Among the recommendations in the London Health Commission’s 2014 report "Better Health for London", chaired by Professor Lord Ara Darzi, was a call to concentrate specialist emergency care into fewer, higher-volume hospital units rather than spreading it thinly across the capital. A decade on, this is one of the few areas where London has hard, published outcome data to show what centralisation actually delivered — not just a policy intention, but measured mortality and survival figures. This piece sets out what changed, what the evidence says, and where the model’s limits show up.
The Recommendation Behind the Reconfiguration
The logic was straightforward: for conditions like stroke and major trauma, speed and specialist volume both matter. A patient treated by a team that sees severe cases daily, with immediate access to scanning, intervention and multidisciplinary staff, tends to fare better than one taken to the nearest general hospital. London had already begun testing this model before the 2014 report was published, and Better Health for London’s contribution was to endorse the approach as a template for how specialist emergency care across the city should be organised, rather than a one-off pilot.
What Was Actually Delivered
On stroke care, London’s services underwent what has been described as a radical centralisation starting in 2010. Instead of being taken to the nearest hospital, stroke patients across the capital were directed to one of eight 24/7 Hyper Acute Stroke Units (HASUs), positioned so that no Londoner was more than a 30-minute ambulance journey from one. After initial treatment, patients moved on to one of 24 Stroke Units for rehabilitation. Five hospitals stopped providing acute stroke services entirely as a direct result of the reconfiguration — a politically difficult change, since it meant closing a service some local hospitals had offered for years.
On trauma, London stood up the UK’s first regional trauma system in April 2010, covering a population of around 10 million people. Hospitals were designated either Major Trauma Centres (MTCs) — including the Royal London, Imperial, King’s College and St George’s — or Trauma Units handling less severe injuries. MTCs carry 24-hour access to specialist teams, radiology and operating theatres, with consultant-led trauma teams meeting patients on arrival rather than routing them through a general take.
The Evidence, a Decade On
What sets this reconfiguration apart from many other health-system changes is that it has actually been studied. The table below summarises the main published findings.
| Metric | What the evidence shows |
|---|---|
| Stroke service structure | Care consolidated from many individual hospital admissions into 8 designated Hyper Acute Stroke Units, all open 24/7 |
| 90-day stroke mortality | UCL-led research published in the BMJ (2014) found a 1.1 percentage point reduction in death rates, equivalent to an estimated 96 extra lives saved per year; a 2019 follow-up analysis found the improvement was sustained |
| Average stroke hospital stay | Reduced by roughly a day and a half per patient, an estimated saving of 12,766 hospital bed-days a year across London |
| Rapid brain imaging after stroke | 72.1% of London patients received a brain scan within 3 hours of admission, against 65.2% in Greater Manchester, a comparator system with a different model |
| Trauma network structure | UK’s first regional trauma system, launched April 2010, covering roughly 10 million people, with named Major Trauma Centres receiving the most severely injured patients |
| Major trauma survival (London) | A 2015 study found survival rates rose by 50% over five years, an estimated 610 additional lives saved; a further analysis reported a 23% reduction in mortality among major trauma patients |
| National trauma outcomes post-2012 rollout | A 2018 study found England’s major trauma centres, modelled on the London network, saved an additional 1,600 severely injured patients since 2012 — a 19% increase in the odds of survival for patients who reached hospital alive |
Taken together, this is one of the more consistent evidence trails behind any single recommendation in the 2014 report: independent, peer-reviewed research, from more than one research group, over more than one time horizon, pointing the same direction.
What Stalled, and What the Model Doesn’t Solve
The stroke and trauma reconfigurations succeeded partly because they were narrow and measurable: a defined patient group, a defined pathway, and outcomes that are relatively easy to track (did the patient survive, how long did they stay). Not every service the 2014 report touched on had that shape. Broader debates about consolidating other hospital departments across London have tended to be slower and more contested, in part because the case for change is harder to make in a single, citable statistic, and in part because closing a local service is a harder political conversation than opening a new specialist unit alongside it.
It is also worth being precise about what the stroke and trauma evidence does and does not show. The published research demonstrates that centralisation improved outcomes for the specific conditions studied, under London’s particular geography and ambulance times. It is evidence for a model, not proof that concentrating every kind of care always produces better results everywhere. Population density, ambulance response times and the availability of specialist staff all shaped why this worked in London; a smaller city or a more rural area could see a different balance of costs and benefits.
Why This Still Matters
Ten years after the report that helped popularise this approach, stroke and trauma centralisation stands out precisely because it can be checked. The commitments were specific, the outcomes were measured by independent researchers, and the results have been revisited more than once rather than claimed and forgotten. For a report remembered mainly for its broad ambitions on prevention and the wider determinants of health, this narrower, harder-edged piece of delivery is arguably where the 2014 recommendations are easiest to hold up against what actually happened.