Reconfiguring services is back on the agenda what does the evidence tell us

Reconfiguring services is back on the agenda what does the evidence tell us
Reconfiguration is back on the menu everywhere.

Some would argue it was never off the menu.

I often get asked – what does the evidence tell us – and have answered this in a number of guises over the years. I don’t profess to have THE answer, but here are some thoughts on evidence and the contextualisation of it.

Regardless of any future model, the Five Year Forward view sets out a vision that future reconfiguration may be necessary.
There is a gradual shift of the rhetoric about the future of hospital care (for eg the RCP Future Hospital Initiative, the Five Year Forward view). This shift of rhetoric is away from the “all-or-nothing” approach to the future of acute hospital care towards hospitals that are more integrated with primary and tertiary care.
Achieving the recommendations of the Royal Colleges and other bodies may not be possible with the current (and potential future) workforce and growing need. Some service change may be required.
Those NHS hospitals are under mounting financial pressure and face a major productivity challenge is well documented. A common belief is that the reconfiguration of hospital services, primarily through rationalising services across sites and shifting services into the community will help resolve these pressures.
Monitor had expressed a view that “The evidence suggests that reconfiguring services and integrating care more effectively across providers could yield productivity improvements in the region of £2.4 billion to £4 billion by 2021”. It is of note that little evidence is put forward in this Monitor document to support this.

Here I haven’t got into the economics of shifting care into the community. Maybe in a future blog……
This blog is a very superficial overview of the available evidence on potential service design options for the future. It is based on experience and some (but by no means all) of the available evidence.



Introductory thoughts – what level of reconfig – procedure to whole system?
Much of the literature here pertains to whole hospital type reconfiguration rather than specific bits of a hospitals business. A great deal of the literature on reconfiguration is drawn from overseas, where the underpinning system and cultures are different. Finally much of the available literature relates to adult services rather than paediatric care. Some judgement and care is therefore required in interpreting available evidence.
There’s two levels of this question (maybe 3) – procedure level, specialty level, whole org level




Procedure level – highly procedure specific. Need to take great care, and many unanswered qs. I haven’t considered procedure specific volume / outcome relationship issues in the blog

On the question of procedure specific, a colleague of mine posted a note once on a systematic review she had conducted on volume/outcome in relation to cancer and the evidence was at best equivocal. Problems included:

lack of multivariate analysis

low cut off for high volume – either for individual surgeon or institution, because there were too few very high volume institutions or surgeons

problem that 30 day post-operative death rates were low and falling after the 70s, other outcomes – local recurrence and surgical complications were not well documented

a lot of the emerging professionally driven audit data that could have analysed volume/outcome didn’t – ?failure to do this analysis or publication bias (no volume effect shown and because this was counter-intuitive the analysis was excluded)

However there was and remains a most effective use of clinical resources argument: WTD, team working, training – juniors + new technologies, ITU, MDT decision making, and how all these can be delivered in the context of the (much loved by the public) DGH model and estate we already have.

two key refs – dug out by Mark Lim

Davoli – review of reviews

Mark summarised the key data – as follows








Specialty level – linkages between specialties and services become critical. So you wouldn’t have xxxx without yyyyy being onsite etc.

NHSE have done loads of very helpful work in this space
The Morris et al work on stroke reconfiguration is the benchmark here.

Most of it published on the NIHR website. It’s amazing. Go read it.

Ramsay et al. Stroke. 2015;46:2244-2251

Morris et al. BMJ 2014;349:g4757

Are two of the key refs

The Boston Consulting /MONITOR report looks worth a read ( I haven’t yet)

I’ve heard some folk say some less than complementary things about this but I’ve not been through it so you will need to draw your own conclusions
Little clear evidence to suggest that, as the range of services a hospital offers expands, the cost per unit reduces (so-called economies of scope)







Macro is where the fun starts.

Key refs

Concentration and Choice is the classic in this field.

There were three parts to this that are quite hard to find in the Internet. They are there somewhere!
The CRD 1996 bulletin is still worth a read
As is this more recent paper by Harrison

NB – paywalled

However, the daddy – or in this case the mummy is the Imoson study

This is definitively is the state of the art

It’s like gold dust. Consider it a trade secret!
There was a Kings Fund version of this, but simplified messages
SELECTED key points

NIHR study on service reconfiguration based on the experience of the NCAT.

The NIHR study chimes will with a broader body of evidence on reconfiguration.
The study neatly summarises the available evidence on reconfiguration of health care at service and system level, using a number of clinical areas that are the most common targets of reconfiguration.
The key points raised in the Imoson study can be summarised as follows:
• Since 2007 the primary drivers of service reconfiguration have been medical workforce and financial pressures. Quality drivers have been subsidiary and often linked to workforce numbers.

• The future economic environment for the NHS, alongside a reduction in hospital doctor training numbers, suggests that the pressures to reconfigure hospital, mental health and community services can be expected to grow.

• There will be further policy pressures to reconfigure emergency and specialist services. The strengthened role of the CQC in hospital quality inspection, especially if inspecting against College standards, could be an additional catalyst for service reconfiguration.

• The NHS is continuing to concentrate many district general hospital services to resolve financial and workforce pressures. However, many proposals are not implemented owing to public opposition.

• This study found no evidence to suggest that major reconfiguration will deliver the savings anticipated.

• There is a significant gap in the evidence about safe staffing models and the appropriate balance of junior and senior medical as well as other clinical staff. There is an urgent need to carry out research that will help to fill the current evidence gap.

• There is also an absence of evidence about safe staffing models and the interplay between staff numbers, skill mix and outcomes. We found that the advice provided by the NCAT reflects the current evidence, but one of the NCAT’s most valuable contributions has been to encourage greater clinical engagement in service change.

The Imoson study found strong evidence that some specialist service reconfiguration (for example vascular surgery and major trauma) can significantly improve clinical outcomes.
There is obviously a wealth of information about the benefits and dis-benefits of reconfiguring services that is not picked up in the Imoson study. One might characterise some of this as opinion rather than evidence. There are a number of major evidence gaps that the Imoson neatly highlights. Chief of these is the hypothesis that reconfiguration will save money, from any perspective. The evidence doesn’t bear this out – either through centralising hospital care or replacing them with community services.
Imoson finds that there is good evidence to support the centralisation of many specialist services. There is a clear evidence base underpinning the volume / outcomes hypothesis in many surgical interventions – however this is often procedure specific and there isn’t the evidence to generalise from a procedure to a sub speciality. It is also highlighted that the benefits often rely on much more than a simple link between volume and outcomes. The ways of working within and across specialties can be just as important (a message that is true for all services, specialist or not).
Workforce is the other most often cited main driver for service change. The desire to move from a consultant-led to a consultant-delivered service. While there is strong evidence to support the benefits of more consultant-delivered care, particularly for high-risk patients, there is little evidence to say how many senior staff are needed, of what type and for what time periods.


My general sense of the general lay of the land is

1) the evidence is conflicted

2) nuanced

3) can be read in many ways, depending on what your agenda is

4) and varies intervention by intervention

5) in general there is some relationship between volume and outcomes, ESP once case mix adjusted! That can be related to operator characteristics (more you do better you get), case mix, system characteristics (operator only as good as the system she operates in),

6) immense care needs to be taken to avoid selective interpretation! There are competing interests and conflicts of interest (on all sides) that make this hard

7) big shift (on account of mixed and not compelling evidence) needs to be taken in broader population context……the example that springs to mind is vascular review and implications for service change having implication for loss of rapid access to the renal guys for fistula etc etc….which of the various competing goals carries more importance. No evidence on that one….value judgement needed, maybe backed with Epi, but usually not Big change does carry potential for unintended consequences and harm…..

8) volume utcome debate also needs contextualising in geography, paramedic and pre hospital care and all that. Care needed re selective interpretation

3) Context is critical …….read ‘Why Strategy Matters Now’ – Porter and Lee. NEJM
Porter and Lee recently set out the five key questions on strategy that all provider organisations (and payers) might ask. The conceit to this piece that as funding gets tighter, providers can rely less and less on tariff income covering costs and thus different models are needed. Success in the future will require a relentless focus on populations, value and systems; and not on simply maximising volume. Good operational performance remains important but new models are needed.


Porter and Lee suggested six questions of some importance to providers and networks of providers – these are set out below, based on the NEJM article but adapted slightly for the purposes of this work in South Yorkshire.


An important précis to these six questions is the notion of what IS the organisation – is it individual provider or a network of providers.


1“What is our goal?”

Considering the organization’s fundamental purpose and definition of success.

Is the organisation the primary concern or the broader system.


2 “What businesses are we in?”

volume, outcomes, value.

Increasingly value for the population (treated and untreated) will be seen as important. What does this look like in the context of surgery for a population of c500,000 people aged 0-17, in the broader context of paediatric care.


3 “should our scope should be narrowed or broadened”

what set of conditions and patient populations should we compete in?

Can we meet every need of every patient in our catchment who is referred for treatment.

Should we specialise.

Should we collaborate with other providers to specialise in mutually beneficial areas.


4“In every business where we choose to compete, how will we be different?”

what is our unique proposition.

Can we and should we compete for customers. Might this lead to price or quality erosion and poorly utilized capacity.

Should we develop managed networks across all our capital in the population we serve.

In Porter and Lee language this would be an integrated practice units (IPUs). These were first set out in the HBR article on “the Strategy that Will Fix Health Care”.


5 “What synergies can we create across our existing business units and sites?”

All complex organizations require strategies at two levels — for each business unit and for the overall corporation.

larger multisite organizations can amplify patient value through system integration.

Condition-level strategies and system-level strategy should go hand in glove.

Can we sweat our capital assets and make the most of available human labour by collaborating across multiple organisations the collectively serve a larger population.

This may require difficult choices about which unit will specialise in certain forms of care to eliminate duplication and “excess” capacity; inherent in this is shutting down some sites, and shifting care to lower-cost settings. This will require us to confront issues of ego and politics.



6“What is our appropriate geographic density and scope?”

Does the system have the appropriate concentration and types of services and sites?

Would establishing off-site ambulatory care locations enhance value?

Does the organization’s geographic footprint maximize value?

How broad a region is needed to assemble the volume in a particular condition required to achieve superior value for patients?

Are mergers necessary to build the needed volume, or should the organization expand through partnerships and affiliations?

Such decisions must always revolve around increasing value, rather than revenue alone.

Expanding, merging, and partnering are not strategies, but potential tools for improving value at the condition and system level


Porter and Lee suggest these six questions are interdependent, and the choices must reinforce each other. Making difficult choices should be with the primary goal of value and sustainability not cost control.



Wider evidence base

NB – some of my references appear to have disappeared here. Email me if you want them.
A Bristol analysis in 2012 examined whether this promise of large scale reconfiguration led to improvements in care. Between 1997 and 2006 in England a significant number of general hospitals were involved in a merger. The study examined the impact of mergers on a large set of outcomes including financial performance, productivity, waiting times and clinical quality. The study found little evidence that mergers achieved gains other than a reduction in activity. It was concluded that further merger activity may not be the appropriate way of dealing with poorly performing hospitals.

The Bristol study concluded that mergers did reduce available capacity, they had little or no impact on clinical quality or productivity. They also appeared to deliver some negative outcomes, such as increased treatment waiting times and a decline in financial performance post-merger.
Fulop conduced in depth interviews with ninety six senior staff in 13 organisations involved in reconfiguration to explore the impact of mergers and the effect on management costs. The study found a number of negative consequences of large scale service change, including loss of management focus, delayed (or non) implementation of developments and difficulties in merging very different cultures. There was limited evidence of any cost savings.
These may be issues of poor execution and implementation, poor match between perceived benefits and actual effect or other untested factors at play.
A recent JAMA editorial pulled together some evidence from the USA on hospital consolidation and concluded that larger size is neither a necessary nor sufficient condition for hospital systems to trim waste and enhance quality. In fact, studies show that greater competition, not consolidation, is more likely to hold down costs and lead to better care.
Tsai and Jha argue that quality improvement comes not from size, but from leadership. Smaller institutions can implement inexpensive but highly effective quality improvements, such as surgical checklists, as well if not better than larger organizations can.
Though not an issue in the UK on account of national tariff, a 2012 evidence synthesis found that larger consolidated providers led to higher prices and a greater concentration of provider power.
The extent to which these lessons are transferable from macro service change to any change in paediatric surgery service models of pathways needs careful thought.
Recurring themes in the literature on reconfiguration
There are some recurring themes in the literature on reconfiguration drawn from across the world. These are:
• Reproviding services is expensive – we don’t understand the economics of alternative models of care, particularly “community care”, we should build the transition and implementation costs of any reconfiguration into the estimation of costs and benefits.

• Reconfiguration doesn’t always mean fewer staff – the extent to which reconfiguration of surgical care will reduce demand / need is unknown. When combined with a move to 7 day care, it is likely to require the same or more staff.

• Bigger doesn’t mean cheaper – The available evidence suggests that the optimum size of a hospital in terms of releasing economies of scale is only 200 beds – about half the size of a small District General Hospital (DGH) – and that when a threshold of about 650 beds is reached diseconomies of scale begin.

• There is also no clear evidence to suggest that, as the range of services a hospital offers expands, the cost per unit reduces (so-called economies of scope).

• Monitor, concerned that smaller hospitals may be financially disadvantaged, have not found any correlation between trust size and financial performance.

• Reconfiguration can lead to financial failure – Finally, the reconfiguration of clinical services can present a governance and performance risk and be a contributory factor in financial failure.

Reconfiguration and the Working Together Programme
It seems accepted that status quo is not an option. It seems accepted that the further development of a managed clinical network is a definite direction of travel. Whether that would require provider level reconfiguration is not clear.
The drivers of any reconfiguration should be set out very clearly, and many have argued that the “technical” case for change is often considerably less important than winning the trust of non NHS stakeholders.


It seems clear that the key drivers to this programme are not necessarily about cost, but are about ability to provide safe surgical care services to children in South Yorkshire and surrounding areas; to maintain safe staff cover, especially out of hours and across multiple sites.


The main options for this seem to be managed clinical network with no real change to organisational infrastructure, hub and spoke mode (with or without lead provider contracting) , tartan model .
Any significant service change is likely to be resisted (by commisisoners and providers, who may choose to focus on here and now operational problems).
These are not necessarily mutually exclusive. Within each potential configuration there are a number of service level innovations, some of which are highlighted in this report
If a hub and spoke model is adopted there may be a necessity to define issues such as age and other cut offs that would define the criteria for referral to a specialist in a specialist centre. Similarly network agreement would be needed on determining for example that all elective work should be done at the specialist centre with emergency work being undertaken in a spoke; or whether this might apply to children under a certain age. Specialist advice would be needed in determining how to arrive at such cut offs.

There may be barriers to networks – such as:

• a competitive environment and commercial relationships between providers

• a lack of financial support outside the boundaries of a single provider (the trust board or the CCG is the statutory body – thus ultimately the body which will carry responsibility)

• Rigid contractual arrangements


This would need to be set against much of the available evidence that suggests managed clinical networks MUST be

• An interconnected system of providers

• Not limited by boundaries

• Multidirectional flow (not hub & spoke)

• Contractual agreements specifying service requirements and outcomes


The predilection of the current surgical and anaesthetic workforce, or provider trusts, towards any significant service change or network development is not known. Similarly the predilection of the public is not known.

The implications of any reconfiguration should be considered as broadly as possible. For example impacts on travel times and costs, family life, social services, district nursing, language support and schooling provision. It may be useful to consider these separately for inpatient care and follow-up/outpatient care.

Recommendations – networks, standards and models of care, reconfiguration
Networks of care are recommended almost universally. It is recommended that the established and developing clinical standards (RCS, RCoA, NCEPOD and other recommendations) are used, as the basis for this formal establishment of a formal managed network.
There may be resource, clinical, workforce planning, service and provider configuration issues to consider. Many aspects of a managed network can be established with no reconfiguration.
No specific recommendations are made about service configuration, as this is the point of the Working Together Programme. The two main viable options are the hub and spoke model (lead provider or current contracting framework) or the tartan model (some specialise in x, some in y).
A number of innovations around the network of care are possible. Some of them are set out here. These should be considered on their merit. For each potential innovation the key questions are:
• What are the potential benefits of these model

• What are the potential risks, limitations and trade-offs? Trade off between choice / travel time and ability to maintain safe cover.

• What incentives or rules would be needed for these models to work across local (or even regional) health economies and across different types of providers?




One comment

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s