A patient who cancels creates a record. A patient who does not show creates a record. A patient who simply stops coming creates nothing at all.
There is no event. Nobody called. Nothing was cancelled. The appointment that should have been booked in six weeks was never booked, so there is no row anywhere describing what did not happen.
Which is why attrition is the largest revenue category in most practices and the one nobody has a number for.
What is lapse detection?
Comparing the date a patient was expected to return against whether an appointment exists, and producing a list of those who passed the date with nothing booked. It turns a silent absence into a name somebody can act on.
Why nothing catches it
Every alert in a practice management system is triggered by something occurring. A cancellation fires because somebody cancelled. A denial fires because a remittance arrived. Even a no-show fires, because the appointment existed and its status changed.
A lapse has no trigger. The absence of a future booking is not an event, it is a condition, and conditions are only visible to something that checks for them deliberately.
So the practice finds out months later, when somebody notices volume is soft, or when a provider mentions they have not seen someone in a while. By then the patient has either found care elsewhere or stopped seeking it.
The date already exists
This is what makes the mechanism cheap rather than difficult.
Most treatment involves a cadence. Weekly, fortnightly, monthly, quarterly. That expectation is usually documented in the treatment plan, sometimes in the appointment type, sometimes in a clinical note.
Which means the expected return date is derivable from data the practice already holds. Nothing new has to be collected. The comparison is a date against a booking.
How the check works
Three parts.
An expected return date on every active patient. Derived from the treatment cadence rather than entered by hand. Where a plan says fortnightly, the date is fourteen days after the last completed visit.
A comparison against booked appointments. Does a future appointment exist for this patient. Not did they have one, does one exist now.
A tolerance. A patient a few days past their expected date is normal life. Two weeks past for a weekly cadence is a signal. The window has to differ by cadence, because being a month late matters enormously for weekly therapy and not at all for quarterly review.
What comes out is a list of names, each with a last visit date, an expected return date, and how far past it they are.
Why the list has to be short and current
The same principle that governs every other detection mechanism, and the reason most attempts at this fail.
A list of everybody who has ever lapsed is a database, and nobody works a database. It reads as a project, it needs a block of time nobody has, and it gets deferred.
A list of who crossed the line this week is a task. Ten or fifteen names, workable in an hour, and the same hour next week.
There is also a recovery argument for currency. A patient two weeks past their expected return is contactable and often just busy. A patient six months past has made other arrangements, and the call is a different and much less successful conversation.
Early is not a nice-to-have here. It is most of the value.
What the list tells you beyond the names
Run it for a few months and it stops being a recall exercise and starts being a diagnostic.
Lapses concentrate by provider. Not evenly, and rarely for reasons anyone would guess. Usually it traces to whether follow-ups get booked before the patient leaves rather than to anything clinical, and the per-provider view is what makes it visible.
They concentrate by visit number. The gap after a first appointment behaves differently from the gap after a twelfth. Knowing which one you lose people at tells you where to put the effort.
They concentrate in time. A spike in lapses traces to something datable. A scheduling change, a staffing gap, a provider leaving, a week where nobody was rebooking at checkout.
None of that is visible in a total. All of it is visible in a weekly list with names and dates.
The prevention that follows
Detection finds the ones who lapsed. It does not stop the next ones, and the thing that does is upstream of everything above.
Book the next appointment before the patient leaves the building or ends the call. A patient with a date in hand does not enter the lapse list at all.
Which makes lapse detection the safety net rather than the answer. Necessary, because rebooking will never be perfect, and not a substitute for the rebooking itself.
What this means for you
Pick one cadence, the most common in your practice, and run the check for a single month.
Count how many patients passed their expected return date with nothing booked, and multiply by your average revenue per visit and the visits a typical course of care contains. The figure is usually larger than anybody expects, because nothing has ever presented it as a figure.
Grab 30 minutes with us. Prep nothing. You will see who stopped coming and roughly what it cost.
Questions people ask
What is lapse detection?
Comparing the date a patient was expected to return against whether an appointment exists, and listing those who passed the date with nothing booked. It turns a silent absence into a name somebody can act on.
Why does patient attrition not show up in reports?
Because it produces no event. A cancellation and a no-show both create records. A patient who simply stops coming creates nothing, and reports summarise records. There is nothing to summarise.
Where does the expected return date come from?
The treatment cadence, which is usually already documented in the plan or implied by the appointment type. Weekly means seven days after the last completed visit. Nothing new needs collecting.
How far past the date should trigger a follow-up?
It depends on the cadence. Two weeks past matters enormously for weekly therapy and not at all for quarterly review. Set the tolerance per cadence rather than using one number for everybody.
Does lapse detection replace rebooking at checkout?
No. Booking the next appointment before the patient leaves prevents the lapse entirely. Detection is the safety net for the ones that slip through, and rebooking will never be perfect, so both are needed.