NHS 111 is the front door of urgent care in England: a free, 24-hour triage line and online service designed to route callers toward the right level of help — A&E, an urgent treatment centre, a GP, a pharmacy, or self-care — before they ever reach an emergency department. It is one of the few pieces of system design explicitly built to relieve A&E pressure. The question worth asking, a decade on, is whether it actually does.
How the Triage Decision Is Made
Callers to NHS 111, and users of the 111 online tool, are worked through a structured set of questions built on NHS Pathways, the clinical algorithm behind the service. It does not diagnose; it classifies urgency and routes accordingly. Depending on the answers, a caller can be booked directly into an A&E slot, an urgent treatment centre, an out-of-hours or same-day GP appointment, referred to a pharmacy, told to expect a nurse callback, or advised on self-care. NHS England has said that over half of callers end up speaking to a clinician — a nurse, doctor, pharmacist, or paramedic — rather than a call handler working from the script alone. In principle, that clinical layer is what should catch the cases that don’t need an emergency department.
The Case for Diversion Working
The headline evidence is favourable. NHS England reported in 2019 that the 111 service had saved more than 12 million unnecessary A&E visits between its 2011 launch and September 2018, based on an analysis suggesting that roughly 28% of callers would otherwise have gone to A&E. A separate Health Foundation analysis of three London areas, focused on callers aged 15 and under, found that when a GP or nurse reviewed the case by phone, those callers were around 14% less likely to attend A&E than callers simply advised to manage at home — and around 68% less likely to attend a minor injury unit specifically. On paper, that is exactly the effect a triage-first system is supposed to have.
The Case Against: Where the Evidence Gets Uncomfortable
The same Health Foundation work found no significant reduction in attendance at major A&E departments — the clinical reassurance effect showed up for minor injury units, not for the departments actually under the most pressure. Earlier evidence is less encouraging still: a controlled before-and-after study of the original 111 pilot sites, which included London, found no overall change in emergency department attendance in the service’s first year, alongside a small increase in ambulance incidents. A Nuffield Trust review covering 111’s first three years found that the number of callers advised to go to A&E or sent an ambulance actually rose, both in raw numbers and as a share of all calls. Some analyses have also pointed to call handlers, particularly non-clinical ones, defaulting to more risk-averse advice — which tends to push borderline cases toward, not away from, emergency care.
"111 First" Added a New Layer, Not Necessarily a Cleaner One
The "111 First" initiative, introduced in late 2020, tried to sharpen the model by encouraging patients to call or go online before attending A&E at all, with direct booking into urgent care slots. Surveyed users generally found it useful when it worked as designed. But emergency department staff also reported that their own in-person triage processes kept running in parallel — meaning 111 First became an additional assessment step layered on top of existing hospital triage, rather than a replacement for it, with the delays that implies for some patients.
What the Mixed Record Actually Means for System Design
None of this makes NHS 111 a failure. It is a genuinely large-scale attempt to insert clinical judgement between a symptom and a hospital visit, and the data shows it does change behaviour for a meaningful share of callers, particularly for the least severe presentations. What the evidence does argue against is treating a single call-before-you-go service as sufficient, on its own, to relieve pressure on the departments that matter most. The Health Foundation finding — real reduction for minor injury units, none for major A&E — is the clearest single data point on this: triage at the front door reshuffles low-acuity demand more reliably than it reduces the caseload landing on the busiest departments.
| Source | Scope | Finding |
|---|---|---|
| NHS England (2019) | National, 2011–2018 | 12M+ visits diverted; ~28% of callers would otherwise have attended A&E |
| Health Foundation (2018) | Three London areas, under-16s | Clinical review: 14% less likely to attend A&E overall; 68% less likely for minor injury units; no significant effect on major A&E |
| 2013 pilot evaluation | Original pilot sites incl. London | No overall change in A&E attendance in year one; small rise in ambulance incidents |
| Nuffield Trust (2017) | National, first 3 years | Share of callers advised to attend A&E or sent an ambulance rose over time |
For anyone designing or commissioning urgent care capacity across London’s boroughs, the lesson is not that call-first triage is the wrong idea — it is that its measurable effect is concentrated at the lower-acuity end of demand. Relieving pressure on the departments that are actually overwhelmed still depends on what sits alongside NHS 111: same-day GP capacity, urgent treatment centres people trust enough to use, and community pathways that can absorb the cases a phone algorithm correctly identifies as not needing a hospital bed, but cannot itself provide a place for.