Why We’re Measuring the Wrong Thing in Elective Recovery
February 16, 2026
.svg)
June 4, 2026
Almost every Integrated Care Board (ICB) in England can point to some form of Advice & Refer. On paper, it is widely available. In practice, it is rarely consistent.
There is a fundamental difference between having A&R available and operating a coherent, system-wide model that actively shapes demand. One is a tool. The other is infrastructure.
As the NHS continues to grapple with record elective backlogs, still in excess of 7 million, and sustained workforce pressure, A&R is increasingly positioned as part of the solution. NHS England has explicitly framed Advice & Refer as a mechanism to reduce unnecessary referrals and ensure patients receive care in the most appropriate setting, first time. But that potential is only realised when A&R moves beyond local variation and becomes embedded at system level.
So what does “good” actually look like?
In many ICSs, A&R has developed organically. A motivated consultant sets up a responsive dermatology service. A GP federation adopts it enthusiastically. Another specialty, in another trust, responds sporadically, or not at all.
The result is a patchwork.
From a clinician’s perspective, this inconsistency creates friction. From a system perspective, it creates risk. Practices cannot rely on predictable response times. Clinical decisions are made with varying levels of specialist input. And commissioners lack a clear view of performance across providers.
A strong ICB-wide model addresses this head-on through standardisation.
That does not mean imposing rigid uniformity. It means establishing a consistent baseline: agreed response service-level agreements (SLAs), structured submission formats, clear specialty coverage, and defined lines of clinical accountability. These are not bureaucratic add-ons. They are the foundations of safety and trust.
There is a good reason for this emphasis. Research highlighted by NHS Confederation has shown that more digitally mature organisations tend to deliver more efficient care, including lower costs per patient episode and shorter lengths of stay. While A&R is only one part of that maturity, the principle holds: consistency and standardisation are what allow digital tools to translate into measurable system performance.
Without that, A&R remains dependent on local enthusiasm, and local enthusiasm is not a scalable model.
Even where A&R is working well locally, many systems struggle to answer a simple question: what is it actually doing at system level?
This is where most models fall short.
Good A&R is not just about enabling communication between primary and secondary care. It is about generating a clear, system-wide view of demand and decision-making. That requires unified reporting across providers.
At a minimum, an ICB should be able to see:
Without this, A&R remains an operational tool. With it, it becomes a demand management lever.
The difference is not academic. If commissioners can see, for example, that a high proportion of cardiology queries are still converting to outpatient referrals, that is not a failure of A&R, it is a signal. It may indicate unclear referral criteria, insufficient specialist capacity, or gaps in primary care confidence. But without visibility, that insight is lost.
The broader NHS policy direction reinforces this. Organisations such as NHS Providers have repeatedly emphasised the importance of system-level data in enabling ICBs to manage performance, allocate resources effectively, and redesign services. A&R data, when properly captured and shared, is one of the clearest windows into how demand actually flows.
In that sense, data is not a byproduct of A&R. It is the mechanism through which it delivers value.
One of the more uncomfortable realities of informal A&R is that it often happens anyway, just outside formal systems.
Emails, corridor conversations, phone calls. These interactions can be clinically helpful, but they are rarely auditable. There is no consistent documentation, no clear delineation of responsibility, and limited visibility if something goes wrong.
At a small scale, this is manageable. At a system scale, it is a governance risk.
A well-designed ICB-wide model brings these interactions into a structured, auditable environment. Every query is documented. Every response is attributable. Communication is standardised and automatically recorded within the patient record.
This is not simply about compliance. It is about protecting clinicians as much as patients. Clear documentation reduces medico-legal exposure and ensures that decision-making is transparent.
Crucially, governance should not be retrofitted after incidents occur. It needs to be embedded into the workflow itself, making the safest option also the easiest one.
Perhaps the most important shift in a mature A&R model is its relationship with pathway design.
In less developed systems, A&R sits adjacent to formal pathways. It is available, but optional. Clinicians use it when they feel it is helpful, but it does not fundamentally change how referrals are structured.
In a high-performing model, that relationship is reversed.
A&R becomes part of the pathway itself. Referral criteria are aligned with A&R usage. Specialists and primary care clinicians collaborate to define when advice should be sought and when a referral is appropriate. Over time, this reshapes demand.
For example, in specialties such as dermatology or musculoskeletal (MSK) services, well-implemented A&R models have demonstrated that a substantial proportion of cases can be safely managed in primary care with specialist input, reducing unnecessary outpatient activity. NHS England has actively promoted this approach as part of its elective recovery strategy, recognising its potential to reduce pressure on hospital services.
This is where A&R moves from being a communication tool to a front-door filter for demand. And that shift is what ultimately delivers system-level ROI.
A recurring failure mode in NHS transformation is over-reliance on individuals. A strong clinical lead champions A&R within a specialty. Response times improve. Adoption increases. But when that individual moves on, performance drops.
That is not resilience. It is fragility.
A good ICB-wide model is designed to function regardless of individual turnover. It works across multiple trusts, maintains consistent standards, and can absorb fluctuations in demand—whether seasonal spikes or longer-term growth.
This requires more than good intentions. It requires infrastructure: shared governance frameworks, aligned incentives, and consistent service expectations across organisations.
It also requires recognising that scale introduces complexity. What works in a single trust may not translate directly across an ICS without adaptation. But without designing for scale from the outset, even the most successful local models will struggle to deliver system-wide impact.
Ultimately, the test of any A&R model is whether it can demonstrate impact clearly and consistently.
A mature, ICB-wide approach should be able to answer a set of straightforward but revealing questions:
If these questions cannot be answered with confidence, the model is not yet delivering at system level.
This is where many systems still have work to do. Not because A&R is ineffective, but because its impact is not consistently measured or articulated.
Advice & Refer has been part of the NHS for years. What is changing now is the expectation placed upon it.
It is no longer enough for A&R to exist. It must perform.
At ICB level, that means treating it not as a digital add-on, but as a core component of demand management, supported by standardisation, underpinned by governance, and driven by system-wide data.
When those elements are in place, A&R stops being a workaround. It becomes infrastructure.
And that is what “good” looks like: not pockets of excellence, but consistent, measurable impact across the system, shaping demand, supporting clinicians, and improving patient flow at scale.
See how you can implement A&R at a system-level using Cinapsis.
.svg)
