Every practice knows the faces. The patient who is in three times this month. The one whose notes run to a volume, whose problems never quite resolve into a single diagnosis, and whose name appears on the rota in a way that everyone notices and nobody quite discusses.
In improvement work across general practices in the Black Country, the cohort analysis produced a figure that is worth sitting with: 3.9 per cent of patients accounted for 26 per cent of appointments.
That number is a description of concentration, not a verdict on the patients inside it. The temptation when a figure like that appears is to treat high attendance as a problem of demand to be managed. In my experience it is more often a signal of unmet need that the appointment is being used to express.
Data finds the cohort. Clinicians explain it.
A list of high utilisers is easy to produce and almost useless on its own. What turned the list into something actionable was clinical interpretation: sitting with practice teams and asking what was actually happening for those patients. Some had several long-term conditions and fragmented care across services. Some had a single problem that had never been resolved because each appointment addressed the presenting symptom rather than the underlying issue. Some were socially isolated, or dealing with housing, income or caring responsibilities that no appointment can fix but which reliably generate appointments.
Interventions that add work will fail
The most common failure in this kind of work is designing a new service that adds a layer without removing anything. A care coordinator who is added on top of an unchanged system produces more contacts, not fewer. The interventions that held up were the ones that replaced something: a named clinician with authority to make decisions, longer appointments instead of several short ones, and a route to community support that did not require a GP appointment to access.
Continuity is an intervention
Continuity of care is often discussed as a preference. In this cohort it functions as treatment. A patient seen by the same clinician across several contacts accumulates a story rather than repeating an opening chapter, and repetition is expensive in every sense.
What I am not claiming
This work did not produce a measured reduction in appointments, and I am not claiming one. The cohort finding is a description of the pilot population, not an outcome of any intervention. Anyone presenting this kind of work should be clear about which of those two things they are holding.
What it means for the future of primary care
Primary care is being reorganised around access, and access is measurable, which is why it dominates. But the concentration of demand suggests that the larger prize is in the small number of patients whose care is currently expensive, fragmented and unsatisfactory for everyone involved. That work is slower, harder to attribute, and does not fit neatly into a target. It is also where the clinical value sits.
One caution. A label like frequent attender can travel from a spreadsheet into a corridor and become a reason to withdraw care rather than a reason to organise it better. The point of identifying the cohort is to give those patients a better route through the system, not to move them out of it.