A physician once asked me, holding a report that showed his rooms full and his template booked solid, why anybody thought the practice had an access problem. Down the hall, the person answering the phone was telling a new patient the first opening was six weeks out. Neither of them had the schedule wrong.
That gap is where I stopped trusting most of what a practice management system reports about access. As practice manager of a large medical practice in Cincinnati, I can pull room utilization, wRVU by provider, cycle time, no-show and cancellation rates. All of it earns a place on a dashboard. None of it answers what a patient and a referring office are both asking, which is how soon somebody can be seen.
A full schedule and a long wait are different problems
Utilization is a rear-view measure. It reports how completely last week's slots got used, and a practice can run high utilization with bad access, because a schedule fills from wherever demand is, including from a queue that has been sitting there since spring. Utilization also improves as access degrades. The longer the wait, the more reliably every slot books, right up to the point where the no-shows start.
Third-next-available asks something else. It is the days until the third open appointment of a given type with a given provider, taken today and taken again next week.
The first opening is usually somebody's cancellation
The intuitive version of this is next-available, and it will lie to you. The first hole in a booked schedule is nearly always an artifact: this morning's cancellation, a no-show slot, a procedure that got moved. It reports churn, not capacity. Report next-available in a leadership meeting and the room concludes access is healthy on a day when it is not.
The third opening filters the noise, because three accidental holes clustered in one stretch of calendar stop being accidents. Three is arbitrary the way useful operational rules are arbitrary, far enough out to survive a bad Tuesday and near enough in to move inside a quarter. Take it per provider and per visit type. A practice-level figure is mostly reassurance. Two physicians at four days and one at thirty average to a respectable twelve, and the average is what travels upward while the third physician's new patients go elsewhere.
A backlog is a project. Short capacity is a budget line.
The two get treated as one problem, and the treatments are opposite, which is why the trend matters more than the reading. Long but flat means the practice is carrying a backlog, work that arrived faster than it was served at some earlier point and never got worked off. Backlog does not respond to hiring. It responds to a work-down, extra sessions run until the queue drains and then stopped, and stopping is the part people skip. A line that climbs week over week is demand exceeding capacity, and no template fixes arithmetic. That one belongs in a budget conversation, next to panel size.
Carve-outs feel like control and spend the access you have
Every template collects them. Slots held for post-ops, slots held for a referring group, a block only one nurse can room, an urgent hold that releases at 48 hours. Each rule was written for a reason somebody could defend, and each takes a piece of the day out of circulation. Accumulate enough and the practice sits on open capacity that nobody calling this morning can be offered.
Wave versus modified wave gets argued as though template shape were the lever. It matters at the margins, for cycle time and room turnover, though neither of those adds a slot to the day. The question with weight behind it is how much of the day is reserved for a category instead of for a patient. Take the number before a hold goes in and again a month later. A hold that cannot survive that comparison was protecting a preference.
In this market, access is a referral strategy
An independent group in Cincinnati competes for referrals against employed physicians sitting one click away inside somebody else's EHR, in a market with four large systems and private equity buying up specialty practice. The referring office is not reading your quality data. It is watching how fast its patient gets in and whether a note comes back. Six weeks loses referrals to six days, and loses them quietly, because nobody calls to announce they have stopped sending you patients.
The tri-state adds an edge. Patients cross in from Northern Kentucky and Southeast Indiana, and telehealth is not the free capacity people hope for, since a provider has to be licensed and enrolled where the patient is sitting. Credentialing can constrain access exactly the way staffing does, and only one of those is solved by hiring.
What the number is actually for
Third-next-available does not tell you what to do. It tells you whether the thing you did worked, which is rarer than it should be in operations, where changes get judged by whether the people who made them felt better afterward. Move a carve-out, add a session, reassign a panel, and within a few weeks the line moves or it sits there. I still read utilization and wRVU every month, because a practice has to stay solvent. But those tell me how the schedule is treating us, and no patient has ever called to ask me that.