Hospital service reconfiguration: the battle for hearts and mindsBMJ 2012; 344 doi: http://dx.doi.org/10.1136/bmj.e953 (Published 16 February 2012) Cite this as: BMJ 2012;344:e953
- Helen Barratt, research training fellow,
- Rosalind Raine, professor of health care evaluation
- Correspondence to: H Barratt
Proposals to reconfigure NHS hospital services are always contentious. In 2001, Dr Richard Taylor won a parliamentary seat on the strength of his campaign against the decision to close acute services at Kidderminster Hospital.1 More recently, plans to centralise surgical services for children with congenital heart disease have prompted opposition across the country.2 The Independent Reconfiguration Panel (IRP) advises ministers about re-shaping hospital services and, by the end of 2010 it had undertaken 17 full reviews of contested proposals for health service change in England and offered written advice on several others.3
During the 2010 general election campaign, the Conservative party promised to stop “centrally dictated” reorganisations of NHS services.4 Writing in the Daily Telegraph on 21 May 2010,5 days after the coalition government came to power, the Secretary of State for Health, Andrew Lansley, declared a moratorium on such programmes. All proposals—even those at the implementation stage—were put on hold and planners were instructed to demonstrate that plans met four new tests: support from general practitioner (GP) commissioners, strengthened public engagement, consistency with patient choice, and clear evidence for change.
One high profile reconfiguration affected by the moratorium was the proposed downgrade of acute services at Chase Farm Hospital in north London (box). We provided academic input to the review of these proposals. In this article, we draw on this experience to reflect on the difficulties of applying the Lansley tests and propose ways to strengthen the evidence base to aid decision making.
Chase Farm reconfiguration
Proposals to downgrade services at Chase Farm Hospital in north London were first raised more than 15 years ago.6 They included replacement of the emergency department with an urgent care centre and consolidation of women’s and children’s services at Barnet and North Middlesex Hospitals, which are six to seven miles away. Other places will be grappling with similar decisions.7
In July 2009 work began on implementing the proposals, including building work on the North Middlesex site to accommodate increased patient flows. However, implementation was put on hold while the proposals were assessed against the four new Lansley tests.
May 2010: A panel of clinicians (mostly local GPs) was convened to review the clinical evidence underpinning the plans. The panel concluded that no change was not a possibility because the current situation was both “unstable and unsustainable” and would result in declining quality of care and worsening health inequalities.8 All local GPs were invited to comment on the panel’s conclusions. Patient and public engagement was sought through a series of public meetings and a supplement in local newspapers explaining the rationale for the proposals. Local involvement network representatives were involved in assessing the strategy against the patient choice test
January 2011: The strategic health authority, NHS London, concluded that the four tests had been met and that the reconfiguration could recommence
March 2011: The health secretary held a closed meeting with local MPs and Enfield council representatives, who were opposed to the proposals. Afterwards, Mr Lansley invited this group to submit alternative options to maintain services at Chase Farm
May 2011: The secretary of state referred Enfield Council’s report to the IRP. The report did not include an alternative plan, but requested more investment in healthcare for the borough9
July 2011: The IRP concluded that the council’s submission did not present a credible alternative to the current proposals and that the “status quo has real downside risk in terms of the current safety and sustainability of local services”9
September 2011: Andrew Lansley announced that he accepted the IRP’s recommendations and that the proposals to downgrade services at Chase Farm could go ahead10
Clarity on the clinical evidence base
The research evidence used to justify reconfiguration focuses on volume-outcome relations. A growing body of research shows improved patient outcomes when a range of procedures— including surgery for colorectal cancer 11 and elective repair of aortic aneurysm12—are carried out in larger units that serve bigger populations. However much of this research has been done in the US, where organisational differences may limit transferring the findings to the NHS.13 Furthermore, many of the studies are poorly controlled for the effects of confounding variables.
In contrast, an English study of retrospective data from ambulance call-outs for immediately life threatening symptoms found that increased journey distance to hospital was associated with increased mortality.14 After case-mix was adjusted for, every additional 10 km in straight line distance was associated with a 1% absolute increase in mortality. These findings reflect the performance of emergency services between 1997 and 2001, and journey time may be less relevant now that paramedics commonly start definitive treatment.15 In addition, the results may apply only to patients at high risk of immediate death. Indeed, preliminary data suggest that centralising specialist hyperacute stroke units across London is associated with reduced mortality compared with national rates.16 This is likely to be because patients receive rapid access to thrombolysis. Findings such as these highlight the need for a thorough understanding of the interaction between geographical access and patient outcomes.
These data are largely drawn from observational research. Although such studies have limited power to demonstrate causality, experimental designs of major service reorganisations are usually not feasible.17 In these circumstances, the case for change often relies on a combination of observational research evidence and expert clinical consensus.
Consistency with patient choice
This test is defined by the Department of Health as the extent to which proposals affect patients’ ability to choose between providers, settings, and interventions.18 However, research suggests that choice of provider is not, in fact, a priority for patients having elective care; respondents placed it as the 11th most important aspect of their healthcare in a list of 16 items.19 20 It may be even less important for patients needing emergency care.21
Moreover, the germane question is whether the plans are consistent with patient choice when people are presented with a trade-off between the perceived advantages and disadvantages of the current service compared with reconfigured services. Such trade-offs have not been formally examined but are likely to be influenced by community loyalty to local hospitals. Hospitals have an important social role, helping the public to maintain trust in the NHS.22 Mixed methods (qualitative-quantitative) research exploring determinants of public opinion and the importance of the various components of the trade-offs (for example, proximity versus improved survival) would help us gain a better understanding of public and patient priorities. This could in turn inform future service reconfigurations.
Strengthened public engagement
Meaningful engagement requires that the public should be able to affect decisions.23 Intensive stakeholder engagement and clear messages about the need for change help build a legitimate case for reform.24 In reality, commissioners have been criticised for consulting on service redesign after decisions have been made.3 To meet this test, commissioners are recommended to seek the views of local involvement networks and health overview and scrutiny committees. Beyond this, however, strengthened patient engagement remains challenging to quantify.
The previous government introduced several reforms aimed at increasing public involvement and local scrutiny of healthcare, and the IRP, established in 2003, was part of this.25 The panel provides independent advice on reconfiguration proposals when local agreement cannot be reached, and its members include clinicians, management representatives, and lay members with experience in delivering health service change. One of the main reasons why reconfiguration proposals are referred to the panel is that commissioners have failed to convince affected communities of the clinical case for change.3 However, there is little research or consensus on the best methods to secure public engagement.26
In terms of the economic rationale for change, the public are often deeply suspicious that reforms are a cover for spending cuts.27 This may be a particular concern now that the NHS is under pressure to make efficiency savings.28 In reality, the economic case for reconfiguration is seldom clear cut because it depends on the costs that are included (such as transaction costs and, where relevant, repayments on private finance initiatives). In addition, long time frames, uncertainty about the future, and changing conditions make it difficult to predict future costs. Even if the changes are likely to save money and not harm health outcomes, public acceptability may be tempered by other important determinants of patient and carer experience such as travel times and costs.
Effective public and patient engagement requires explicit presentations of the clinical and financial risks, benefits, and implications of service change. The way that such information is framed is also important because it affects the way that it is interpreted.29 The media sometimes exploit this by using sensationalist language to raise fears about large numbers of lives being put at risk, often with little or no supporting evidence.30 As part of strengthening engagement, commissioners and providers must therefore disseminate transparent, comprehensive information in a form that can be understood by all sections of the affected community.
Support from GP commissioners
The fourth test reflects the coalition government’s commitment to devolve decision making power to general practitioners. At the time of the Chase Farm review, commissioning groups were not yet established in the three main affected boroughs, so this test was assessed by inviting all local general practitioners to indicate whether they agreed with the recommendations. Response rates and percentage levels of support were then described for each borough.31 However, in south Hertfordshire, which is also affected by the Chase Farm reconfiguration, the two general practice consortiums were asked to submit a written indication of their corporate views about the proposals, as suggested in the Department of Health guidance.
Since the Chase Farm decision commissioning groups have been broadened to include other health professionals, and the government will need to think again about how best to obtain the views of local GPs. A dichotomous indication of GP support is relatively simple to obtain. However, this gives no indication of the strength of their views. Such data could be obtained by asking GPs to provide more in depth information and respond to questions using a Likert scale. This approach is obviously more resource intensive and requires more sophisticated analytical techniques.
Additional drivers for change
Reconfiguration schemes are commonly driven by additional factors that are not scrutinised by the Lansley tests. These should also be taken into account when evaluating the rationale for change. First is the shift towards greater provision of services in community settings because of factors such as increasing day case surgery and a higher burden of chronic disease as the population ages.28
Workforce related factors that may affect patient safety are also important. The implementation of the European Working Time Directive tends to drive reconfiguration proposals in acute services with high emergency workloads.32 It is argued that consolidation of acute services is required to ensure both a critical mass of junior doctors to maintain adequate standards of patient care and sufficient numbers of patients for satisfactory clinical training.33
Several royal colleges are also lobbying for higher levels of senior staffing to promote safer practice.34 35 They argue that achieving this without service reorganisation would require a substantial increase in consultant numbers. Once again, it is not possible to apply experimental methods to prove such assertions. We therefore have to rely on observational data, and here there is some evidence of worse outcomes in patients who are admitted to hospital during evenings and weekends when fewer consultants are available.36
The Department of Health guidance does not set thresholds for meeting the four new tests, arguing instead that the process should be locally led and designed.18 Furthermore, many of the requisite data are qualitative and require value judgments about their importance, relevance, and representativeness.
Research evidence will always comprise but one determinant of reconfiguration decisions, and there will be a trade-off with other factors, including local and national political concerns. Although organisations including the British Medical Association, the NHS Confederation, and the King’s Fund have called for the IRP to become the final arbiter, thus distancing politicians from decisions,37 it is naive to expect politicians not to support their constituents’ concerns, even in the face of clinical consensus, particularly in marginal constituencies.
The relevance of the research evidence is also likely to be contested by stakeholders with different perspectives and values. It is therefore crucial to pay close attention to the transparency, comprehensibility, and comprehensiveness of the evidence and to ensure it meets the needs of all decision makers
Furthermore, there is room to strengthen the evidence base through a programme of national research that is generalisable to local circumstances. For example, a better understanding of the relation between geographical access and patient outcomes would be valuable. We would also benefit from a better understanding of what concerns the public and the trade-offs that patients and their families are prepared to make when considering major service change.
Another problem is that the four tests could produce conflicting outcomes. For example, the evidence may suggest clinical benefits from reconfiguring services on fewer sites. But this could reduce patient choice of provider. It may be more acceptable and transparent to define the relevant components of benefit and patient choice and their order of priority with local stakeholders before plans for reconfiguration are made.
Finally, changes set out in the NHS Bill could raise additional challenges. The extent of involvement of multidisciplinary clinical commissioning groups, clinical networks, and senates, as well as the potential role of the National Commissioning Board, is currently unclear and will need to be considered. Without regional structures to guide the process, it is unlikely that single commissioning groups will have the resources to conduct the necessary analyses required to assess the case for change or wield sufficient power to initiate reform across large geographical areas. Decisions concerning service integrations involving multiple hospitals may also be influenced by the promotion of competition between trusts.
Cite this as: BMJ 2012;344:e953
We thank Jonathan Wolff and Steve Morris for their advice and comments.
Contributors and sources: HB is a speciality registrar in public health and RR is UCL Partners programme director in population health. The article is based on a review of the evidence underpinning the consolidation of acute hospital services and the authors’ experience of providing independent academic input into the review of the Barnet, Enfield and Haringey clinical strategy. The views expressed are those of the authors. Both authors drafted the article. HB is guarantor.
Competing interests: Both authors have completed the ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare no support from any organisation for the submitted work and no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; they provided academic input to the review of the Chase Farm decision.
Provenance and peer review: Not commissioned; externally peer reviewed.