Understanding the population using their services gives NHS providers the insights needed to reduce non-attendance and narrow inequalities in access.
Every NHS organisation is under pressure to do more with less. Demand keeps rising, workforce constraints haven’t gone away, and health inequalities are widening. These factors are showing up directly in day-to-day operations. One challenge continues to eat into clinical capacity that providers can’t spare: Did Not Attend (DNA) appointments.
DNAs have traditionally been managed as an operational headache. An empty clinic slot is a wasted slot, lengthening waiting lists and disrupting the flow for the rest of the day. That framing isn’t wrong, but it misses the bigger story. Missed appointments are rarely random events. DNAs tend to reflect the same structural and behavioural patterns that shape whether a patient can access care at all. Understanding those patterns is where real reductions start.
From DNA prediction to population intelligence
Predicting the likelihood of non-attendance is nothing new. Many trusts already use models to identify appointments that may be at risk of becoming DNAs, often incorporating some demographic context such as a deprivation measure linked to a patient’s postcode. The challenge is not the modelling itself. It’s the information those models are built on.
Most conventional DNA models work from the inside out. They rely on attendance histories held within provider systems and describe patients using a relatively limited set of demographic indicators. That can work well when looking at people who have engaged with services repeatedly. It offers far less insight into first-time referrals, people who interact with services infrequently, or the practical realities that make a particular appointment difficult to attend.
We call the alternative “population intelligence”. It’s not a replacement for population health management (PHM) as the NHS already understands it. Rather, it’s PHM viewed through a broader lens, and it differs from how PHM is often applied in two important respects.
The first is scope. Population intelligence begins with an understanding of the entire population a provider serves, not simply the people already receiving care. Every household within a catchment can be characterised, regardless of whether its residents have ever attended an appointment.
The second is depth. Rather than stopping at measures such as deprivation deciles, it incorporates a much wider range of lifestyle and contextual factors, including household composition, financial resilience, digital exclusion, car ownership and transport access, caring responsibilities and communication preferences. These are often the factors that determine whether an appointment is realistically attendable. They are also the factors that tend to be absent from internal attendance data.
When delivered through an AI interface such as Ask Aida, this intelligence becomes something operational teams can explore and question directly, rather than a specialist analytical exercise that produces a report every few months.
Looking beyond the DNA numbers
Viewing attendance through a population intelligence lens can reveal patterns that are difficult to see through operational data alone. By combining appointment information with geodemographic and behavioural insight, providers can identify which communities experience the highest levels of non-attendance, understand what is driving those patterns, and see which services and pathways are most affected.
Analysis often shows that a relatively small number of neighbourhoods account for a disproportionate share of missed appointments. Just as importantly, the reasons can vary significantly between communities. Higher DNA rates are frequently concentrated among populations experiencing financial hardship, insecure employment, social housing, limited digital access, transport barriers and, in some cases, caring responsibilities that make fixed appointment times difficult to manage.
More affluent populations miss appointments as well, but the underlying causes are often different. Work commitments, travel and straightforward diary clashes tend to feature more prominently than structural barriers to access. That distinction matters because the response needs to match the problem.
Where non-attendance stems from competing priorities or simple forgetfulness, reminder messages and straightforward rebooking options can have a significant impact. Where patients are dealing with unreliable transport, insecure work patterns or digital exclusion, reminders alone are unlikely to address the root cause. NHS England’s guidance on reducing DNAs emphasises that understanding the reasons behind non-attendance is essential, because different groups face different barriers.
Patient-level costing adds a different layer of understanding. It does not explain why appointments are missed, but it does show the financial and capacity impact of those missed appointments. When costing data is combined with population insight, providers can see how much resource is tied up in different patterns of non-attendance. That is often the step that turns analysis into action. Population intelligence helps explain the problem; patient-level costing shows the potential value of solving it.
Connecting DNA reduction to neighbourhood health
This is where work on DNAs aligns with a much broader direction of travel across the NHS. The Neighbourhood Health Framework, published by DHSC and NHS England in March 2026, calls on systems to organise more planning and delivery around local geographic communities, with a stronger focus on tackling inequalities at neighbourhood level. Understanding DNAs at neighbourhood level is not separate from that agenda. It is one way of putting it into practice.
A trust that can identify the communities most affected by non-attendance, understand why those patterns exist and demonstrate the impact on access is building exactly the kind of population-level evidence base the framework expects systems to develop over the coming three years.
Turning data into actionable insight
Most healthcare organisations already have access to much of the information needed to better understand attendance behaviour. More often than not, the challenge lies in bringing together appointment data, patient-level costing and population segmentation, rather than gathering new datasets.
This is where tools like CACI’s Synergy and Acorn come in. Synergy, a patient-level costing platform, attaches cost to activity. Acorn provides postcode-level geodemographic and lifestyle segmentation for the whole population, including both patients and non-patients. Used together, they help providers connect the cost of non-attendance with a clearer understanding of who is missing appointments and why, without the need for a major new data infrastructure programme.
For example, a trust modelling this combination might find that two neighbourhoods in its catchment account for 40% of a directorate’s DNA-related lost capacity, with digital exclusion and transport access as the dominant drivers in one, and shift-pattern employment in another. That distinction changes what the intervention looks like: a transport-linked reminder and rebooking service in the first case, evening or weekend slots in the second. Neither is visible from attendance data or costing data alone; both come from combining the two with population-level segmentation.
Because Acorn operates at postcode level, organisations may be able to incorporate these insights within existing governance arrangements, subject to their own information governance requirements. Used effectively, this approach allows providers to move beyond retrospective reporting and towards identifying where risk is emerging, and where targeted intervention is likely to have the greatest impact, before appointments are lost.
The scale of the problem
NHS England estimated the cost of missed GP appointments at approximately £216 million in 2019, while around 8 million outpatient appointments were missed in 2023/24 alone, representing an estimated £1.2 billion annual cost to the NHS. The precise figures differ depending on the year and the methodology used, but the overall picture remains the same. Even relatively modest reductions in DNAs can release meaningful clinical capacity for most providers.
Reducing DNAs is not simply about operational efficiency. It is about making services easier to access, reducing inequalities and getting greater value from resources that are already under pressure. When providers understand who is missing appointments and the reasons behind that behaviour, they are in a much stronger position to design interventions that reflect the realities of people’s lives, rather than relying on a single reminder approach for populations whose circumstances differ markedly.
Understanding the opportunity is one thing. Working out how to get there, from mapping demand through to identifying hotspots and anticipating future risk, is another.
Discover the full four-step framework
Our guide, From cost to impact: A four-step guide to population intelligence, sets out how NHS providers can combine patient-level costing data with geodemographic and behavioural insight to pinpoint hotspots, understand risk and forecast future demand – without redesigning the entire service.
In our next blog, we’ll look at how NHS organisations can move from insight to action by tailoring interventions to the specific needs of different patient groups.
