Following research into the use of emergency beds by over-65s, Candace Imison considers what can be done to realise the vision of ‘care closer to home’.

The realisation of the vision of “care closer to home” and the accompanying shift of resources from hospital to community seems remarkably elusive.

People over the age of 65 experience more than two million emergency admissions to hospital a year, using more than 51,000 hospital beds. In many cases the admission is avoidable or the length of stay in hospital longer than necessary. 

It is estimated that more than 30 per cent of people admitted as a medical emergency do not need to be in a hospital bed.

There are places that have dramatically reduced their use of hospital beds and Torbay is perhaps the most well known example. Torbay reduced hospital beds from 750 (1999-2000) to 502 (2009-10) and now uses less emergency beds per head of population (for people aged over 65) than anywhere else in England.

At the King’s Fund we have completed some research that explores the variation in the use of emergency beds by people over the age of 65, and examines the factors that might be driving this. 

We found a fourfold variation between primary care trust areas in the use of hospital beds by people aged over 65 admitted as an emergency. Average lengths of stay vary from six days to 13 days. For patients over the age of 85, who had been admitted from home but needed to be discharged to supported accommodation, there was nearly a fivefold variation in average length of stay, from 11 days to 51 days.  

Across England, more than 7,000 hospital beds would be released if all PCTs achieved the rate of admission and average length of stay of the lowest 25th percentile. This gives the opportunity to invest £462m in community services from the resources released.

The variation in the use of hospital beds is influenced by many factors. Of all these factors, age is the strongest driver. Men and women aged under 65 use approximately one-fifth of a bed day per year, whereas men and women over 85 use more than five bed days per year, 25 times more. However, analysing the relative importance of the different drivers is complex because they are often interrelated.

There is also long-standing evidence of so-called “supply-induced” demand, whereby the use of services is driven more to their availability than to underlying demand. A facet of this is the link between access and use. There is an inverse relationship between the distance to a hospital and its use. A striking finding from our research was that all the areas with the lowest use of hospitals were rural.

We explored the variation that exists between PCT areas by investigating four groups of PCTs:

  • the 10 with the lowest bed use;
  • the 10 with the highest bed use;
  • the 10 urban PCTs with the lowest rate of bed use;
  • the 10 PCTs with the greatest reduction in their rate of bed use from 2006-07 to 2009-10.

Our research found no clear links between hospital bed use and access to community services such as GPs, community nursing and social care. For example, while the areas with the lowest use of beds had distinctively fewer single-handed practitioners and practices, the urban areas with low bed use were not that different in their GP profile to the areas with the highest bed use (figure 1). 

Graphic showing proportion of lone GPs as percentage of all practices for 4 PCT groups

Source: Al Grant


We did find that areas with higher proportions of older people had lower rates of emergency bed use (figure 2).

Graphic showing percentage of PCT population aged 65 and over for 4 PCT groups

Source: Al Grant


Our hypothesis is that in areas with a relatively high proportion of older people, more attention may have been paid to service improvements and developing more integrated models of care. Many of the PCTs with the lowest rates of emergency bed use have been leaders in the integration of health and social care.

However, when we looked at care trusts we found a mixed picture, which shows that organisational integration alone will not deliver improved performance. Overall, our analysis, along with other evidence, suggests that the key to improvement lies in changing ways of working across a system, rather than piecemeal initiatives.

So what can be done? The starting point is to understand the current position. Our supporting data tables rank all PCTs and enable each area to see how it sits compared with others. Having established the opportunity for improvement what action is needed?

The learning from Torbay suggested the following:

  • have a clear vision (across the whole health economy) based on making a positive difference to service users and be sure to keep this in sight at all times;
  • start from the bottom up by bringing together frontline teams and align these teams with general practices and their registered populations; 
  • consider how simple and inexpensive innovations such as the appointment of health and social care coordinators can make a major impact.

In an acute hospital setting there is also much that can be done to reduce length of stay and avoidable admissions. This includes:

  • ensuring there is a senior medical clinician to review patients in the emergency department;
  • use of dedicated assessment and observation wards;
  • frequent medical ward rounds;
  • providing specialist support for geriatrics, stroke, heart failure and dementia patients;
  • patients being under the care of one rather than multiple doctors.

An emergency admission to hospital exposes older people to clinical and psychological risks and can increase their dependency. Our research demonstrates a significant opportunity to reduce these risks and realise the vision of “care closer to home”.

Candace Imison is deputy director of policy at the King’s Fund.

Find out more