Annual Wellness Visit Outreach at Panel Scale

Annual wellness visit outreach is the rare project where the benefit is covered, the patient wants it, the schedule has room, and it still does not happen. The obstacle is not demand or capacity. It is that nobody has the list, and building the list by hand across a geriatric panel takes longer than anyone has.

So the visit happens for the patients who happen to come in for something else and get asked, which selects for the people already engaged with the practice. The patients who would benefit most from a scheduled touchpoint are the ones least likely to generate the sick visit where somebody remembers to offer it.

A covered benefit with a large gap between eligible and used

The annual wellness visit has been a covered Medicare benefit for well over a decade, and uptake has never matched eligibility.

In the Medicare Current Beneficiary Survey, 60% of Medicare beneficiaries living in the community in 2022 reported having an annual wellness visit. Claims-based studies of fee-for-service beneficiaries have generally found lower figures than surveys do, and either way the shortfall is measured in tens of percent of a panel that is entitled to the visit and not receiving it.

For a practice, the interesting part is the shape of the gap rather than its size. This is not a population refusing care. It is a population that was never asked on a day when booking was possible.

That makes it an outreach problem with a known solution, and it is the kind of outreach that pays for itself, because the visit is billable, it surfaces work that leads to other appropriate visits, and it puts a date on the calendar for a patient who otherwise appears only in a crisis.

The list is a date calculation, not a judgment call

The reason this is automatable is that eligibility for the visit is arithmetic.

A patient is due based on time since their last one. That fact lives in the appointment and claims history, and it can be computed for every patient on the panel every night without anyone forming an opinion about it. Reading the patient list against appointment history produces a due-and-not-booked list, ranked by how far past due each person is.

That ranking matters more than it sounds. Working alphabetically means the practice contacts a patient who became eligible last week before one who has not had the visit in three years. Working by overdue days puts the largest gaps first, which is both the better clinical outcome and the better use of the calls.

The automation’s role stops at the boundary of the calendar. It identifies who is due, calls, offers real times, books against the correct appointment type, and confirms. What the visit contains and what comes out of it are the clinician’s, entirely.

The complication: booking it wrong costs more than not booking it

This is where panel-scale outreach goes wrong in practice, and it is a billing failure disguised as a scheduling one.

The wellness visit and the annual physical are different things with different coverage, and patients use the words interchangeably. A patient asked whether they want their yearly check-up will say yes to either. If that lands on the wrong appointment type, the practice either bills something the patient did not expect to pay for or performs a visit that will not be covered as booked. Both outcomes are worse than the empty slot, because both end with an unhappy patient and a phone call.

The timing rule compounds it. Because eligibility depends on how long it has been, a visit booked slightly too early fails on a date rather than on anything anyone did in the room. The patient was seen, the work was done, and the claim does not survive.

The automation handles both mechanically and neither by judgment. It computes the earliest date the visit can be booked and refuses to offer a time before it, and it selects the appointment type from the eligibility math rather than from what the patient called it. Where the record is ambiguous, and on a long-tenured geriatric panel it often is, the patient goes to a staff work queue rather than onto the schedule.

The second complication is who answers the phone. On a geriatric panel a meaningful share of calls are answered by an adult child or a caregiver, and the practice’s rules about who may schedule on a patient’s behalf are usually informal. Writing those rules down before running outreach at volume is the difference between a helpful program and an awkward one. The automation follows the rule; it does not invent one on the call.

Reaching a population that does not answer unknown numbers

Contact rates on this panel are lower than on any other, and a program designed around one attempt will report that the outreach did not work.

Older patients screen unknown numbers, use landlines that nobody carries, and often prefer a call at a time of day that no scheduling department chooses by default. None of that is a reason to skip the outreach. It is a reason to design for it.

What works is a bounded sequence across different days and different times of day, mail or portal contact for the patients whose phone attempts fail, and a real path to a person for anyone who calls the number back. A returned call that lands in a phone tree undoes the whole effort.

It also helps to make the offer concrete. Asking a patient to call and schedule converts poorly across every population and worst across this one. Offering two specific times converts, because the decision is small and immediate.

How to know it worked

Three numbers describe this program and the first one is the one practices skip.

Start with the size of the due-and-not-booked list before any outreach runs. Without that baseline the program can only report activity, and activity is not the result. The list itself is usually the most persuasive artifact the project produces.

Then track the share of that list booked, and separately the share that completed the visit. The gap between those two is where reminders and rescheduling need attention, and it behaves differently on this panel than on a younger one.

The last number is the count sent to staff for ambiguity rather than booked automatically. If it is very small, the rules are probably too permissive and some of those bookings will fail at billing. If it is very large, the historical record needs cleanup before scale is realistic. Watching that number is how the practice tunes the program rather than guessing at it.

Key Takeaways

Produce the due-and-not-booked list before starting anything. Most practices have never seen it, and it is both the baseline and the business case.
Rank outreach by overdue days rather than alphabetically, so the largest gaps in the panel get contacted first.
Select the appointment type from eligibility math, never from the words the patient used. Wellness visit and annual physical are not interchangeable at billing.
Compute the earliest eligible date and refuse to offer a time before it. A visit booked days early fails on the date, after the work is already done.
Write down who may schedule on a patient’s behalf before running outreach at volume. On a geriatric panel, caregivers answer a lot of these calls.
Design for multiple contact attempts across different days and times, and give anyone who calls back a path to a person rather than a phone tree.
Watch the count routed to staff for ambiguity. Too few means the rules will produce billing failures; too many means the historical record needs cleanup first.

The annual wellness visit is one of the few pieces of work where the practice, the patient, and the payer all want the same outcome, and it goes undone because the list nobody has time to build is the only thing standing in the way. An AI team reading the patient list and appointment history inside athenaOne can build that list nightly, call the patients who are furthest past due, book against the appointment type the eligibility math supports, and hand the ambiguous ones to staff before they become denied claims.

Sources