Developing a generic business case for an advanced chronic liver disease support service

Abstract

Introduction

Liver disease deaths are rising, but specialist palliative care services for hepatology are limited. Expansion across the NHS is required.

Methods

We surveyed clinicians, patients and carers to design an ‘ideal’ service. Using standard NHS tariffs, we calculated the cost of this service. In hospitals where specialist palliative care was available for liver disease, patient-level costs and bed utilisation in last year of life (LYOL) were compared between those seen by specialist palliative care before death and those not.

Results

The ‘ideal’ service was described. Costs were calculated as whole time equivalent for a minimal service, which could be scaled up. From a hospital with an existing service, patients seen by specialist palliative care had associated costs of £14 728 in LYOL, compared with £18 558 for those dying without. Savings more than balanced the costs of introducing the service. Average bed days per patient in LYOL were reduced (19.4 vs 25.7) also intensive care unit bed days (1.1 vs 1.8). Despite this, time from first admission in LYOL to death was similar in both groups (6 months for the specialist palliative care group vs 5 for those not referred).

Conclusions

We have produced a template business case for an ‘ideal’ advanced liver disease support service, which self-funds and saves many bed days. The model can be easily adapted for local use in other trusts. We describe the methodology for calculating patient-level costs and the required service size. We present a financially compelling argument to expand a service to meet a growing need.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Deaths from liver disease are increasing, but supportive and palliative care at the end of life is patchy and largely inadequate.

• There is growing recognition that supportive and palliative care should be available in parallel with other care in advanced chronic liver disease.

• Establishing new services in time of restricted funding is challenging.

WHAT THIS STUDY ADDS

• We provide a framework for an advanced liver disease support service based on surveying practitioners and patient representatives and present evidence of benefit.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• We provide a costed ‘off the shelf’ business case for an advanced liver disease service, which can be adapted to local trusts in order to facilitate roll-out of these services more widely across the UK.

Introduction

There has been a dramatic rise in deaths caused by liver disease in the UK over the past 50 years.^{1 2} As patients with cirrhosis approach the end of their life, increasingly frequent hospital attendances contribute to an escalating symptomatic, personal and financial burden. In turn, this creates huge costs for healthcare systems, which paradoxically fail to optimise quality of life and seldom use supportive and palliative care³ In England, about 70% of deaths secondary to liver disease (80% for alcohol-related liver disease) occur in hospital, compared with less than 40% for cancer⁴ and other chronic conditions such as lung disease.⁵ Supportive and palliative care aims to reduce suffering, improve quality of life and facilitate personalised planning. Provision of palliative care in advanced chronic liver disease (ACLD) is, however, challenging. The trajectory of physiological decline can be unpredictable, patients are typically young, and the potential for recompensation or transplantation can mean there is a reluctance among clinicians to engage—even when the possibility of cure is remote.

Advanced liver disease services are being increasingly used as a means to afford a wider breadth of expertise to patients with advanced liver disease, to improve co-ordination of care between hospital and community and to inform advance care planning. While there is no set quorum, membership may include hepatologists, specialists in palliative care, community nursing, alcohol support, dietetics and physiotherapy/occupational therapy. A multiprofessional approach allows patients individual needs to be addressed, for example, in the use of physiotherapy for patients with frailty, expert dietetic input for malnutrition or specialist palliative care (SPC) for patients requiring complex anticipatory care or who have symptoms that are difficult to control. This can be used to reach a consensus on a reasonable ceiling of care in the event of sudden deterioration.

Centres which lack such services may miss opportunities for advanced care planning, leading to unplanned or unwanted admissions, lengthy hospital stays and (in some cases) a futile or unwanted intensive care unit admission towards the end of life. Intensive use of resource and ‘medicalised’ dying can contribute to high-cost burdens as well as poor patient experience. Patients and families may miss out on optimal holistic support, including symptom control, psychological and spiritual support.

Provision of advanced liver disease services is patchy across the NHS and a clear concept of what ‘good’ looks like is lacking, as are data on potential cost-effectiveness. These factors limit wider roll out, particularly in the current financial climate.

We have determined the constituents of a core advanced liver disease service through a detailed survey of clinicians involved in the care of patients with liver disease. We describe potential costs and how services could be developed practically. We aimed to provide the financial case based on real-world experience and provide a generic business case for centres wishing to establish advanced liver disease support services.

Methods

Designing and costing a palliative and supportive care service for liver patients

Using a two-stage survey, we sought opinion and then agreement on the shape of an ‘ideal’ ACLD supportive care service.

We surveyed members of the British Association for the Study of the Liver end-of-life Special Interest Group. This group is made up of liver and palliative medicine healthcare professionals, including physicians, nurses and allied healthcare professionals within the fields of hepatology, gastroenterology, palliative medicine and primary care. Patient and public involvement was received from ‘LIVeR North’ and the British Liver Trust.

Initial questions (see online supplemental appendix 1) were agreed by an expert working panel from a mix of level 1, 2 and 3 centres and were based around existing models for advanced liver disease multi disciplinary teams.

Round 1 consisted of open questions relating to the components, design and name of the ideal service, core data for collection and recommendations for quality ndicators.

Following analysis of round 1 responses, a more detailed questionnaire was distributed to the group to ascertain key areas of agreement/disagreements using Microsoft Forms. A summary of questions is shown in online supplemental appendix 1. Agreement was considered achieved when a majority backed a particular statement.

An ‘ideal service model’ was then defined and costed using standard NHS salary points and tariffs to establish the price per wholetime equivalent or programmed activity for the required staff. We determined base costs for a defined minimal core service, which can be scaled appropriately depending on the population served.

Service costs in last year of life

We examined the potential cost-effectiveness of the defined ACLD MDT service by examining the real-world costs associated with liver disease patients in their last year of life (LYOL). At University Hospital Southampton (UHS), we have the beginning of an advanced liver disease support service, supported mainly by the good will and interest of our SPC colleagues. This meant that a proportion of our patients dying in the time period studied had received input from the SPC services.

Data were extracted based on admissions (using Data extracted from UHS Business Intelligence software) from 1 January 2021 to 31 December 2022 filtered to patients with ICD-10 coding as a primary or a secondary diagnosis, and where patient has subsequently died (see box 1)

Box 1.

• I85 oesophageal varices.

• I85.0 oesophageal varices with bleeding.

• I98 oesophageal varices in disease classified elsewhere.

• I85.9 oesophageal varices without bleeding.

• I98.2 oesophageal varices in diseases classified elsewhere without bleeding.

• I98.3 oesophageal varices in diseases classified elsewhere with bleeding.

• K70.3 alcoholic cirrhosis of the liver.

• K72.9 hepatic failure, unspecified.

• K74.6 other and unspecified cirrhosis of the liver.

• K76.6 portal hypertension.

• K76.7 hepatorenal syndrome.

• Tariff costs for procedures T46.1 and T46.2 (drainage of ascites) based on 2021/2022 costing data

Costing of patients in the last year of life was performed using patient-level costing reports. Costs from each admission to UHS were collated and the total summed data were also collected on total admissions (elective and non-elective), ICU admissions and total bed days. More details relating to methodology are shown in online supplemental appendix 2.

Patients discussed at least once by the palliative care MDT before they died were identified from MDT minutes and compared with those patients who died without having been reviewed by palliative care.

Average costs were calculated and used for comparison between groups for analysis.

Patients were excluded if they died on their index admission and were not from the Hospital’s primary catchment area (ie, did not have opportunity to receive palliative care from within the institution prior to death). Patients with a non-liver cause of death, but with concomitant background liver disease, were not included.

Based on the percentage of the total population of liver disease patients who died and who had been seen by a SPC healthcare professional and the benefits this bought, we modelled the effect of different percentages receiving palliative care on costs for the service and savings from its introduction.

Results

Service design

Thirty-one respondents including patient representatives, hepatology, palliative care, alcohol nurses, dieticians, palliative care doctors and hepatologists, returned surveys. Components were included where majority agreement was reached.

It is important when introducing a new service that it has an acceptable name for its users. A number of themes came through strongly in this regard, notably avoidance of the words ‘end stage’ and ‘palliative’.

The words ‘advanced liver disease’ and ‘supportive’ were favoured.

It was preferred that the service be led by hepatologists in collaboration with SPC healthcare professionals (rather than vice versa) and should incorporate a nurse-led ascitic drain service, an advanced liver disease MDT and an advanced liver disease clinic with easy ‘phone access for patients.

Core staff were identified as essential to those components and %WTE required of each allowed costings to be produced for this ‘dream team’. Full details of the service design are shown in online supplemental appendix 3.

Key information to be collected in each patients case and the recommended quality indicators of the service are shown in boxes 2 and 3.

Box 2. Core information.

• Aetiology, demographics, prognostic scores relating to liver disease.

• Complications of advanced liver disease, number of unplanned admissions in last 6 months, potential candidate for transplant.

• Opinion about appropriate ceiling of care, resuscitation decision, patients opinion repreferred place of death and ceilings of care.

• Caregiver details, physical symptoms and strategies to optimise, psychological symptoms and strategies to optimise.

• Having large volume paracentesis, suitable for TIPS? suitable for LTAD?

Box 3. Recommended quality indicators.

• Mortality review for all liver deaths essential—proportion discussed to be recorded.

• Proportion dying who had been assessed at Multi Disciplinary Team in last year.

• Proportion of LVPs carried out as day case.

• Number of non-elective admissions in last year of life.

• Proportion of patient’s dying who had advanced care plans and cardio pulmonary resuscitation status documented.

• Proportion of patient’s dying who had preferred place of death documented and achieved was desirable but not essential.

• Patient and family questionnaires to be collected.

Work force costs

NHS costs of a baseline level of service are shown in figure 1 and are based on a monthly MDT meeting, a monthly Clinic (two new, 2 Follow Up) with a full time hepatology advanced nurse practitioner (costs for both band 6 and 7) who provides the ascitic drain service and the telephone help line. It is then easy to provide the costs for a larger service with more frequent MDTs and clinics.

Figure 1.

Baseline level of service—staff costs.

Resource use

Of 168 patients who were admitted from 1 January 2021 to 31 December 2022 with a primary or secondary diagnosis relating to advanced liver disease who subsequently died were included. Of these only 16% (27/168) received SPC support.

Patients who received palliative care support were associated with reduced costs of £3830.34 per patient for non-elective admissions in the last year of life. There was also an estimated saving of 6.3 non-elective bed days per patient (table 1).

Table 1.

	Patients discussed in a palliative care MDM	Patients not discussed in a palliative care MDM
Average number of non-elective admissions	3.1	2.9
Average non-elective bed days	19.4	25.7
Average cost of non-elective admissions	£14 728.16	£18 558.50
Highest non-elective admissions total costs	£37 387.53	£86 825.43

MDM, multi disciplinary meeting.

18.5% of those seen by the SPC team (SPCT) had an ICU admission with a mean length of stay of 6 days, compared with 30% of patients not seen by them (mean length of stay 6.1 days).

Survival from first admission to death was 6 months in those seen by a member of the SPCT compared with 5 months in those who were not.

By extrapolating these numbers to model the effect of a greater proportion of patients being seen by palliative care, and so that this can be matched to service costs, we produced estimated savings from scaling up our base case service (table 2).

Table 2.

Base example expanded clinic and MDT multiplied, Hep ANP full time across all groups

	Monthly model	Twice monthly model	Weekly model	Twice-weekly model
New activity
New outpatient attendances	24	48	96	192
Follow-up outpatient attendances	24	48	96	192
Elective ascitic drains	104	104	104	104
Non-elective bed day savings	151.2	302.4	604.8	1209.6
Financial implications
Staff costs	£98 245	£109 495	£143 046	£224 786
Income from new activity	£143 896	£151 864	£167 800	£199 672
Non-elective admissions savings	£91 928.16	£183 856.32	£367 712.64	£735 425.28

While implementation of these models may result in financial savings for the health economy as a whole, funding models will have to be agreed between providers and their commissioners due to the financial implications, whether that be cost or savings, for each of them. Where savings are made, these can be shared between both provider and commissioner. Different funding models are described in online supplemental appendix 4.

Discussion/conclusions

Patients with life-limiting conditions, not just liver disease, spend a disproportionate amount of their final days in hospital. This is costly for the patient as the focus is often on unnecessary treatment escalation that results in prolonged admission, possibly futile interventions and escalation to ICU that could be avoided with timely involvement of SPC. This is also costly for the health service as a whole, which (with the best of intentions) diverts a huge amount of its resources to people in their last days but does not always help them in a holistic way. Despite this, there is a dearth of work looking at the health economics of introducing SPC services⁶ and certainly no large-scale randomised trials.⁷

We have used an expert panel to design an ideal service for support in advanced liver disease. We have then costed this service. We then used real-world data to show the impact on costs of care when compared between those receiving whose care involved SPCTs and those not.

The final business case (online supplemental appendix 4) will be publicly available on the NHS futures website and also the BASL website where it is also possible to access a number of other useful tools for setting up services to support patients with liver disease who may be approaching end of life, identifying patients who may benefit and assessing and treating symptom burden. Full text of the business case is included here in online supplemental appendix 4. We have produced a business case, which is generic and easy to replicate and can be used in trusts around the United Kingdom to support the introduction of this service. We have identified funding models that may be acceptable to payers and providers alike. We have a methodology, whereby other trusts could ‘run their own figures’ thereby understanding the local impact of such a service.

We compared costs, bed days, ICU bed days between those who died with and without SPC, but we have deliberately not provided statistical analysis on these figures as this is not a requirement of NHS business cases and because costs measured are local to the service studied. There will inevitably be differences in service provision between trusts. This was not a randomised trial, but it was based on real-world events. Likewise, we have used the result of our survey to provide recommendations for quality indicators rather than dictating key performance indicators. Patients who received SPC in this cohort may have fundamental clinical/demographic differences, which triggered the clinician to refer to palliative care, although ICU care was still received by some (so the two are not mutually exclusive and patients were not excluded).

There is clear financial benefit to trusts demonstrated here and also benefit in terms of bed day savings and ICU stay, which is in keeping with previous work from some members of our team in different NHS regions.⁸ With current NHS pressures and an emphasis on ‘flow’ within hospitals, the freeing of beds provides capacity benefits for other admissions, including to ICU . Despite less time in hospital/ICU, we found survival was slightly longer in those receiving SPC.

We do not know whether this would apply in every trust because of non-standardised care pathways, for example, in a smaller district general hospital, less use may be made of ICU with perhaps patients being sent elsewhere. We aimed to produce a generic case, with a structure for local data to be easily added in order to emphasise benefit for local patients in local trusts. We have not shown which components have the biggest impact on monetary and bed day savings. The business case model also allows for removal of components, which may not be deemed necessary (or are maybe already present).

In our extrapolated models, we have assumed that referring a greater proportion of patients to a service such as this will continue to extend its benefits. This may or may not be the case. What the upper limit of SPCT referrals is in terms of showing these benefits, we do not know. We recognise that even if a compelling case is made, it can be difficult to find the staff to recruit, especially less than whole time.

No attempt has been made to demonstrate benefits in terms of quality of life for our patients in this study. There is data available for this already^{9 10} although not extensive.^{7 9} The case for improved QOL with SPC services in liver disease has been made elsewhere¹¹ and randomised controlled trials are in progress.9 10 12 13 The aim of the work was to provide the economic argument for introduction of these services, with the expectation that patient benefit will follow with the resources in place. This work describes the costs for trusts setting up this service, which we infer leads to patient/carer QOL gains and may result in savings elsewhere in the system.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information.

Ethics statements

Patient consent for publication

Not applicable.

Ethics approval

Not applicable.