Health CommitteeWritten evidence from Nuffield Trust (LTC 96)


1.1 Many health systems face the challenge of rising prevalence of chronic health problems coupled with increasing longevity, resulting in more people surviving into old age with multiple chronic conditions. The UK has been no exception and over the past decade there has been a plethora of policy initiatives designed to reduce the risk of emergency admissions, including case management by community matrons,1 telephone coaching, telehealth,2 virtual wards3 or integrated care. At the national level, guidance has been developed on best practice for the treatment of common chronic conditions and financial incentives devised to improve the management of chronic disease in primary care.

1.2 A benefit of having a single payer of healthcare with universal coverage is that several years of inpatient data for the whole population are available and can be used, through time series analyses, to explore whether in combination national and local preventive care initiatives are having an impact on the rates of emergency admissions. This can be a useful approach given that new initiatives in the NHS are often grafted onto a range of older ones which are still developing, and evaluations of individual programmes might not be undertaken over long enough periods, nor take account of the interplay of different policies and initiatives.

1.3 Having a single payer also makes it possible to discern the aggregate balance of spending across different sectors of the NHS in England (ie hospital, primary and community services) from the annual accounts of strategic health authorities, primary care trusts, NHS and foundation trusts, which are consolidated into annual accounts produced by the Department of Health and Monitor. The following submission draws on findings from these data sources, as well as recent modelling work quantifying the funding pressures facing the NHS in England over the decade to 2021–22, and a two-year in-depth study of commissioning practice in three high-performing primary care trust (PCT) areas (Calderdale, Somerset and the Wirral).

The scope for varying the current mix of service responsibilities so that more people are treated outside hospital and the consequences of such service re-design for costs and effectiveness

2.1 Both the Health Committee4 and the National Audit Office5 have previously pointed out that the more appropriate use of acute hospital services is critical for the future achievement of efficiency savings. To achieve this, effective community health services are important, as is the quality of primary care. One important gauge of these services’ effect on the health system are rates of admissions for “ambulatory care sensitive conditions” (ACS)—clinical conditions for which the risk of emergency admission can be reduced by timely and effective ambulatory care, meaning mainly primary care, community and social services, and outpatient care.

2.2 A 2013 Nuffield Trust observational study of hospital admission data spanning the period between March 2001 to April 2011 revealed that the number of emergency admissions per year for ACS conditions had increased over the decade by 40%, rising from 701 995 admissions per year to 982 482 annually.6 Emergency admissions for ACS conditions were more common among the oldest and youngest age groups When the age-standardised rates of emergency admissions for ACS conditions were compared, the overall increase was 25% indicating that not all of the decade long rise in admission rates was due to the changing demographic structure of the population. The overall trend also masked differences between the three broad categories of ACS conditions (acute, chronic and vaccine preventable). For acute ACS conditions the age standardised rate of increase was 44%, while vaccine preventable rates (such as for influenza) went up by 136%, albeit from a much smaller base.

2.3 The rate of admissions for chronic ACS conditions held roughly steady. Among this group, the conditions with the highest rises in the absolute numbers of admissions were chronic obstructive pulmonary disease (COPD) and convulsions and epilepsy. The number of admissions for diabetes grew by 95% and contributed an extra 16 996 admissions. Meanwhile age-standardised rates of admissions for congestive heart failure and angina showed marked and significant reductions (-27% and -41% respectively).

2.4 There are a number of possible explanations for the observed increases in the chronic group of ACS admissions. For some ACS conditions changes could be due to the differences in the underlying prevalence of disease; changes in health-related behaviours such as smoking or improvements in the effectiveness of treatments for some diseases for example, statins for angina. Other possible explanations include changes in the way health systems operate, for example changes in the thresholds for admission to hospital. It may be the case that patients admitted now are less sick than they were 10 years ago, as suggested by the rise in short stay admissions and lower rates of mortality shown in an earlier Nuffield Trust study.7 Or that admission decisions are in part influenced by the perceived lack of alternatives to inpatient care (for example an absence of social care or community based alternatives to hospital).

2.5 Some caution is required as the evidence connecting the rising trend in ACS admissions to specific policies is weak. For example the introduction of the 4 hour A&E target and changes to the GP out of hours contract (both in 2004) are commonly cited as increasing the overall level of emergency admissions, yet our study suggests that emergency admissions continued on an underlying upward trend at this point, with no obvious acceleration in the rates of ACS or non-ACS emergency admissions. Indeed some of the more common ACS conditions show trends that closely mirror the rise in rates of emergency admission for non-ACS conditions. This suggests that the organisation and financing of the health system itself are perhaps more important determinants than the changing health needs of the population over the decade—specifically the expansion of acute sector capacity from 2002. We know from the consolidated NHS accounts that community health services have experienced particularly rapid growth in spending since 2006–07. Acute spending also continues to increase but spending on primary care, and in particular GP services, has been falling in recent years in real terms.8

2.6 Thought needs to be given to how national and local policies are impacting on the pattern of admission for patients, either because of suboptimal preventive ambulatory care, or changes in thresholds for admission to hospital. It is clear that there is a role for much more comprehensive, independent and transparent analysis of policies and local initiatives that might make an impact on quality and costs. This work is beginning and could be modelled in line with emerging evidence from research findings, for example the impact of: the increased funding of community services since 2006/07; service reconfiguration; integrated care;9 workforce skill mix; changes to social care funding; and greater use of telehealth and telecare.10

The readiness of local NHS and social care services to treat patients with long-term conditions (including multiple conditions) within the community

3.1 In its June 2011 summary report the NHS Future Forum called for the commissioning of integrated care for patients with long-term conditions, complex needs and at the end of life. It was an important acknowledgement that the ageing population and increased prevalence of chronic diseases requires a decisive shift away from the current concentration of resources in acute hospital settings towards greater preventive care, self-care, more consistent standards of primary care and community based health and social care services that are also well coordinated.11 The Department of Health subsequently approached The King’s Fund and the Nuffield Trust for help in supporting the development of its national strategy on integrated care. In our resulting report we identified a set of organisational and policy barriers that need to be addressed.12

3.2 They included:

The institutional divisions between primary and secondary care in the NHS and between health and social care that often make it difficult for multidisciplinary teamwork to happen.

The absence of a shared electronic patient record enabling care that can cross organisational boundaries.

NHS regulation that focuses too much on the performance of organisations and not enough on performance across organisations and systems.

3.3 Some local areas have attempted to overcome these and other obstacles. For example the North West London Integrated Care Pilot (ICP) is a large scale innovative programme designed to improve the coordination of care for people over 75 years of age and adults living with diabetes.13 Multidisciplinary groups of local care providers segment patients according to risk; share care plans across care settings; identify patients needing intensive case management, and monitor care plan implementation. Sophisticated governance structures have been put in place, with all staff involved in the pilot, together with representatives of local patient organisations and providers coming together under an Integrated Management Board (IMB). Financial incentives have also been designed to support the aims of the pilot in relation to care planning, collaborative care management, and the development of innovative alternatives to hospital admission.

3.4 Likewise NHS organisations in Trafford, a borough of 215,000 people in Greater Manchester have been trying to develop closer collaboration between community based primary, general acute medicine, specialist outpatient and diagnostic care to enable more care to be delivered outside hospital. Initial implementation of the programme involved nine “vanguard” general practices working with community, acute and social care to redesign selected care pathways, share data, identify patients at risk of unplanned hospitalisation, and generally to act as a test bed for implementing and evaluating integrated care.14

3.5 Neither area is unique in attempting new ways to deliver care closer to the patient. However in terms of scope and ambition they are certainly not the norm. It is perhaps instructive that in both cases there were substantial initial prompts; in the case of Trafford, it was a an ambitious PCT and groups of clinicians that funded and drove the need for change, while in North West London the set up phase was helped by financial and other support from NHS London; enabling investment in IT, pilot leadership, coordination of multidisciplinary groups and project management. While recent Government announcements pledging to align policy initiatives in favour of more integrated care for a set of “pioneers” are welcome, it is also likely that other local examples of innovation will emerge as a result of the current spending squeeze and it is important that these are also supported and evaluated.

3.6 A key determinant will involve the extent to which commissioners are supported to balance the possible tensions that can arise when policies in support of competition and choice (and the associated procurement laws) interact with the planned shift to greater coordination of community based services. These ought not to be mutually exclusive policies but in practice there is a risk that the scope for regulatory action is perceived by commissioners as potentially being so wide and far reaching that it inhibits service redesign and innovation. Second, those commissioners wanting to redesign services appear to be faced with a high burden of proof. Emerging Competition and Cooperation panel guidance suggests that commissioners and providers will have to show that their actions produce benefits that are “significant, quantifiable and evidence based” to set against the theoretical costs of any reductions in competition and patient choice.15 This is a high threshold of evidence to meet, particularly when the effects of large scale integrated care projects can take several years to emerge fully.

The practical assistance offered to commissioners to support the design of services which promote community-based care and provide for the integration of health and social care in the management of long-term conditions

4.1 Earlier this year the Nuffield Trust published a major National Institute of Health Research-funded study exploring what commissioners actually do when commissioning care for people with long-term conditions, and how this might be improved.16 The research was based on 15 months of detailed observation from November 2010 to January 2012 in three commissioning communities: Calderdale, Somerset and the Wirral. These sites were selected because on various indicators they appeared to be at the forefront of commissioning practice. The research focused on the commissioning of care for people with long-term conditions: diabetes in all three sites (to allow comparison), and a second condition chosen by each primary care trust: dementia in Calderdale and the Wirral, and stroke in Somerset.

4.2 A huge amount of effort was observed going into commissioning across all six service developments, in terms of strategic planning, service review, and hands-on work (typically by middle managers within primary care trusts) to implement and support change. The research further revealed that the commissioning was not seen as a two-sided transaction across the “purchaser-provider split”, at least where long-term conditions are concerned. Local commissioners were instead developing an alternative approach based on collaborative working between providers and commissioners. While there are clear benefits to this, in terms of clinicians driving change, there are also risks; including potential conflicts of interest, for example providers using senior clinicians to help ward off challenges from commissioners, or providers being actually involved in design of services for which they subsequently tender.

4.3 The study concluded that commissioning is a labor-intensive activity, often concentrated on relatively marginal rather than mainstream service developments. It was also noted to be much more relational than transactional in nature, with commissioners appearing reluctant to engage in challenge of providers and decommissioning, preferring to focus on service planning and design, involvement and consultation with local stakeholders, and implementation of new developments. In the new commissioning system, given recent management cost reductions of some 40%, clinical commissioning groups will have to be very selective about how they allocate their time and effort. Commissioners’ traditional role of as convenors of the local health system may also have to adapt, or at the very least be scaled back—they will need to be brave enough to “cut and run” and make difficult decisions when they feel they have undertaken enough consultation and engagement, even if this flies in the face of NHS (provider and commissioner) culture that favours extensive involvement and consultation as a way of reviewing and making changes to services.

4.4 The implication of this is that done well, commissioning can be a lonely role, especially when it involves the “tough work” of changing services—work that is likely at times to threaten professional and provider interests, or prove unpopular with some in the wider community. Commissioners will need intelligence from commissioning support units to challenge providers on quality and value for money, and where necessary, use procurement to let contracts for services for their populations. This should include high-quality public health and needs assessment advice, sophisticated and real-time data about services, accurate comparisons with national benchmarks, efficient payment and invoicing systems, and support for modelling and planning future care. In addition, commissioners will need support in undertaking public consultation, accessing and analysing patient and public experience data, providing local system leadership, and handling procurement within a cultural context of collaboration.

The implications of an ageing population for the prevalence and type of long term conditions, together with evidence about the extent to which existing services will have the capacity to meet future demand

5.1 Population growth and demographic change are fundamental drivers of future health care activity. Clearly, as a population grows, demands for health care will rise. Additionally, a population with a higher proportion of older people will have a greater need for health care. While health service utilisation tends to increase with age, there is additional use of health services as people approach the end of their life, regardless of age. This increase is less pronounced for people aged over 85 years, but is still evident. Use of health services is thus dependent not just on ageing, but also on the number of people who are expected to die at any particular time.

5.2 Our December 2012 report mapping the medium term NHS productivity challenge used the principal ONS projections from 2008 for population, mortality and fertility to create a projection to 2021–22.17 The population is projected to grow for all age groups of 50 years and above, for both males and females. The probability of having an inpatient admission related to a chronic condition is higher for older age groups than younger age groups. Admissions linked to chronic conditions are therefore likely to rise as the population ages. However, recent trends suggest that this probability is also rising within age bands. If this trend continues, the impact on the NHS due to chronic conditions will amplify the effect of population growth alone.

5.3 For primary care, we examined trends in the number of times a year individuals visit their GP. The number of consultations at a GP practice per person per year rose from 3.9 in 1995–96 to 5.5 in 2008–09. Over the same period, the proportion of these consultations occurring with a GP, rather than a practice nurse, fell from 75% in 1995–96 to 62% in 2008–09. We have assumed that both of these trends will continue over the period 2010–11 to 2021–22, so that by 2021–22 there will be 7.8 consultations per person per year; 55% of which will be with a GP. We applied a similar technique to the mental health model to predict how many people would require mental health services each year, based on the number of people using services between 2008–09 and 2010–11. This projection would see the proportion of people aged over 65 years using mental health services growing by just over 50% by 2021–22.

5.4 Further protection of the NHS budget is not guaranteed due to the uncertainty of the economic outlook for the country. Our analysis concluded that without unprecedented, sustained increases in health service productivity, including more effective management of chronic conditions, funding for the NHS in England will need to increase in real terms between 2015–16 and 2021–22 to avoid cuts to the service or a fall in quality. Yet with no clear signs of economic recovery, it is conceivable that NHS funding may be frozen for further years. If this is the case, a major rethink of how health services in England are funded and organised will be needed. It is crucial, therefore, that the NHS becomes much better at preventing and managing long-term conditions.

24 May 2013

1 Steventon A, Bardsley M, Billings J et al (2011). An evaluation of the impact of community based interventions on hospital use. Nuffield Trust

2 Steventon A and Bardsley Steventon A and Bardsley M (2012). The impact of telehealth on use of hospital care and mortality. Nuffield Trust

3 Lewis, G, Bardsley M, Vaithanathan R et al. “Do ‘virtual wards’ reduce rates of unplanned hospital admissions and at what cost?” A research protocol using propensity matched controls. International Journal of Integrated Care. Volume 11, 30 June 2011

4 Health Select Committee (2012). Public Expenditure: Thirteenth report of session 2010–12. The Stationery Office.

5 National Audit Office (2012). Progress in Making NHS Efficiency Savings Report by the Comptroller and Auditor General, HC 686, Session 2012–13. The Stationery Office

6 Bardsley M, Blunt I, Davies S et al. “Is secondary preventive care improving? Observational study of 10 year trends in emergency admissions for condition amendable to ambulatory care”. BMJ Open 2013;3:e002007.doi:10.1136/bmjopen-2012-002007

7 Blunt I, Bardsley M and Dixon J (2010). Trends in emergency admissions in England 2004–09. Nuffield Trust. London

8 Jones N and Charlesworth A (2013). The Anatomy of health spending 2011–12: A review of NHS expenditure and labour productivity. Nuffield Trust

9 Ling T, Bardsley M, Adams J. “Evaluation of UK Integrated Care Pilots: research protocol”. International Journal of Integrated Care. Vol 10, September 2010

10 Steventon A and Bardsley M (2012). The impact of telehealth on use of hospital care and mortality. Nuffield Trust

11 Ham C, Smith J, Eastmure E (2011). Commissioning integrated care in a liberated NHS. Nuffield Trust, London

12 Goodwin N, Smith J, Davies A et al (2012). Integrated care for patients and populations: Improving outcomes by working together. King’s Fund & Nuffield Trust

13 Nuffield Trust & Imperial College London (2013). Evaluation of the first year of the Inner North West London Integrated Care Pilot. Nuffield Trust, London

14 Shaw S and Levenson R (2011). Towards integrated care in Trafford. Nuffield Trust, London

15 Nuffield Trust (2013). Parliamentary Briefing: NHS procurement, patient choice and competition (Response to draft regulations). London

16 Smith J, Porter A, Shaw S et al (2013). Commissioning high quality care for people with long term conditions. Nuffield Trust; Shaw S, Smith J, Porter A, Rosen R, Mays N (2013) The work of commissioning: a multisite case study of healthcare commissioning in England’s NHS. BMJ Open, available at

17 Roberts A, Marshall L, Charlesworth A (2012). A Decade of Austerity? The Funding Pressures facing the NHS from 2010–11 to 2021–22. Nuffield Trust

Prepared 3rd July 2014