ArticleHome Care Industry

EVV Is Visit Verification. It Is Not Call Documentation. Don't Confuse Them.

EVV verifies six things about a visit: who, what, where, when, and how long. It does not capture why a visit was short, refused, or rescheduled. That explanation is usually a phone call.

SurfacerIQ TeamOctober 12, 20267 min read

EVV Is Visit Verification. It Is Not Call Documentation. Don't Confuse Them.

If your agency bills Medicaid for personal care or home health services, you run an electronic visit verification system. You have probably spent real money on it, trained caregivers on it, and argued with a vendor about GPS drift. What you may also have done — quietly, without deciding to — is let EVV become the place you look when you need to know what happened with a patient.

That is a mistake, and it is worth being precise about why. EVV requirements for home health are narrow by design. The statute says exactly what the system has to capture, and it is a short list.

What EVV Actually Requires

Section 1903(l) of the Social Security Act, added by section 12006 of the 21st Century Cures Act, defines an electronic visit verification system as one under which visits are electronically verified with respect to six things. Per 42 U.S.C. § 1396b(l)(5)(A), those six elements are:

  • The type of service performed
  • The individual receiving the service
  • The date of the service
  • The location of service delivery
  • The individual providing the service
  • The time the service begins and ends

Read that list again and notice what is not on it. Nothing about why the visit was shortened. Nothing about what the family asked for. Nothing about the fall the daughter mentioned on the phone two days earlier. Nothing about the caregiver who called the office at 6:40 a.m. to say she could not get into the house.

EVV answers one question: did a specific person deliver a specific service to a specific patient at a specific place and time? It is an attendance and billing-integrity control. It was built to detect visits that were billed but not delivered. It does that job reasonably well. It was never built to be a clinical or operational record of the interaction, and CMS has not asked it to be.

CMS also did not prescribe how you get there. CMS guidance notes that section 12006(c)(2) provides that section 1903(l) "cannot be construed to require the use of a particular or uniform EVV system." States chose their own models — open vendor, closed vendor, state-mandated platform — which is why agencies operating in multiple states often run more than one EVV workflow and why your EVV data is rarely comparable across state lines.

The Deadlines and the Money

The compliance timeline is already behind us. Per Medicaid.gov, states had to implement EVV for personal care services by January 1, 2020 and for home health care services by January 1, 2023. Non-compliance costs the state, not the agency, through an escalating reduction in federal medical assistance percentage.

The schedule in CMS's FAQ is specific:

ServiceFMAP reduction schedule
Personal care services0.25 points for quarters in 2019 and 2020; 0.5 points in 2021; 0.75 points in 2022; 1 point in 2023 and each year thereafter
Home health care services0.25 points for quarters in 2023 and 2024; 0.5 points in 2025; 0.75 points in 2026; 1 point in 2027 and each year thereafter

States can avoid the reduction by showing both a good faith effort to comply and unavoidable delays. CMS publishes per-state determinations — see the home health compliance status table and the personal care table — and not every jurisdiction is listed as compliant.

Why this matters to you operationally: when a state is under FMAP pressure, it pushes EVV exception rates downward onto providers. That is when agencies start getting claim denials and corrective action plans tied to EVV mismatches. The denial is about a timestamp. The explanation almost never is.

Where the Gap Shows Up

Consider a hypothetical that will feel familiar. A patient's EVV record shows a 42-minute visit on Tuesday against a 60-minute authorization. The state's edit flags it. Your biller pulls the record and sees a short visit with a valid clock-in and clock-out at the right address. There is nothing in EVV to explain it.

What actually happened, in this illustrative case, is that the caregiver arrived, found the patient refusing care, called the office, spoke with a scheduler for four minutes, was told to document and leave, and did. The explanation exists. It exists as a phone call. Nobody wrote it down, because the scheduler was handling six other things and the disposition code she picked was "other."

You now have a billing exception you cannot defend, and a refusal-of-care event your clinical team never heard about. One timestamp problem and one patient-safety problem, both from the same unrecorded conversation. We wrote about the clinical half of this pattern in what a missed visit sounds like before it shows up in your EVV data.

The structural issue is this: EVV captures the visit that happened. The conversations capture the visit that didn't, the visit that changed, and the reason for both. Those are different datasets with different owners, and in most agencies they never meet.

What the CoPs Expect, and What They Don't

Be careful here, because this is where vendors overstate things. Nothing in the EVV statute requires you to record a phone call. Nothing in the Medicare home health Conditions of Participation requires you to record a phone call either. If someone tells you a federal rule mandates call recording for home health, ask them to cite it.

What the CoPs do require is that information move. Under 42 CFR 484.60(c), the HHA "must promptly alert the relevant physician(s) or allowed practitioner(s) to any changes in the patient's condition or needs that suggest that outcomes are not being achieved and/or that the plan of care should be altered." Paragraph (d) requires the agency to assure communication with all practitioners involved, integrate their orders, and involve the patient and caregivers in coordination of care.

A refusal of care reported by phone is a change in the patient's needs. The obligation to alert the physician attaches to the fact, not to the channel it arrived on. Your EVV system will not help you prove you met it. Neither will a disposition code.

Separately, 42 CFR 484.50(e) requires the agency to investigate complaints from patients, representatives, caregivers, and family, and to "document both the existence of the complaint and the resolution of the complaint." Complaints overwhelmingly arrive by phone. The documentation standard is explicit and the medium is not.

This post is general information about regulatory requirements, not legal advice. EVV rules vary meaningfully by state, and you should confirm your own state's model and exception policies with counsel or your state Medicaid agency.

Do This Week

You do not need a project plan. You need to find out whether the two datasets can be joined at all.

  1. Pull last month's EVV exceptions. Take the top 25 by frequency — short visits, missed visits, location mismatches, manual edits.
  2. For each one, try to find the conversation. Was there a call to or from the office within 24 hours on either side? Can you find it in under two minutes? Count how many you can and cannot.
  3. Check the manual-edit reason codes. If more than a third of your edits resolve to a generic reason, your edit taxonomy is not telling you anything. That is a 30-minute fix with your EVV admin.
  4. Look at who gets the call. A refusal-of-care call that lands with a scheduler and stops there is a clinical escalation failure, regardless of what EVV shows. Write down the actual escalation path, then ask two schedulers what they think it is.
  5. Pick one measurable join. Even a shared patient identifier and timestamp between your phone system and your EVV export lets you ask "what was said around this exception" without a manual hunt.

Our home care call compliance checklist covers the record-retain-redact decisions that follow once you start treating calls as documentation rather than overhead. If you are still reviewing a sample of calls rather than all of them, this piece on why 2% sampling is a liability is the companion read.

The Point

EVV is a good control doing a narrow job. Treating it as your record of what happened with a patient is an unforced error, and the gap tends to surface at the worst time — during a claim review, a survey, or a complaint investigation, when the question is not "was the visit delivered" but "what did you know and when."

SurfacerIQ works in the category of conversation intelligence for healthcare contact centers: making the calls around a patient searchable and reviewable so that the operational record is not limited to what fits in a timestamp. That is the capability gap this post describes.

If you are trying to figure out where your EVV exceptions are actually coming from, talk to us.

See SurfacerIQ in action

Calls in. Tickets out. Automatically. See how it works on a real call.