EVV hard-edits, explained.

If you bill personal care services in a state with mature EVV, you've probably seen this rejection code at least once: 8204 — Visit verification mismatch. Or its cousin: 8211 — EVV missing or incomplete.

These are hard-edits, and they will reject your claim before it even sees an adjudication queue. Same-day, often within minutes of submission. The 837 file goes in, the 277CA comes back, and your billing person spends the rest of the morning figuring out why.

What a hard-edit actually is

A hard-edit is a claims-processing rule that rejects a claim outright — no payment, no further review, no adjudication path — based on a check the system can run mechanically. Soft-edits flag for review. Hard-edits stop the train.

For EVV, the hard-edit is the state Medicaid system asking: does the visit on this claim match a verified visit in our aggregator? If it does, the claim proceeds. If it doesn't, the claim rejects. There is no human in the loop.

The match is usually keyed on five fields: Medicaid ID, service date, procedure code, provider NPI, and a verified visit ID. If any of those don't tie out — even by a single character — the claim rejects.

The five most common hard-edit causes we see

• Visit captured in the EVV vendor, not submitted to the state aggregator. The most common cause. The aide checked in, the EVV vendor logged it, but the daily push to the state aggregator failed silently — usually a stale token or a schema-version mismatch. The visit exists in your vendor's system; it does not exist in the state's.
• Service date mismatch. The aide checked in at 11:54 PM. The claim says service date "tomorrow." Or vice versa. Time-zone bugs in EVV vendors are real and embarrassingly common.
• Procedure code mismatch. The visit was logged as T1019 (PCS, 15 min). The claim was billed as S5125 (attendant care). Some states cross-walk these. Most don't. Hard-edit.
• Patient-aide mismatch. The aide who checked in is not the aide named on the authorization. This often happens when an agency reassigns a patient mid-day and forgets to update the auth.
• Late visit submission. Most aggregators give you a window — usually 7 days — to submit the visit before the claim's hard-edit window closes. Miss the window, hard-edit.

Why "we'll catch it on the back end" is a bad strategy

The temptation is to let the hard-edits happen, work the rejections, and resubmit. This works when you have one or two a week. It does not work when you have forty a day. We've seen agencies bury an FTE in resubmission work — a person whose entire job is reconciling 277CAs against the aggregator.

That's not a billing problem. That's a visit-capture problem masquerading as a billing problem.

What "submission directly to the aggregator" buys you

The cleanest way to avoid hard-edits is to skip the middleman. If your visit verification system pushes directly to the state aggregator the moment the visit is verified — no overnight batch, no daily push, no CSV upload — the visit is in the aggregator before your billing person ever opens a claim file.

That's how we built VisitLock's EVV layer. The token is signed at the moment of verification and submitted in the same transaction. There's no window for it to drop. There's no reconciliation step. The hard-edit cannot fire because the data the hard-edit checks is already there.

The five states where we run direct-to-aggregator (FL, OH, NC, TX, IL) see hard-edit rates from VisitLock customers in the low single digits — almost all of which are authorization-data issues, not visit-capture issues.

What to do this week

If you're seeing more than one hard-edit per ten claims, three things to check today:

• Pull a 30-day report from your EVV vendor. Compare the visit count to your state aggregator's portal. If they don't match, you have a silent submission failure.
• Check your aide-to-patient authorizations. Pull any visit where the billed aide doesn't match the authorized aide.
• Look at your time-zone setup. If your phones are reporting in UTC and your claims are billed in local, you've got a midnight problem.

Hard-edits are loud, but they're not random. They're a signal. The signal is that visit data and claim data don't agree — and that disagreement is exactly the gap fraud lives in.

Kara Mendez is co-founder and CEO of VisitLock. She previously ran operations at a 200-aide PCS network in Florida.