First do no harm: lessons from service reconfiguration in London

London has been in the eye of the storm of health service reconfigurations. Work undertaken following Lord Darzi's 2007 review has begun to tackle well known weaknesses in how services are provided and a recent report from The King’s Fund on improving health and health care in London identified progress made in implementing the review's recommendations.

We also drew attention to the amount of unfinished business in London and asked who would lead this work following the abolition of the strategic health authority (SHA) and primary care trusts.

At a conference of health service leaders on progressing health care in London held at the Fund last week, the SHA's chief executive, Ruth Carnall, reflected on her experience of leading service change in the capital. Ruth emphasised the importance of articulating a clear and strong case for change to underpin work of this kind. She also highlighted the role of clinical leaders in taking forward this work with the support of experienced managers.

The example of stroke care by Tony Rudd, where outcomes have improved through the concentration of specialist stroke care in eight hyper-acute units, is a case in point. The progress made in stroke care illustrates another lesson, namely the need to work at scale to make some changes happen. In this case, collaboration both between providers and between providers and commissioners across London was required to secure agreement on where specialist services should be located.

Ruth also reflected on the disappointments and challenges experienced along the way. These included work still outstanding in some key areas such as cancer care, mental health and primary care. Even more important was the setback that occurred after the 2010 election when Andrew Lansley stepped in to halt the programme of work the SHA was leading. Notwithstanding the Health Secretary’s intervention, much of this work continued because NHS leaders and clinicians judged it was the right thing to do.

Important lessons from London's experience include the length of time it takes to bring about changes in care and the risks that arise when national and local politicians intervene inappropriately. In the case of stroke care, several hundred lives might have been saved had changes occurred sooner. Similar arguments apply to changes in the role of local hospitals which, in the case of Chase Farm in north London, took around 20 years to be finally agreed. These lessons take on added importance at a time when the provider landscape in London is, in Ruth Carnall's judgement, simply unsustainable.

On the issue raised in our 2011 report – who will lead work on service reconfigurations from next April – the jury is still out. In her presentation, Anne Rainsberry, regional director of the NHS Commissioning Board in London, reiterated the need for change to be led both bottom up and top down. In north west London, this is being done through the emerging clinical commissioning groups (CCGs) collaborating with each other to make the case for major changes in hospital services with the support of the NHS Commissioning Board. While this example is encouraging, Anne acknowledged that not all CCGs were yet in a position where they can work in this way.

One of my reflections on the day was the need for politicians to think very carefully about their involvement in service changes because delays and compromises can cost lives. While the urge to protect constituents’ interest is strong and understandable, it should be resisted where there is a clear clinical case for change. In considering what are ultimately life and death issues, councillors, MPs and ministers would do well to adopt one of the fundamental precepts of medical practice – 'first do no harm'.

Keep up to date

Subscribe to our email newsletters and follow @TheKingsFund on Twitter to see our latest news and content.


#3353 Dr Michael Crawford

Writing from a non-metropolitan perspective it seems that a thorough evaluation of the hyper acute stroke service would be an opportunity to test out the validity of some of the assumptions underpinning service reconfiguration in London; it is certainly necesary to do this before looking at recoinfiguratuion elsewhere.

Moynihan and colleagues showed that the St George's hub and spoke service was assocated with a quntupling of the thrombolysis rate but more detailed information about who does not receive this tratmetn is appropriate.

If you look at the map of stroke hubs in London it is clear that teaching hospital vanity rather than patient access was the main factor in their location. Five famous teaching hospitals (plus temporarily a 6th) are crowded around the central area with three token hospitals in the wide area outside the North & South Circular Roads. It is essental that future research does not look at institutional activeity but at the service delivered to incident cases in the population. Is there a distance decay effect? What is the effect of socioeconomic deprivation? This might be complex with poorer people having a completely different access pattern from that enjoyed by the affluent.

There is a lot to learn from cancer services. In the UK, NHS cancer institutions have earned an international reputation for excellence but when one measures UK residents' survival at the population level the NHS falls well behind the more widely disseminated services in Continental Europe.

We would not dream of inroducing a new treatment without assessing it objectively with proper clinical trials. We happily introduce new patterns of service based on observational studies and the say-so of politicians and academics.

Barry Moynihan, et al. J R Soc Med. 2010;103:363-369.

#3380 Mary E Hoult
community volunteer

I have real concern with some of the statements in this article and do wonder what is happening to the democratic process !! we in the Community expect representation from our MPs and nobody would argue that there needs to be change but how it is managed and communicated is the key to it's success.
Bullying MPs and the Community is not the way forward.

#3405 david robson
chair of trustees
Greenwich and Bexley Community Hospice

The title of the blog starts with 'First do no harm' and in her talk Ruth mentioned the failed South London merger.She gave as one of three major causes of the failure that the wrong plan was adopted due to 'political pressure' and she expressed regret that this was so.It is right to note that considerable harm has been done especially to the staff whose efforts at developing teamwork and common policies have involved a great deal of effort and time.This effort and commitment will most likely be thrown away in a new reorganisation.The phrase 'lions led by donkeys' comes to mind.

#3858 Tony Rudd
London Stroke Clinical Director
NHS London

Responding to Dr Michael Crawford, I can reassure him that the London Stroke project has undergon evaluation. A health economic study has been conducted by a team from UCL Partners and we are currently also conducting a study funded by NIHR comparing the reconfiguration in London with the one that happened in parallel in Manchester. The economic study has been submitted for publication and shows that the redesign has not only saved lives, but despite the investment made, has resulted in saving considerable amounts of money. The process by which the hyperacute centres were chosen was rigorous, based on a combination of the quality of the bids put in by the trusts (not all bid to provide the service) and on geographical considerations. Until the reorgansiation the vast majority of hospitals providing any sort of specialist acute care were the inner London teaching hospitals. One of the key criteria that was set in the development of the model was that all residents of London should be within 30 mins travel time of a hyeracute unit and this has been achieved with an average travel time of about 15 minutes from home to hospital. To describe BHRT, SLHT and Northwick Park as token hospitals is grossly unfair. The work that has gone on in those hospitals over a very short space of time to get them to be able to deliver the highest quality of care has been remarkable but has required considerable investment. All the other hospitals delivering on going care after the hyperacute phase in acute stroke units have also made fantastic gains and these include inner London teaching hospitals and outer London DGHs.Their work is at least as important as the HASUs. Comparing us with Europe we are probably now delivering care equal to the best and a lot better than most. 100% of hospitals in England now have a stroke unit, nearly 90% of stroke patients are now treated on one, nationally we are thrombolysing over 8% of unselected patients and in London this is nearer 18%. I don't think Dr Crawford should be so negative about what has been a fantastic success story.

#3902 Chris Clark
UK Director Life After Stroke Services
Stroke Association

The reconfiguration of stroke care in London has been a remarkable achievement, and the leadership of the SHA, support across the PCTs and dedication of all the clinicians involved was a model.

I would also suggest that the SHA were also a model in embracing and including the voluntary sector and people affected by stroke throughout the exercise. I believe this made a difference to the planning.

However looking at areas such as Greater Manchester we are conscious that further attention is required to improve long-term care. The job has not yet been completed.

#6602 Dr Michael Crawford

The precise questions I put were:-
"It is essental that future research does not look at institutional activity but at the service delivered to incident cases in the population. Is there a distance decay effect? What is the effect of socioeconomic deprivation? This might be complex with poorer people having a completely different access pattern from that enjoyed by the affluent." (typo corrected!)

Have these been addressed in the evaluation?

#12630 Dr Michael Crawford

I know it’s been three weeks since the original posting but there is an important issue here. The Hippocratic headline of the blog is “First do no harm” and the point about the use of stroke care as an example is that the policy has done good. Has it done enough good?

When one starts a new service there will be good done because of the virtues of the treatment on offer (thrombolysis) and because of the resources attached – “considerable investment” in Tony Rudd’s phrase and no doubt the ongoing care work attracted some investment.

As a consequence there will have been low-hanging fruit to be gathered; the real challenge is about access to the upper branches. Performance of individual institutions does not measure this, one has to look at the people who lose out. My own interest is in cancer services where the leaders of high-performing Cancer Centres have found the truth hard to swallow that the NHS as a whole performs extremely poorly through its failure to make diagnostic and treatment services available throughout the social spectrum. The configuration of centralised radiotherapy services with peripatetic consultants visiting other hospitals and centralised thoracic surgery with referral from general hospital physicians dates from the Hospital Plan of 1960; it did not achieve geographical equity. It would be wrong to fail to ask this question about the new stroke services especially if they are to be a beacon for future configuration of metropolitan health services.

So the questions I asked in my original post remain valid. If the service was not designed around a criterion of 30 minutes maximum travel time but around the placing of 8 centres so as to minimise the travel time for any London resident, what would it look like? One does not need a deep understanding of statistics to know that the phrase “an average travel time of about 15 minutes from home to hospital” can conceal a lot. If “average” means “mean” it implies that 100% of patients who should have access did so; a great achievement; no doubt the standard deviation will be in the published report. If there were some patients who should have had access but did not attain it, the mean cannot be calculated and we need to know the proportion who did not get to the hyperacute unit, the median and interquartile range of the travel time.

The stroke service in London is a great achievement and I hope it is true that it leads Europe. Its laurels should not be rested upon!

#13387 Tony Rudd
London Stroke Clinical Director
London SHA

I do not think that the new system for stroke care in London has resulted in persisting inequalities in access to high quality care. We know from the SINAP acute stroke audit run by the Royal College of Physicians that there is about a 97% concordance between HES data for stroke and the entries on SINAP. There are very few patients that are being coded as having stroke using routine hospital statistics that are not being included in the SINAP audit and about 95% of admitted stroke patients in London are going directly to a HASU. There are only a tiny number of patients whose travel times from the place of stroke to a hospital with a HASU exceeds 30 minutes (less than 5%) and these cases are regularly reviewed to identify a cause. The main reason for delay is difficulty establishing a diagnosis and not related to travel delays. The location of the hospitals is not a significant factor in whether a patient gets access to high quality care.
Yes, if were designing healthcare in London from scratch we would almost certainly not place all the hospitals where they currently are. We have to work with the resources we have and it would have not made sense to cease hyperacute stroke care in all of the central London teaching hospitals. Patients do sometimes need access to neuroscience centre facilities. Established teams of experienced specialists (not just doctors but nurses, therapists, radiographers etc) are easy to destroy but not easy to create and so it did make sense to focus care where possible on the established high quality services. However, as already stated we did need to address the geographical inequalities that existed prior to the reorganisation and we believe we have done that in the most efficient and sustainable way. The quality of care being provided continues to improve, particularly with regard to community care. Most districts now have early supported discharge teams and the quality and quantity of longer term rehabilitation teams has also undergone a transformation. .
I have no doubt that centralisation of hyperacute care was the correct decision and a model that should be adopted for other diseases that require access to 24/7 specialist care such as vascular surgery and acute kidney injury. Maintaining the initial improvements into the longterm will always be a challenge but so long as the support of the clinical networks and hopefully a clinical director is maintained I am optimistic that not only can we keep the existing quality but also make further gains.

#15095 Dr Michael Crawford

It is very encouraging to have an assessment demonstating the success of the project based on evidence, not just assertion.

The comment that this is "a model that should be adopted for other diseases that require access to 24/7 specialist care such as vascular surgery and acute kidney injury " is an extrapolation too far. Stroke is an illness whose presenting symptoms are characteristic enough to be shown in a national advertising campaign but in spite of this "the main reason for delay is difficulty establishing a diagnosis. "

Leaking aneurysms and AKI are in a different league diagnostically so reconfiguraton would demand its own very careful evaluation to ask the questions I posed above. Lessons from cancer ae that there are differences between tumours with characteristic symptoms and those that are hard to identify against the background noise of symptomatology - patients with endometrial cancer get better access than those with ovarian cancer, rectal cancer is better than colonic cancer in getting appropriate treatment. The differential is greater for poorer people and there is some evidence that for the economically dsadvantaged group, travel time is a significant issue.

#17216 Gerry Ramsden
Healthcare Scientist

I agree wholeheartedly with Michael Crawford.Whilst this model of treatment can give exceptional treatment in the urban areas of England it does not address the service provision to areas where the demographic is different.
The article speaks of "30 minute" delay, I the are of England I live in it is at least 30 minutes before a patient can arrive at the treatment unit. More care needs to be given to all our patients wherever they are based rather than the same old London-centric thinking that is the downfall of most of the healthcare processes in the urban,semi-rural and rural populations trying to access good quality local healthcare.
Perhaps thesecareas should look to other countries to provide models of healthcare appropriate to their setting,: Sweden, Canada. In Sweden there is a treatment pathway that include paramedic assessment and treatment before attending hospital units.
The NHS and heathcare provision is changing and the management and treatment pathways have to change as well.
A a scientist I am told to measure, change, measure again, assess, this process has been lost in many movers and shakers leaving a directionless local NHS.

Add new comment