This site is intended for health professionals only

How our care plans for the elderly halved hospital costs

The problem

In most areas of the UK, general practice is seeing an increase in elderly and extremely elderly patients, all of whom will at some point become unwell and require medical care. Given the inevitability of declining health, it is worth selecting patients most at risk of admission to hospital and having a discussion with the patient and their spouse, family and friends.

This structured discussion has been shown to reduce hospitalisation, length of stay and some of the clinical chaos that can occur at the end of life. We saw specific clinical examples of patients who collapsed at nursing homes and were inappropriately resuscitated, and others who were clearly dying in hospital while undergoing futile investigations and treatment. This led to our development of what has been termed anticipatory care planning (ACP).1

We decided to carry out a pilot study of use of ACP in our practice. The Nairn practice has the highest number of patients over 65 of any practice in NHS Highland, accounting for 19.5% of our practice list. If we had not developed the ACP process, our admissions would have been comparable to the control practice in our study – and would have increased over time, with more patients dying in hospital.

What we did

We identified patients at risk of admission and developed a team who would be able to interview them and their families at home. We also had a small amount of responsive care worker time, to help prevent admission and facilitate early discharge when clinically appropriate to do so.

  • Identifying patients

In Scotland there is the risk prediction tool SPARRA (Scottish Patients At Risk of Readmission and Admission),2 and in England and Wales the King's Fund has produced the Patients At Risk of Readmission (PARR) tool. We also developed the Nairn Casefinder, which took primary care data from the GP Vision system and secondary care data from outpatient and inpatient episodes of care to weigh the risk of admission.

We were able to accurately select patients with long-term conditions – especially COPD, heart failure, cancer and dementia – who were at high risk. We focused on two categories for our care plans: all patients in care homes and the 1% of the practice population at high risk.

Each new ACP was paid on an ‘item of service' basis at £75 per patient, and a review was paid at £25. In the last two years NHS Highland spent in excess of £400,000 on ACPs, capped for a maximum of 2% of patients per practice where it was offered.

The scheme has been very successful, thanks to support from NHS Highland and fellow GP practices and is being extended to a greater percentage of patients on the practice list.

  • Care planning

For the initial study we employed a nurse, an occupational therapist and a social worker, all part-time. Eventually we settled on using a social worker to ask patients questions (a key part of the care planning process), but in smaller practices it would be reasonable to use a GP or a nurse for this. Initially, the questions asked were around:

• Demographics and contact details for the next of kin, relatives and community nursing, and where applicable social work and the key holder if the patient was in sheltered housing. Also, having a baseline functional status is important – for example, is the person mobile? What is their level of cognitive impairment? How much help do they need with the activities of daily life?

• Who was caring for the person, or were they a carer themselves? Incapacity was also discussed, which gave the opportunity for patients and carers to discuss the Continuing Power of Attorney for Welfare and Finance.

• If there was an acute medical or surgical deterioration in the patient's condition, what would they like to happen to them? Would they want intervention or not? For example, we discussed with one patient what they would like to happen if their abdominal aortic aneurysm ruptured – did they want surgery or just to be kept comfortable?

• What is the patient's understanding of their illness trajectory, and how much do they know about their illness? Do they understand what is likely to happen over the next three to five years?3

• In the event of a sudden deterioration in their health did the patient want to stay at home, or go to a GP community hospital or to the district general hospital? At the end of life and in palliative care, this discussion with patients and relatives significantly increased the likelihood of the patient dying at home or in a homely setting rather than in a hospital. This was one of the key findings from our pilot.

• In the event of a sudden deterioration and collapse, what treatment would the patient like to receive?

– to be kept comfortable (for palliative care, we referred to the Liverpool care pathway)4

– IV therapy or hospital admission for symptomatic relief

– IV fluids/antibiotics or oxygen

– high-dependency care or non-invasive ventilation

– admission to ITU (no ‘ceiling' therapy).

  • Data transfer

Once the ACP had been agreed by the patient, it was scanned into their GP Vision record using Docman. Data sharing across agencies is vital with care plans – the ACP was entered onto the ADASTRA system to ensure the information was available to the out-of-hours service, in A&E and on the wards in admitting hospitals.

Despite investment this has not yet been perfected, but it is hoped the Scottish Key Information Summary will be the online repository for ACPs in due course.

Lessons learned

  • Interviewing

The approach to interviewing needs to be sensitive, caring and compassionate. The interview is key to gaining the confidence of the patient, carers and family. It goes into areas that traditionally are taboo in our society – but patients invariably welcome the opportunity to be able to discuss these issues and future problems.

For example, discussing ceiling therapy was initially regarded as controversial, but patients were relieved at having the opportunity to discuss this aspect of care.

Consultation allows patients to prepare all parties concerned for an inevitable deterioration in their health.

  • Power of Attorney

Discussing the importance of Power of Attorney for welfare and finance with patients and carers – particularly where there is likely to be cognitive decline – has been essential. One of the formative aspects to the project was the 80% mortality rate for patients waiting for a guardianship procedure in one community hospital.


The project started in 2006 in Nairn and was sponsored and supported by NHS Highland. Given its success it was scaled up in 2009/10, and over 4,000 ACPs had been developed by 2011. Over three years, NHS Highland has invested £400,000 in ACPs and we are seeing a reduction in admissions and length of stays in the over-75s, which contradicts the demographic trend.

The original pilot study of 100 patients and 100 controls had an 15-16% mortality rate. Of the survivors, there were 42% fewer admissions (P = 0.002) and 52% fewer bed days (P = 0.020) in the study population. The cost of hospitalisation for this group was reduced by 49% (P = 0.029). The control population had 11 patients die in hospital, while the study population had three deaths in hospital (P = 0.007).

By 31 March 2011, the 5,329 patients with ACPs demonstrated a 29% reduction in emergency new admissions and a 47% reduction in occupied bed days. Patients with a SPARRA score of greater than or equal to 50% but no ACP in place showed an increase in both emergency admissions (+59%) and occupied bed days (+63%).

For patients with an ACP, there is more chance of dying at home rather than in hospital. We also found there are better-managed home visits, more structured and organised meetings with families and less work in resolving guardianship cases.


Subject to local negotiations, there will be continued funding from NHS Highland and we are hoping for ACP to be incorporated into mainstream funding in 2013. We will aim to target ACPs at care homes and up to 5% of the practice population.

Dr Adrian Baker is a GP and clinical lead in Nairn in the Scottish Highlands


1 Baker A, Leak P, Ritchie LD et al. Anticipatory care planning and integration: a primary care pilot study aimed at reducing unplanned hospitalisation. Br J Gen Pract 2012;62:e113-20

2 The Scottish Government. Scottish patients at risk of readmission and admission (SPARRA) made easy. 2010.

3 Murray S, Sheikh A and Thomas K. Advance care planning in primary care. BMJ 2006;333:868-9

4 The Marie Curie Palliative Care Institute. Liverpool care pathway for the dying patient. 2009.