This site is intended for health professionals only

At the heart of general practice since 1960

Read the latest issue online

CAMHS won't see you now

Options when a partner leaves

The Government's answer to spiralling emergency admissions, in the form of US-style case management models pioneered by Evercare, was reported as falling at the first hurdle after an interim report concluded the pilots would make only a dent in admissions. But were such reports justified? The CEO of UnitedHealth Care, which runs Evercare, and the leader of the evaluation team debate the issues.

Dear Martin,

I'm pleased your centre is evaluating case management of the frail elderly, but I worry that your interim evaluation has been widely reported as saying 'Case management doesn't work'. You probably share my worry. I'm particularly worried because your study was mainly of Evercare, the programme run by UnitedHealth Europe, of which I'm the chief executive.

There are at least three problems with the phrase 'case management doesn't work.'

First, case management covers many different processes, ranging from case managers who have only telephone contact with patients through to highly-trained nurses who have regular clinical contact with patients.

Second, there's the problem of what constitutes 'working'. I think you'll agree there is abundant evidence that patients and their carers feel great benefit from case management: they have a trusted friend to guide them through a complex system. The debate is more whether case management will reduce unplanned admissions, a target set for all primary care trusts. Our evidence from the US is that Evercare will, but that it takes more than a year to have that impact.

Third, there is ­ and I'm sure you'll agree ­ a traditional confusion between 'absence of evidence' and 'evidence of ineffectiveness'. We were often accused of confusing the two when I was editor of the BMJ.

We certainly don't have evidence that case management doesn't work, and more evidence suggests that it does ­ even in reducing unplanned hospital admissions.

I hope you might agree that a more interesting question than 'Does case management work?' is 'Can case management lead to a reduction in unplanned hospital admissions?'.

Do you?

Best wishes


Dear Richard,

The conclusion 'Case management doesn't work' wasn't based on our evaluation of Evercare itself, but on an analysis of existing hospital admission records.

What this showed was that existing approaches to case management, including Evercare and the new community matron policy, are simply not targeting groups at high enough risk of re-admission. So the hoped-for reductions in admissions are, in our view, unlikely to materialise.

We also pointed out that Evercare as it's being implemented in the UK is very unlike the Evercare model that led to a major reduction in admissions in the US, so we probably won't see similar benefits.

So, Evercare is unlikely to 'work' in terms of reduced admissions.

Which is why we suggested that GPs and PCTs should be looking to other benefits of the new policies. We entirely agree that it's important to see what these will be.


Dear Martin,

I still think that case management can reduce hospital admissions. This is what we need to do in order to achieve reductions in hospital admissions.

1 Recognise that programmes take some time ­ perhaps around 18 months ­ to achieve their effect. Patients in England were in the Evercare programme for an average of only eight months.

2 Improve the way we identify those at highest risk, using data not just on hospital admissions but also on clinical condition, drugs, social conditions, use of other services, and trends in use of services. We have such a tool already.

3 Introduce through commissioning many more intermediate care services ­ for example, step up beds, rapid response teams, diagnostic centres, and teams to give intravenous antibiotics at home. This would avoid the stark choice between home, where the patient may live alone in deprived circumstances, and an acute hospital.

4 Recognise that most admissions occur out of hours ­ and so provide nurse services round the clock seven days a week or provide a strong link with out-of-hours services.

Evidence from the US ­ not only on patients in residential homes but also on patients in the community ­ shows that it is possible to reduce hospital admissions with these kinds of measures. Don't you think that it could work here as well?

Best wishes


Dear Richard,

I think you are right. But you must be a very frustrated man, since few of those things, for example more intermediate care services and round-the-clock nursing, are generally available to back up your Evercare nurses in the UK. Is there anything you can do about this?

I'm interested that you have a tool to identify those at highest risk of admissions. As you know, we think the tools that community matrons are going to use will fail to identify those at greatest risk of admission.

Would you be willing to share your tool with the NHS?

Best wishes,


Dear Martin,

I'm not frustrated ­ at least not about the NHS. I never expect everything to be fine at once. Improvement is often slow and difficult with steps forward as well as backwards. That's why evaluation is so important. I know here that I'm preaching to the converted.

PCTs, as you know, have to reduce unplanned hospital bed days. Some of them hoped that case management alone would help them to reach their target, but we now realise that more is needed. Case management is necessary but not sufficient.

Many trusts are already developing intermediate care services and the like to reduce unplanned admissions.

The result should be a shift of resources from hospitals to the community, something that will benefit patients and which I'm sure you will applaud.

We are already working with some trusts and strategic health authorities on risk stratification and will be happy to work with more. The model is a function of age, sex, burden of illness (which includes several measures), previous use of services, and trend.

The model has to be calibrated for each community and depends on being able to access information not just from hospitals (which is comparatively straightforward) but also from the community.

It would be interesting if somebody ­ perhaps you or you and us together ­ were to calculate the value of getting as many people as possible in the right place on the famous risk triangle. It must surely be high.

Best wishes


Dear Richard,

I'm glad you agree that more than case management will be needed to meet the admission targets. Our research has been important in drawing this to people's attention.

Unfortunately, PCTs' expectations have been raised by some pretty extravagant claims ­ some of them, I have to say, from UnitedHealth.

I also think its disappointing that you are not willing to share your tools more freely with the NHS (ie unless PCTs contract with you).

The NHS has invested heavily in UnitedHealth, and I think it would be good if, when the NHS contracts with the private sector in future, this allowed for wider sharing of intellectual property.

Best wishes,


Behind the Evercare headlines

·Evercare is a US model that identifies elderly people at risk of emergency admission and then builds a care plan around their needs and responds promptly to deteriorations in health. The advanced nurse practitioner 'community matron' is at the heart of care delivery.

·While the Government is reported to have spent £3.4 million to run the nine UK pilot Evercare schemes and is enthusiastic about the results Evercare can achieve (50 per cent reduction in US admissions ), a Government-commissioned study led by Professor Roland said the evidence to date left many questions over the effectiveness of these schemes in the UK.

·Despite the equivocal results, Health Secretary John Reid has announced plans for a nationwide roll-out with 3,000 community matrons by 2007.

Many others do not now share ministers' enthusiasm for the pilots:

The RCGP said: 'This vindicates our long-running concerns over community matrons and what seems to be a growing trend towards fragmentation.'

The BMJ said: 'Plans for community matrons to help keep older people out of hospital may be based on misleading data.'

GPC chair Dr Hamish Meldrum said: 'Matrons are unlikely to be cost-effective.'

Richard Smith

is chief executive officer of UnitedHealth Europe and former editor of the BMJ

Martin Roland

is director of the National Primary Care Research and Development Centre, and a GP in Manchester ­ he is also spokesman for the academic team evaluating the Evercare pilots

Rate this article 

Click to rate

  • 1 star out of 5
  • 2 stars out of 5
  • 3 stars out of 5
  • 4 stars out of 5
  • 5 stars out of 5

0 out of 5 stars

Have your say