“We verify every patient’s insurance before they come in.” Every front desk says it, and it’s true.
The check runs, the day before every visit, automatically, and it comes back with a flag that says approved. Then medical claim denials arrive for coverage anyway, and the office can’t see how both things can be true.
They’re both true. An eligibility check, meaning a question sent to the insurer asking whether this patient has coverage, comes back with a flag. The flag answers that question and no other: a policy exists. Whether the policy pays for this visit, at this practice, on this date, for this service, is four more questions, and the flag was never asked any of them.
Here’s the map: a patient is booked, and the night before, the system asks the insurer about the patient’s coverage and stores the answer. The desk sees the flag, the visit happens, and the claim goes out and is paid or refused. The flag sits at the second step, and the whole front door of the money path depends on someone reading past it.
What the flag actually says
Approved means the insurer found a policy for this person that was active on the day the check ran, and the message ends there.
It’s useful, because a patient with no policy at all gets caught, and that’s a visit that would have produced a claim to nobody. But a patient with a policy is most patients, so the flag says approved on most visits, including the ones the insurer is about to refuse.
The answer the practice needed is in the response behind the flag. Every eligibility response carries the plan name, the coverage dates, the services covered, and what the patient owes. The flag is a summary of one line of it, and the other lines are where the denials, the insurer’s refusals to pay, come from.
The four things approved does not tell you
The plan: one insurer sells dozens of plans, and a practice is in network with some of them, meaning under contract at an agreed rate. The flag reads approved for a policy under a plan the practice has never billed. The claim goes out and comes back out of network. Nothing on the flag was wrong.
The date: the check ran the day before the visit, or a week before, or when the card was first scanned. Coverage that ends on the last day of the month is active on the twenty-eighth and gone on the first. Nothing happens the day a card expires, and a flag from last week reads approved on a visit the insurer won’t pay.
The service: a policy can be active and still exclude the visit being scheduled. A plan that doesn’t cover a particular assessment, a limit on the number of visits, a service that needs authorization, the insurer’s advance permission, before it will pay. The flag says the policy is active, but the refusal says the service wasn’t covered under it.
The switch: a patient changed jobs, changed plans, or moved to a spouse’s coverage in January and never mentioned it. The card on file is the old one, so the check ran against the old policy, found it active for a few more weeks, and said approved. The claim went to an insurer that no longer covered the patient.
None of these is a billing error: each is a question the practice could have answered before the visit, from the response it already had.
Why the response never gets read
Because reading it is a person’s job and the flag is a machine’s.
The check runs and the flag appears automatically, but the response behind it sits in the system, one screen deeper, in a format built for a biller. The desk has forty patients tomorrow and a flag that says approved next to each one. Opening forty responses and reading four lines in each is an hour nobody has, so the flag becomes the answer.
At practices we have worked with, the eligibility check reads approved on most of the visits the insurers later refuse for coverage. The check was running on every patient, but the response was never opened. The plan on it was, on visit after visit, not one the practice billed.
Denials get created upstream, meaning early in the chain, and this is where those refusals get created. The billing team inherits a refusal it couldn’t have prevented and a denial rate it gets blamed for.
The check that works
The check that works reads the response instead of the flag, and it does it without a person.
The plan on the response is matched against the list of plans the practice bills, and a plan not on the list is flagged, in words, the afternoon before the visit.
The coverage dates on the response are compared to the visit date, not the check date, and a visit outside the dates is flagged. The service being scheduled is compared to what the response says is covered and what needs authorization, and a gap is flagged.
And the patient’s share on the response, the copay and deductible and coinsurance, is put on the check-in screen so the desk can ask for it.
That’s four comparisons per patient. A machine does them in a second for the whole schedule, while a person does them for the first six patients and stops.
The result is a short list every afternoon, in words the desk can act on. Not approved: this plan isn’t one we bill, or coverage ends before the visit. This service needs authorization and none is on file, or the patient owes this much at check-in.
What to do with the list
Call the patient before the visit: the whole action is that call, and it’s only possible because the list exists a day early.
A plan the practice isn’t in becomes a conversation about cost, or a referral, or a rescheduled visit under the right coverage, before the chair is spent on it.
A coverage gap becomes a question to the patient about a new card, and a service without authorization becomes a request to the insurer with time to get an answer. An insurer that doesn’t pay the claim is a problem for the billing team. An insurer that was never going to pay is a problem for the desk, the day before, when it can still be fixed.
Real situations, and what the flag hid
At one practice we worked with, the denial review started with the billing team and ended at the front desk. A large share of the refusals were for coverage, and every one had a flag that said approved behind it. The plan on the response was one the practice didn’t bill, and the desk had never been shown the plan, only the flag.
At another, coverage denials clustered at the start of each month, because the check ran days before the visits it covered. The check ran on the twenty-fifth for the whole following week, so patients whose coverage ended on the thirty-first read approved and were refused four days later.
At a third, one type of service kept coming back refused, and the eligibility response for those patients had said, on a line nobody opened, that the service needed authorization. None had been requested, and the billing team had been appealing a refusal the desk could have prevented with one call.
What this means for you
If your denials include coverage refusals and your desk checks eligibility on every patient, the flag is being read and the response isn’t. Pull tomorrow’s schedule, open the eligibility response for each patient, and compare the plan on it to the plans you bill.
The number of mismatches is a denial count for a week that hasn’t happened yet, and every one of them is a phone call today.
Grab 30 minutes with us. Prep nothing. You will see how many of next week’s approved patients are booked under a plan you don’t bill.
Questions people ask
What does an approved eligibility check actually mean?
It means the insurer found an active policy for the patient on the day the check ran, and nothing more. An eligibility check is the question your system sends the insurer about a patient’s coverage. Approved doesn’t say you’re in network with that plan, that coverage covers the visit date, that the service is covered, or that the patient hasn’t switched plans.
Why do claims get denied for coverage when eligibility said approved?
The approved result got read, and the full response behind it didn’t. A coverage denial is the insurer refusing to pay because the patient’s coverage didn’t apply to that visit. The plan, coverage dates, covered services, and any authorization requirement (the insurer’s advance permission) are all in that response, and those lines are where these denials come from.
How should a medical practice verify insurance eligibility?
Have the system read the full eligibility response automatically, the afternoon before each visit. Match the plan against the plans you bill and the coverage dates against the visit date. Check the service against what’s covered and what needs authorization (the insurer’s advance permission). Then put the patient’s share on the check-in screen so the desk can ask for it.
When should a medical practice check insurance eligibility?
Check the afternoon before the visit, and compare the coverage dates to the visit date, not the check date. A check run a week early can read approved on coverage that ends before the patient arrives. That’s why coverage denials (refusals because the coverage didn’t apply) cluster at the start of a month when checks run days ahead.
What should the front desk do with an eligibility problem?
Call the patient before the visit, which works because the list exists a day early. If the practice isn’t in the patient’s plan, talk about cost, a referral, or rescheduling under the right coverage. A coverage gap means asking for a new card, and a missing authorization (the insurer’s advance permission) means a request to the insurer while there’s still time.