This report from the Nuffield Trust is an important addition to the literature base on end of life care as it attempts to quantify the costs involved in end of life care, from the perspectives of the various services involved.  Although there are limitations to the cost estimates, which are explicitly acknowledged, the findings will be interesting reading for commissioners, many of whom are currently reviewing their strategies for end of life care.

Why is this important?

End of life care will be a priority area for improvement for many commissioners, focusing on choice of where to die as well as alignment and coordination across services.  The current fragmentation across services is one of the reasons it’s so hard to get a handle on the costs of end of life care.

The report considers the costs across a range of services, acknowledging that this doesn’t give a complete picture (e.g. prescribing costs are missing) but does cover most key services:

• GP consultations in primary care
• District nursing contacts
• Local authority funded social care
• Inpatient hospice care
• Secondary care.

The costs are provided as averages for the final 90 days of life, which is intended to make it easier for commissioners to estimate whole population costs.  It may have been useful to include the ranges here to give an indication of variation across services.  There is also a risk that where sample data was limited, for example, for district nursing contacts (London-based), that the costs may not be representative of services in other geographies or for other populations.

Findings

A summary of the estimated costs is provided for each of the services listed above (Table 9 in the report); for reasons of space, I haven’t listed them here.

• GP consultations:  Of a sample of 21,522 people from 3 PCTs who died between April 2006 and September 2011, around one third didn’t see their GP in their last 3 months of life.  Around half had 3 or more contacts during this time.
• District nursing: Of a sample of 10,799 patients from a single London borough who died between April 2006 and September 2011, nearly two thirds had not received district nursing support in the final 3 months of life; the report notes this differs from a recent survey National Bereavement (VOICES) survey.  For those patients using the service, the findings suggest increasing use during the last month of life.
• Local authority funded social care: Of a sample of 73,243 deaths across 7 PCTs and corresponding local authorities (January 2007 – March 2010), around a quarter to a third of people access social care within the last year of life.  Usage was higher in those aged 95 or over.
• Inpatient hospice care: Costs were based on information provided by Marie Curie Cancer Care; such costs can be difficult to extract from the data, as commissioning may be based on block contracts or bed-day rates.
• Secondary care: Hospital Episodes Statistics data can be linked to Office for National Statistics mortality data to provide a summary of the activity and costs.  The report notes an increase in costs during the final week of life, mainly due to increased unplanned admissions.

The report also explores the impact of the care provided by Marie Curie Cancer Care, analysing the effect on other services.  The authors compare the activity and costs of a group of 29,538 people receiving care via Marie Curie against a matched controls sample.   The findings suggest a difference in emergency admissions, with the patients receiving Marie Cure care spending 2.5 less days in hospital, which equated to a difference of £1140 acute hospital costs per person.  The report explores this further by analysing the utilisation of other services, given the additional days at home.  Additional costs amounted to £653 per person compared to the £1140 for hospital care, suggesting that palliative care nursing delivered in the patient’s home is more cost effective.

Commentary

The authors share their methodology for calculating costs and activity and are very careful to note the uncertainties and limitations of their estimates.   Commissioners may wish to consider analysis to estimate local costs, to take account of local configuration of services and population needs.  There are some  projects now underway which are linking health and social care data.

The findings also point to patterns of utilisation during the last month of life and this highlights a number of questions:

• are processes in place to identify patients who need end of life care?
• what alternatives to A&E and emergency admissions are in place?
• are patients and their carers aware of the services available and their options for support?
• how do we know if services are meeting patients’ and carers’ needs e.g. availability when needed?

Link

Exploring the cost of care at the end of life, Nuffield Trust, 2014.