Every exception is a visit that was worked but can't bill. The difference between agencies that lose a little and agencies that lose a lot is not the exception rate — it's how fast the queue gets cleared, and whether the causes get fixed or just the symptoms.
An EVV exception is a visit record that failed a verification check — a missing clock-out, a location outside the geofence, a caregiver the system can't match. Until it's resolved, the visit can't become a claim. The units sit in limbo: the caregiver has been paid, the payer owes nothing, and a clock is running, because every state sets a timely-filing window after which the visit can never be billed.
That clock is what makes speed the whole game. An exception worked the next morning is a five-minute fix. The same exception ninety days later means reconstructing what happened from memory and documentation — and past the filing window it isn't worth anything at all.
The most common failure mode is structural: exceptions get looked at when billing runs, weekly or biweekly. By then the caregiver doesn't remember the visit, the member can't confirm the time, and the batch is large enough that the biller triages by ease instead of value. A daily pass — even thirty minutes each morning — catches exceptions while the visit is fresh, keeps the queue small enough to finish, and surfaces repeat offenders while there's still a pattern to see.
Not every exception deserves the same attention. A workable order:
— Closest to the filing deadline first. Aged exceptions are the ones about to become permanent losses.
— Highest unit count next. A missed clock-out on a 10-hour shift outranks one on a 30-minute check-in.
— Then by cause, in batches. Twenty exceptions with the same cause — one caregiver, one member's geofence, one service code — clear faster together than interleaved, and batching is what makes the underlying cause visible.
Clearing an exception recovers one visit. Fixing its cause prevents hundreds. If the same caregiver misses clock-ins every week, the fix is a two-minute retraining call or checking whether their app works on their phone — not another manual verification. If one member's address flags every visit, the geofence or the address on file is wrong. If a new hire generates exceptions on every shift, onboarding isn't finishing the EVV setup before the first schedule. A useful discipline: every time an exception is cleared, tag why it happened. A month of tags tells you exactly where the process is leaking.
Every state allows some form of manual correction, and every state constrains it: who may edit, what documentation is required, which reason codes are acceptable, and how much manual verification an agency can show before it draws audit attention. High manual-edit rates are themselves a signal payers watch. The goal is not to get good at editing visits — it's to need fewer edits — but while the causes are being fixed, edits must be done inside the state's rules, with documentation that survives an audit.
Queues that belong to everyone belong to no one. Exceptions touch scheduling, HR, and billing, so they default to being nobody's job. Whatever the team's size, one named person should own the queue's daily clearing and its aging report — even if others do pieces of the work. Ownership is what turns "we should look at that" into a number someone answers for.
Four figures, pulled monthly, are enough to manage this:
— Exception rate: exceptions as a share of visits. Trend matters more than level.
— Time to clear: median days from exception to resolved. Should be days, not weeks.
— Aged out: units that crossed the filing window unresolved. This is the write-off line; the target is zero.
— Repeat causes: share of exceptions from causes seen before. If it isn't falling, visits are being cleared but nothing is being fixed.