Preventing admission of older people to hospital

BMJ 2013; 346 doi: http://dx.doi.org/10.1136/bmj.f3186 (Published 20 May 2013) Cite this as: BMJ 2013;346:f3186
  1. Shaun D’Souza, specialist registrar, medicine for older people,
  2. Sunku Guptha, consultant physician, medicine for older people
  1. 1Peterborough and Stamford Hospitals NHS Foundation Trust, Peterborough PE3 9GZ, UK
  1. shaun.dsouza{at}pbh-tr.nhs.uk

No evidence that managing “frail older” people in the community reduces admissions

The hypothesis that better community facilities for older people (over 65 years) will reduce demand for inpatient care is appealing and has gained wide acceptance. In England, many commissioning groups plan to redirect a substantial proportion of resources into services that will enable medical care for older people in the community. Secondary care trusts will need to take these funding changes into account when planning their services, and some trusts have already reduced their bed numbers and healthcare staff.

In July 2012, the then secretary of state for health in England, Andrew Lansley, first admitted that the shift of focus towards community healthcare, as outlined by the Health and Social Care Bill, would lead to a reduction in inpatient beds. Plans to reduce beds in medical wards, and in units dedicated to older people, rest on the assumption that longer term conditions can be safely treated in the community. Reducing emergency hospital admissions remains a key target, not only because of the high costs of emergency admission compared with other forms of care, but also because emergency admissions disrupt elective healthcare.1 But will better community services for older people really reduce acute admissions to hospital?

Recent substantial changes to the NHS have gone ahead despite a lack of evidence that any model of community care can greatly reduce hospital admissions in frail older people. In 2010 a paper commissioned by the think tank the King’s Fund, on avoiding hospital admissions in all age groups, found no evidence that case management, in the community or in hospital, reduces generic admissions.2 Community matron or case worker models using experienced nurses targeting naive and known users of primary or secondary care in those over 65 years has had no impact on rates of admission.3 4

It is hoped that advances in technology, and the removal of artificial barriers between hospital and community care, will offset any reduction in inpatient capacity. The formation of integrated care organisations is intended to help break down the barriers. These organisations emphasise that most people would prefer to have medical care at home, and that frail older people can be safely treated without the need for hospital admission. However, trials using integrated teams in various forms have so far failed to reduce numbers of admissions compared with usual care. One trial targeted low income older patients in general, but specifically those at risk of using secondary care resources. It used a community multidisciplinary team model comprising primary care doctors, geriatricians, community nurses, pharmacists, physiotherapists, and social workers. It also involved extensive assessments and detailed comprehensive management plans tailored to individual patients, weekly meetings to ensure implementation, at least once monthly patient contact by a team member, and annual review of protocols by the whole team. Despite being so thorough and resource intensive, the trial had no impact on admissions to secondary care compared with usual care.5

Another randomised trial actively screened community dwelling older people (irrespective of their contact with primary or secondary care) for conditions such as depression, falls, urinary incontinence, and cognitive and functional impairment (the so called geriatric giants). The researchers then intervened intensively using specialist services that included geriatric medicine and psychiatry, urology, audiology, rehabilitation, psychology, and social services. However, they found no reduction in admissions compared with the usual care group over a three year follow-up.6

Models of extending secondary care into the community have also been unable to show a reduction in admissions compared with usual care. One trial investigated a hospital based team consisting of a geriatrician, trained nurses, and social workers that offered outreach in the community. Despite active intervention, extensive assessments, and round the clock support during follow-up, admissions were not reduced compared with usual primary care.7 In another trial of older people admitted to hospital by a geriatric team, intervention during inpatient stay and extension of follow-up after discharge did not reduce admissions compared with usual care.8

Furthermore, no convincing evidence exists that increases in the provision of community services reduces length of stay for frail older people. A Cochrane review of “hospital at home” services was unable to pool data on length of stay for older patients admitted with a mix of medical conditions because of substantial heterogeneity.9 There is also no clear association between investment in social care and hospital bed use among older people.10

All trials examining community care used the comprehensive geriatric assessment tool for screening and intervention, either in its complete or modified form. This tool has been shown to improve quality and reduce mortality in older people admitted to secondary care, but it has not been shown to reduce readmissions, including when instigated at an early stage in acute medical units or emergency departments.11 12 Currently, no validated tool has been proved to be useful in reducing admissions or readmissions in community dwelling frail older people.

Given this lack of evidence, it is difficult to understand the widespread belief that admissions to hospital can be reduced by managing frail older people more effectively in the community. Frailty is a dynamic condition that is characteristically associated with sudden profound decompensation secondary to a stressor.13 After such events, patients require dedicated nursing because haemodynamic instability, hyperactive or hypoactive delirium, and falls are common. There is no evidence yet that proactive management in the community can reduce the occurrence of such episodes, most of which result in hospital admissions that would be hard to categorise as avoidable.

We urgently need aids that can more efficiently detect and measure the severity of frailty in community dwelling older people. These would allow research into interventions that could reduce episodic decompensation in older patients, and the development of integrated care models that are appropriately directed. For some conditions, such as stroke and hip fractures, we already know that models that effectively integrate secondary care and community services can reduce length of stay and improve functional outcomes.9

There are some known benefits of good community care services that speak for additional investment in these services. Systematic reviews of home based interventions, despite being complicated by methodological variations and a lack of standardisation of interventions, show that community based patient centred care delivered comprehensively in a sustained fashion with multiple visits reduces long term institutional care.14 15 High levels of patient satisfaction3 4 5 6 7 8 and improvements in self reported health and mental health were found.16

However, there is no evidence that enhancing community care for frail older people will reduce hospital admissions, and demands on secondary care will probably continue to rise. There has been a sustained reduction in the number of acute beds over the past few decades, and most hospitals now average around 90% bed occupancy.17 A further reduction in beds based on the vain hope that enhancing community services will reduce admissions could be potentially dangerous to patient care. It would be more sensible to evaluate the effects of enhancing community services before making decisions to cut more acute care beds.


Cite this as: BMJ 2013;346:f3186


  • Competing interests: We have read and understood the BMJ Group policy on declaration of interests and declare the following interests: None.

  • Provenance and peer review: Not commissioned; externally peer reviewed.