials, which matters: they are invisible in denial reporting unless someone is watching the clearinghouse acknowledgment file.

Turning the checks into a routine

1. 1Assign each of the five checks to a named role, not to "billing" in general.
2. 2Put the checks at the point in the workflow where the information is still changeable — eligibility and benefits before the visit, coding review before submission.
3. 3Watch the clearinghouse acknowledgment file daily so front-end rejections don't age silently.
4. 4Pull your last 100 denials, tag each one to whichever of the five checks would have caught it, and work the largest bucket first.

That last step is the one worth doing this week. Denials are not random, and the tagging exercise almost always shows that a small number of causes are producing most of the volume.