Abstract
Some people receive palliative or end of life care at home, others in hospitals or hospices, or a combination of home and hospice/home and hospital models. This rapid review aims to determine the costs and cost-effectiveness of different service models of palliative care or end of life care. These studies are mostly conducted from the perspective of the healthcare system, disregarding costs related to patients/caregivers economic burden (Perea-Bello et al., 2023).
Research Implications and Evidence Gaps More UK research is needed on cost impacts of new services such as Enhanced Supported Care (ESC). Future research should consider which methods are most appropriate to evaluate palliative care models. Standard methodology, such as the calculation of quality-adjusted life years (QALYs), may not be most appropriate for this end of life population.
Improving QALYs may not be the intended aim of palliative care or end of life interventions, and prolonging death may be inconsistent with patient preferences and wishes. The quality and applicability of the evidence we found in our rapid review were variable, and therefore, uncertainty remains, especially when the perspective of analysis was not stated clearly.
Therefore, it was difficult to ascertain whether all relevant costs were considered. Assumptions on costs were not varied in many studies, and most studies had different time horizons.
Policy and Practice Implications This rapid review has shown that hospital-based palliative care costs are higher than hospice or home-based palliative care. This suggests that home-based palliative care should be available to all patients in a recognisable end of life phase who desire to remain and die at home. Healthcare planners should aim to reduce hospitalisation at the end of life but only if access to quality home care at the end of life is guaranteed. Patients should have a choice about where they prefer to die without moving the costs from the healthcare system to the home caregivers, rendering the costs invisible.
Funding Statement The Bangor Institute for Medical and Health Research, was funded for this work by the Health and Care Research Wales Evidence Centre, itself funded by Health and Care Research Wales on behalf of Welsh Government.
2. BACKGROUND
2.1 Who is this review for?
This Rapid Review was conducted as part of the Health and Care Research Wales Evidence Centre Work Programme. The research question was suggested by the Palliative and End of Life Care Delivery Plan team. The intended audience are palliative care service commissioners and policy makers in Wales.
2.2 Background and purpose of this review
Palliative care is an interdisciplinary medical caregiving approach to optimise quality of life and mitigate suffering among people with serious, complex, and often terminal illnesses (World Health Organisation (WHO), 2023). End of life care is defined by NICE as care that is provided in the ‘last year of life’ (National Institute for Health and Care Excellence (NICE), 2019). This time frame can be difficult to predict, so some people might only receive end of life care in their last weeks or days, and others may have end of life care for longer (NHS England, 2023a).
There are many different models of palliative care, with some people receiving palliative care or end of life care at home and others in hospitals or hospices(National Institute for Health and Care Excellence (NICE), 2019). In the United Kingdom (UK), much of the palliative care provided in the community is delivered by General Practitioners (GPs), practice nurses and district nurses. Patients also move from home to hospital, sometimes to hospices, and back and forth between a variety of services several times (Papworth et al., 2023). In the UK, adult, children and young people’s palliative care are commissioned separately and managed within clinical palliative care networks (Noyes, Edwards, et al., 2013). Children’s palliative care is considered a relatively new subspeciality (established in the 1980s), and neonatal palliative care is the latest advancement of the overall model (Kain & Chin, 2020).
Hospice care aims to improve the quality of lives and wellbeing of adults, young people and children who have a terminal illness (Dreamscape and Hospice UK, 2023; Noyes, Edwards, et al., 2013; Noyes, Hastings, et al., 2013; The Kings Fund, 2018; Ziwary et al., 2017). In Wales, most hospice care is provided at home by third sector services such as Marie Curie. However, it can also be provided in a care home, as an in-patient at the hospice itself, or as a day patient visiting the hospice (Baker, 2020; McBride et al., 2011; National Health Service (NHS), 2022; The National Gold Standards Framework (GSF) Centre in End of Life Care, 2022; Wheatley & Baker, 2007). There are also nurse led service models of palliative care (Dumont et al., 2022; Salamanca-Balen et al., 2018).
For this review, the term ‘model of palliative care’ was defined as any structured care model involving multiple components, including ‘who delivers (e.g., professionals, paid carers) the intervention (specialist or generalist palliative care), where (setting, e.g., hospital), to whom (care recipients), when (i.e. timing and duration), how (e.g., face to face) and for what purpose (i.e. expected outcomes)? (Brereton et al., 2017; Davidson et al., 2006).
Generalist palliative care is provided by professionals from primary care who have good basic palliative care skills and knowledge (National Institute for Health and Care Excellence (NICE), 2023). Specialised palliative care is provided by expert clinicians who provide assessment, advice, and responsive care to people with progressive life-limiting illnesses by managing physical symptoms such as pain as well as psychological and spiritual distress (NHS England, 2023b).
The aim of this rapid review is to investigate the costs and outcomes of different service models of palliative care with a focus on end of life care. The rapid review question was: What evidence is available on the costs and cost-effectiveness of different service models of palliative care, with a focus on end of life care?
3. RESULTS
The rapid review search strategy is presented in Appendix 1 (see Section 8.1). The searches yielded a possible 101 studies, which were sought for full text screening (see Figure 1 in Section 6 for the PRISMA diagram). Fifty-six studies were deemed relevant, of which eight were systematic reviews, and 48 were primary studies.
Eight systematic reviews (SRs) were included as part of this rapid review (RR). These SRs were conducted by authors in Belgium (Simoens et al., 2010), Canada (Mathew et al., 2020), Ireland (Smith et al., 2014), Switzerland (Gonzalez-Jaramillo et al., 2021), United Kingdom (UK) (Bajwah et al., 2020; Gomes et al., 2013; Salamanca-Balen et al., 2018), and the United States of America (USA) (Yadav et al., 2020). Seven of the SRs were deemed of high quality, and one was of moderate quality (Smith et al., 2014) using the Joanna Briggs Institute (JBI) checklist for Systematic Reviews and Research Synthesis (Joanna Briggs Institute, 2017).
In terms of the UK SRs, the Bajwah et al., (2020) SR reported 13 studies which included costs of hospital palliative care (HSPC), n= 9 studies found no difference between HSPC, and usual care and two studies favoured HSPC over usual care. The difference in cost was unclear in one study. Another study reported mixed findings with lower cost of hospitalisation in favour of HSPC but no difference in the cost of emergency room visits. Four studies with full economic evaluations were inconclusive on the cost-effectiveness of HSPC (Bajwah et al., 2020). In the Gomes et al (2013) SR only two UK cost studies were reported from within the time period of this RR, with results suggesting that home based palliative care was less costly than hospital based palliative care (Gomes et al., 2013).
In a 2018 SR, two economic studies were identified. A short-term cost-minimisation study examining a telephone follow-up Clinical Nurse Specialist (CNS) led intervention for patients with breast cancer in the UK, and a cost analysis of effects of practice nurse-led care for chronic diseases. In both cases, the interventions were associated with higher costs compared to usual care although some patients preferred these care models (Salamanca-Balen et al., 2018).The SRs are presented in further detail in the evidence section (Section 7) and are not considered further in this results section as the emphasis for this RR was on the primary studies.
Of the 48 primary studies, most were cost analyses (n=39), n=1 was an SROI study (Hughes, 2021), n=5 were full economic evaluations (Isenberg et al., 2017; Pham & Krahn, 2014; Rosato et al., 2021; Saygili & Çelik, 2019; Sellars et al., 2019) and n=3 were Markov modelling studies (Kim et al., 2022; McBride et al., 2011; Nguyen et al., 2017). Of the full economic evaluations, n=5 were cost-effectiveness analyses (Isenberg et al., 2017; Pham & Krahn, 2014; Rosato et al., 2021; Saygili & Çelik, 2019; Sellars et al., 2019). A detailed summary of each included study is provided in Appendix 9.2.
Nine primary studies focused on hospital palliative care costs, three studies focused on hospice palliative care costs, 15 studies focused on community palliative care costs, and 22 studies either described or compared different models of palliative care. Of the 49 primary studies, two studies described paediatric models of palliative care, and 47 studies described adult models of palliative care. One paediatric study is described within the hospice section, and the other paediatric study is described within the community-based models of care section below.
3.1 Hospital palliative care costs
Nine studies were included which reported the costs of palliative care in hospital settings (Hanson et al., 2008; Isenberg et al., 2017; Kerr et al., 2017; McCarthy et al., 2015; Nathaniel et al., 2015; Pollock et al., 2022; Schneider et al., 2020; Sellars et al., 2019; Tan & Jatoi, 2011). With regards to quality as assessed by the JBI critical appraisal checklist for economic evaluations, n=2 were deemed to be of high quality (McCarthy et al., 2015; Sellars et al., 2019), n=6 were of moderate quality (Isenberg et al., 2017; Kerr et al., 2017; Nathaniel et al., 2015; Pollock et al., 2022; Schneider et al., 2020; Tan & Jatoi, 2011), and n=1 of low quality (Hanson et al., 2008). Most of these studies (n=6) were from the USA, and there were also studies from Australia (Sellars et al., 2019), the Netherlands (Schneider et al., 2020) and the UK (Kerr et al., 2017). N =7 studies were from a healthcare system perspective, and n=2 studies did not define the perspective of analysis. The results of these nine studies will be described under the following themes: palliative cancer care, palliative kidney disease care, general palliative care, and economic evaluations focusing on palliative care.
3.1.1 Cost studies of hospital-based specialist palliative care (cancer)
Two of the included studies focused on specialist cancer palliative care within hospital settings. One was from the USA, and the other was from The Netherlands. A retrospective cost analysis from the USA examined n=120 patients with solid tumour diagnoses admitted to hospital for end of life care. The median total cost required to provide a medical service per patient hospitalisation episode was $12,962. However, the cost per oncology patient was higher at $25,320. When adjusting for patient age at death (median age of 61 years) and days in hospital (median length of stay of four days), advance directives and route of hospitalisation were not associated with a statistically significant difference in hospital costs, with p > 0.24 and p > 0.51 respectively (Tan & Jatoi, 2011).
A cost analysis study from the Netherlands examining hospital costs of patients who died of advanced breast cancer in hospital (n=558) found overall monthly hospital costs (including medication, treatment, diagnostics, and consultation costs) of €2,255 (SD = €492) per patient. The mean cost per patient across the final twelve months of life was €21,641 (SD = €20,147). In the first seven months of admission, monthly costs remained stable, with a mean of €1,984. From the eighth month before death until the final month before death, mean costs per month steadily increased with an average increase of €343 per month, reaching a maximum of €3,614 during the last month before death. Medication costs fell after the third month of admission, while hospitalisation costs contributed to the increased cost at the end of life (Schneider et al., 2020).
3.1.2 Cost studies of hospital-based specialist palliative care (chronic kidney disease)
Two studies were included, which focused on hospital-based specialist palliative care for patients with chronic kidney disease (CKD). The studies were from the UK (Kerr et al., 2017) and the USA (Pollock et al., 2022).
A prospective cost analysis also investigating place of death, from the UK of n=211,215 patients receiving hospital care for CKD found that the mean cost of hospital admissions and outpatient care in the last twelve months before death was £11,916 for people with CKD and £7,832 for patients with other conditions. Costs increased by more than 50% in the final three months before death. The study also assessed the impact place of death had on the cost of hospital care. For people with CKD, the mean cost of hospital care in the twelve months prior to death was £9,877 for those who died at home and £12,160 for those who died elsewhere. The mean cost in the final thirty days of life was £1,077 for those who died at home and £3,206 for those who died elsewhere. People with CKD were less likely to die at home than those without CKD in every age category. Death at home was more common in men than in women in all age groups aged over 30 years old, for those with and without CKD. Overall, 12.5% of men and 8.5% of women with CKD died at home (Kerr et al., 2017).
A recent retrospective cost analysis study from the USA investigated the end of life hospital costs for patients admitted with CKD, cardiovascular (CV) and infection related admissions failure related encounters incurred longer hospital stays and higher costs than either CV or infection related encounters. The median total cost of a single patient encounter was $17,057. When disaggregated into reason for admission, the median costs were $18,469, $17,503 and $16,403 for CKD, CV and infection related admissions, respectively (Pollock et al., 2022).
3.1.3 Cost studies of hospital-based generalist palliative care
Three studies described the costs of hospital-based generalist palliative care. These studies were all from the USA (Hanson et al., 2008; McCarthy et al., 2015; Nathaniel et al., 2015).
A cost analysis study published in 2008 described the impact of interdisciplinary palliative care consultations on hospital costs at the end of life. When compared to a matched control group (matched on diagnosis), palliative care had no significant difference in variable costs across the entire hospitalisation period, with mean costs of $16,748 and $15,926, respectively (P > 0.78). The length of stay between the palliative care and control group was also not significantly different (16.6 and 13.8 days, respectively) (Hanson et al., 2008).
A retrospective modelling study investigating the cost of palliative care from the USA matched PC patients to non-PC patients (separately by discharge status) using propensity score methods). The per patient stay of Individuals who died in hospital after having received palliative care, including palliative care from a palliative care physician and palliative care registered nurse, cost $3,426 less per inpatient stay than the control group (discharge status matched controls). For the cohort of patients dying in hospital, costs without a palliative care consultation were estimated to be $33,075 compared to costs of $29,649 for patients with a palliative care physician and research nurse consultation. Consultations initiated within the first ten days of inpatient stay showed significant savings in both cohorts, with mean savings of $2,696 among patients discharged alive and $9,689 among patients who died in hospital (McCarthy et al., 2015).
A retrospective cost analysis of a hospital-based palliative care unit in the USA found that the mean cost per patient per day was $1,522 in the days before transfer to the palliative care unit (PCU). Following transfer to the PCU, the mean cost fell to $835, a saving of $687 in daily patient costs (this did not account for confounding factors). Among patients who died in the hospital, average daily direct cost per patient in the days after transfer to PCU were statistically significantly lower by $240 as compared with patients being followed by Palliative Care Consultation Service (PCCS) on the general hospital wards (SE = $45, P < 0.001) (Nathaniel et al., 2015).
3.1.4 Economic evaluations of hospital-based palliative care
Two cost-effectiveness studies investigated hospital-based palliative care services. One study was from Australia (Sellars et al., 2019), and the other was from the USA (Isenberg et al., 2017).
A cost-effectiveness study conducted in Australia with a hypothetical cohort of patients with Chronic Kidney Disease (CKD) focussed on Advance Care Planning (ACP). Advanced care planning is a voluntary process of person-centred discussion between an individual and their care providers about their preferences and priorities for their future care. The comparison group was treated with home haemodialysis (usual end of life care). The ACP intervention was, on average, AUS$519 per patient. In the decision analytic model, the average cost per patient for the ACP group was $100,579 (SD = 17,356), and the proportion of patients receiving end of life care according to preferences was 68% (SD = 48). In the no ACP group, the average cost per patient was $87,282 (SD = 19,078), and the proportion of patients having preferences met was 24% (SD = 43). The last twelve months of ACP was more expensive yet more effective in facilitating adherence to patient preferences than usual care. The incremental cost per additional case of end of life preferences being met (incremental cost-effectiveness ratio [ICER]) was $28,421. Using the Australian cost-effectiveness threshold, ACP would be highly cost-effective (Sellars et al., 2019).
Another cost-effectiveness study from the USA aimed to establish the costs of an inpatient palliative care unit (PCU) and conduct a threshold analysis to estimate the maximum possible costs for the PCU to be considered cost-effective. The authors found that PCU was cost saving, but not cost-effective compared to usual care without the inpatient PCU. The PCU can be cost-effective if the variable costs were under $559,800 (an additional $716 per patient encounter per day) (Isenberg et al., 2017). Quality appraisal of the hospital focussed studies can be viewed Appendix 9.2. See Table 1 for cost-effectiveness study tables showing the main results from See Table 2 for the table of inflated and converted hospital costs. (Caution should be taken when reading this table as the time horizons are different in nearly every row). The list of inflation calculators used for inflating and converting costs from the original currency into GBP is presented in Appendix 9.3.
3.1.5 Bottom line results for costs of hospital-based palliative care
Most of the hospital-based palliative care studies found that hospital-based palliative care costs increased in the last thirty days of life. This is due to increased hospitalisation costs as the condition of the patients deteriorates (Pollock et al., 2022; Tan & Jatoi, 2011). However, there was some evidence to indicate that if a palliative consultation was done prior to death, the hospitalisation costs were less as palliative care consultation is followed by decisions to forego costly treatment, resulting in greater cost-savings (Hanson et al., 2008; Isenberg et al., 2017; McCarthy et al., 2015). Advanced care planning is more costly but more effective in facilitating adherence to patient preferences for end of life care (Sellars et al., 2019).
3.2 Hospice palliative care costs
Four hospice focused cost papers were included in the rapid review. Two were from the USA (Gans et al., 2016; Huskamp et al., 2008), and two were from the UK (Hughes, 2021; Mitchell et al., 2020). Three publications were focused on costs, and there was also one Social Return on Investment (SROI) study comparing the social returns from inpatient and day therapy (Hughes, 2021). In terms of quality appraisal, n=1 was of high quality (Mitchell et al., 2020), and n=3 were of moderate quality (Gans et al., 2016; Hughes, 2021; Huskamp et al., 2008). Two of the studies were from a provider perspective (Gans et al., 2016; Huskamp et al., 2008), one was from an NHS perspective in the UK (Mitchell et al., 2020), and one was from the perspective of hospice providers in Wales, UK (Hughes, 2021). See Table 2 below for converted currency inflated to 2023. The list of inflation calculators used for inflating and converting costs from the original currency into GBP is presented in Appendix 9.3.
A 2020 cost analysis study from the UK aimed to examine and estimate the effects of Palliative Care Day Services (PCDS) in three centres across the UK: England, Scotland, and Northern Ireland. As well as costs, quality of life questionnaires were included (including EQ-5D-5L and ICECAP-SCM). The mean cost per attendee per day ranged from £121 to £190 across the three centres (Mitchell et al., 2020). The other UK study conducted an SROI, which yielded a greater social return on investment for those receiving palliative care at home, in comparison with those who received inpatient palliative care at a hospice in North Wales (SROI ratios of £8.97 and £2.81 respectively) (Hughes, 2021).
A cost analysis study from the USA examined patient level cost data. The mean daily cost of hospice stay was $329 in 2008. This mean cost was higher on the first and last days of hospice care. The authors found that the costs would differ for younger and older adults, with younger adults being more costly as some elements of elderly care might have already been covered by residential nursing care costs (Huskamp et al., 2008). Another study from the USA focused on the Medicaid cost of child palliative care in a hospice compared with hospice-like services for children and investigated programme enrolment data. Hospice-like paediatric palliative care programme model was $3331 less costly than the traditional hospice model ($15,643 Vs $12,312) (Gans et al., 2016). Quality appraisal of the hospice focussed studies can be viewed in Appendix 9.2. See Table 3 for the table of inflated and converted hospice costs. (Caution should be taken when reading this table as the time horizons are different in nearly every row). The list of inflation calculators used for inflating and converting costs from the original currency into GBP is presented in Appendix 9.3.
3.2.1 Bottom line results for costs of hospice-based palliative care
A paediatric palliative care programme for children with palliative care needs (who received the Partners for Children (PFC) intervention) cost £4,976 less than usual care without the extra support (Gans et al., 2016). The way that palliative care and end of life care costs are calculated vary. For example, one study calculated the cost per day of hospice care of between £151–£237 (Mitchell et al., 2020). However, other studies have calculated the mean total cost per hospice stay (of varying lengths) and found the mean cost per stay to be £2,483 in cost year 2023 (Huskamp et al., 2008). Due to the various lengths of stay, these palliative care costs are not comparable. This rapid review presents evidence that palliative care is more costly on the first and last days of end of life care in hospices and less costly for elderly people in the USA who live in residential care (due to USA medical insurance systems such as Medicare accounting for health and social care separately) (Comans et al., 2021). The SROI study conducted in Wales clearly showed a better social return on investment for those receiving palliative care at home, in comparison with those who received inpatient palliative care at a hospice in North Wales (SROI ratios of £8.97 and £2.81, respectively) (Hughes, 2021).
3.3 Home-based or community-based palliative care and end of life care
Fifteen papers that describe models of palliative or end of life care delivered in the community setting were included. The perspective of analysis from seven of these studies was not defined. In terms of quality appraisal using the JBI critical appraisal checklist for economic evaluation, n=9 were deemed to be of high quality (Amador et al., 2014; Bentur et al., 2014; Butler et al., 2022; Chai et al., 2014; Gage et al., 2015; Maetens et al., 2019; Pham & Krahn, 2014; Rosato et al., 2021; Spiro et al., 2020), n=3 were of moderate quality (Chen et al., 2018; Johnson et al., 2009; Klinger et al., 2013) and n=3 were of low quality (Enguidanos et al., 2005; Goldhagen et al., 2016; Gordon et al., 2022). One study was from a payer perspective (Chai et al., 2014), four studies were from a healthcare system perspective (Gage et al., 2015; Johnson et al., 2009; Klinger et al., 2013; Rosato et al., 2021), one study was from a payer and a healthcare system perspective (Pham & Krahn, 2014) and one study was from a third party and patient co-payment perspective (Maetens et al., 2019). Thirteen were cost analyses, and two were full economic evaluations (Pham & Krahn, 2014; Rosato et al., 2021). Of the fifteen community-based palliative and end of life care studies, four were conducted in England (Amador et al., 2014; Butler et al., 2022; Gage et al., 2015; Spiro et al., 2020), four in Canada (Chai et al., 2014; Johnson et al., 2009; Klinger et al., 2013; Pham & Krahn, 2014) four in the USA (Chen et al., 2018; Enguidanos et al., 2005; Goldhagen et al., 2016; Gordon et al., 2022), one in Israel (Bentur et al., 2014), one in Belgium (Maetens et al., 2019), and one in Italy (Rosato et al., 2021). Most of the identified studies reported on home-based models of palliative and end of life care (n=13). The remaining two studies assessed the costs of end of life care in residential care homes (Amador et al., 2014) and community-based paediatric palliative care (Goldhagen et al., 2016). Quality appraisal of the fifteen community-based studies can be viewed in Appendix 9.2.
3.3.1 Cost analyses of home-based models
Eleven cost analyses of home-based models of palliative and end of life care were identified. Of these, three were conducted in England (Butler et al., 2022; Gage et al., 2015; Spiro et al., 2020), three in the USA (Chen et al., 2018; Enguidanos et al., 2005; Gordon et al., 2022), three in Canada (Chai et al., 2014; Johnson et al., 2009; Klinger et al., 2013), one in Israel (Bentur et al., 2014), and one in Belgium (Maetens et al., 2019). The study characteristics and main findings from the cost analyses of home-based palliative and end of life care models can be found in Section 6 of this report. The studies are described via country below.
3.3.1.1 England, UK
A realist-informed mixed methods evaluation to determine optimum levels of hospice at home services with an included cost analysis of hospice at home models was conducted in England (Butler et al., 2022). Findings indicated that costs increased with proximity to death (median daily costs for days 0–14 were £104.57, compared with £56.07 for days 29–92).
Median daily formal care costs were significantly lower (£40.43, £27.93 and £12.22 for 0–14, 15–28 and 29–92 days before death, respectively) than median daily informal care costs (£580.00, £449.50 and £348.00 for 0–14, 15–28 and 29–92 days before death, respectively) (Butler et al., 2022).
From an NHS perspective, resource utilisation cost analysis study was conducted in England in 2020 to (i) compare the characteristics of rapid response service (RSS) users and non-users, and (ii) explore differences in the proportions of users and non-users dying in the place of their choice. Preferences for place of death were obtained from hospice records. (Gage et al., 2015). Findings indicated no significant differences in total service costs between users and non-users of the rapid response service. Nevertheless, overall costs were higher for rapid response users who were referred to the service two days prior to death due to the costs associated with rapid response service input and other community costs (Gage et al., 2015). A pilot cost analysis of a visit-based home care service collected day-to-day costs of hospice-at-home services for up to two weeks in the last three months of life (Spiro et al., 2020). Health and social care resource use diaries collected data for n = 333 days of hospice-at-home care completed by n = 30 families, which equated to an average of n = 11 days per family and n = 708 staff visits (equating to n = 604 hours) at a total cost of £20,192 (Spiro et al., 2020).
3.3.1.2 USA
Searches identified three cost analyses of home-based palliative and end of life care models in the USA. Firstly, a retrospective matched cohort study assessed Medicare reimbursement savings for before and after the enrolment into a home-based palliative care programme for older frail adults with advanced medical illness (Chen et al., 2018). There was a statistically significant difference between the intervention and propensity matched control group (p < 0.001) in favour of the home-based palliative care intervention, with annual Medicare saving of $18,251 per patient compared with matched control patients (Chen et al., 2018). Another study from the USA assessed healthcare resource use costs for congestive heart failure (CHF), chronic obstructive pulmonary disease (COPD) and cancer patients who received home-based palliative care compared with usual care at the end of life (Enguidanos et al., 2005). Home-based palliative care resulted in significant cost savings for patients for all patient groups compared to usual care. For example, patients with COPD enrolled in the palliative care group spent $11,325 less on average compared to those in usual care, amounting to a 67% decrease in the cost of care (Enguidanos et al., 2005). A further costing analysis study from the USA aimed to evaluate an adult home palliative care (HPC) programme for multiple insurance product lines conducted from a payer perspective (Gordon et al., 2022) 2021). Enrolment in the programme was associated with medical cost savings of $24,643 per member for the calendar year 2019 (Gordon et al., 2022).
3.3.1.3 Canada
Three studies were cost analyses of home-based palliative care models in Canada. From a societal perspective, in 2014, a cost analysis study (of the last 12 months of life) was conducted to explore the extent of unpaid care costs in total healthcare costs for home-based palliative care patients with malignant neoplasm. Findings demonstrated that unpaid care accounts for a considerable proportion (77%) of monthly home-based palliative care costs in the last twelve years of life, at an average monthly cost per patient of $11 334. For all cost categories, monthly costs increased significantly with proximity to death (Chai et al., 2014). A cost analysis conducted in 2008 evaluated the costs of a pilot interdisciplinary healthcare model of home-based palliative care from a health system perspective (Johnson et al., 2009). Total costs for 434 patients enrolled in the pilot programme were $2.4 million, equating to a cost per patient of $5,586.33 (Johnson et al., 2009).
An evaluation of resource utilisation and costs of an enhanced end of life shared-care project delivered to 95 patients (87% cancer patients; the remaining 13% mainly COPD and heart disease patients) reported total costs of $1,625,658.07, equating to $17,112.19 per patient or $117.95 per patient day (Klinger et al., 2013). Although these costs were higher than previously reported expenditures for cancer patients, the authors concluded that costs remained proportionate to funding allocated to long-term care homes and less than alternate level of care and hospital costs in Canada (Klinger et al., 2013).
3.3.1.4 Israel
A cost analysis conducted in Israel compared resource utilisation and costs of health services in the last six months with and without home-hospice care among a sample of n=429 patients with metastatic cancer (Bentur et al., 2014). Over the last six months of life, the average cost of care for patients who had received home-hospice care was US$13,648, whereas the average cost of care over the last six months for patients who did not receive home hospice care was significantly higher at a cost of US$18,503 (Bentur et al., 2014).
3.3.1.5 Belgium
In Belgium, a cost analysis assessed the quality of care and costs of palliative home care support in the last 720 to 15 days of life (Maetens et al., 2019). The analysis was conducted from a third party and patient co-payment perspective and included inpatient and outpatient cost with a matched comparison group. People (n = 8837) who received palliative home care support in the last 720 to 15 days of life were matched 1:1 by propensity score to 8837 people who received usual care. The provision of home-based palliative care was associated with more physician contacts, an increased chance of a home death and a lower chance of hospital admissions, compared with no palliative home care support. Moreover, cost findings revealed lower mean total costs of care for those who had received home-based palliative care support (€3,081 [95% CI €3,025 to €3,136] vs €4,698 [95% CI €4,610 to €4,787]; incremental cost: −€1,617 [p<0.001]) (Maetens et al., 2019).
3.3.2 Cost analysis of a residential care home model of end of life care
One cost analysis study assessed the costs of community end of life care of older people with dementia in residential care homes with no on-site nursing in England (Phase 1) and then evaluated the costs of an intervention designed to improve end of life care in this setting (Phase 2) (Amador et al., 2014). Findings from Phase 1 revealed that monthly costs per resident were £2,800. The intervention implemented in Phase 2 of the study resulted in a 43% reduction in total service costs and an 88% reduction in hospital care costs (Amador et al., 2014). Further information on study characteristics and main findings of this study can be viewed in Section 6 of this report.
3.3.3 Cost analysis of a community-based paediatric palliative care model
A cost analysis study conducted in the USA used administrative data to calculate hospital utilisation and costs associated with a community-based paediatric palliative care model that involved engaging children and families in their homes, schools, and communities to provide a holistic range of services across primary, secondary and community settings (Goldhagen et al., 2016). The community-based paediatric care model was associated with positive findings in relation to health-related quality of life (HRQoL), was generally high, and hospital charges per child declined by $1,203 for total hospital services (P = .34) and $1,047 for diagnostic charges per quarter (p=0.13). Moreover, there was a decrease in length of stay at hospital from 2.92 days per quarter to 1.22 days per quarter (P < .05) (Goldhagen et al., 2016). Further information on study characteristics and main findings of this study can be viewed in Section 6 of this report.
3.3.4 Bottom line results for costs of community-based models of palliative care
Findings from the eleven cost analyses of home-based models of palliative and end of life care reported either positive (n=6) or neutral findings (n=5) in relation to costs. Informal care makes up a significant financial contribution to models of home-based palliative care (Butler et al., 2022; Chai et al., 2014). Moreover, a residential care home model of end of life care (Amador et al., 2014) and a community-based model of paediatric palliative care (Goldhagen et al., 2016) resulted in fewer healthcare costs.
3.3.5 Economic evaluations of community-based models of palliative care
Two economic evaluations assessed the cost-effectiveness of home-based models of palliative and end of life care (Pham & Krahn, 2014; Rosato et al., 2021). In Italy, a cost-effectiveness and cost-utility analysis of a home-based palliative care approach for people living with severe multiple sclerosis (MS) compared with usual care was conducted (Rosato et al., 2021). The mean baseline-adjusted cost difference was € −394 (95% confidence interval, CI −3,532 to 2,743). The cost-effectiveness analysis considering Palliative Outcome Scale-Symptoms-MS scored revealed no change in costs. In summary, home-based palliative care did not impact QALYs and only produced slightly improved symptom scores with no increase in associated costs (Rosato et al., 2021).
A cost-effectiveness of end of life intervention study identified in the ‘end of life mega analysis’ was conducted from the Canadian health payer’s perspective (Pham & Krahn, 2014). The analysis used multiple data sources, including systematic reviews, linked health administration data, and survey data. Results from the primary cost-effectiveness analysis indicated that in-home palliative care decreased per patient costs by $4,400 and increased the likelihood of dying at home rather than in hospital by 10%, increased the average number of patient days at home (6 days) and quality-adjusted life-days (0.5 days) (Pham & Krahn, 2014). Tables 9-20 in Section 6 of this report provide further information on the study characteristics and main findings of the two economic evaluations of community-based models of palliative and end-of life care. Quality appraisal of the community-based studies can be viewed in Appendix 9.2. See Table 4 for a summary of the cost-effectiveness study relating to home/community palliative care.
See Table 5 for the table of inflated and converted community-based palliative care costs. (Caution should be taken when reading this table as the time horizons are different in nearly every row). The list of inflation calculators used for inflating and converting costs from the original currency into GBP is presented in Appendix 9.3.
3.3.6 Bottom line results for the cost-effectiveness of community-based models of palliative and end of life care
There is a dearth of evidence on the cost-effectiveness of community-based models of palliative and end of life care. However, results from one cost-effectiveness analysis indicated that in-home palliative care in Canada was cost-effective compared to hospital care (Pham & Krahn, 2014). A home-based palliative care approach in Italy for people with severe MS was similar to usual care in terms of cost-effectiveness (Rosato et al., 2021).
3.4 Models of palliative care which are not hospital, hospice or home-based, but a combination of pathways at various times
Of the n=22 combination models of palliative care papers, n=4 included outcomes relating to hospital, hospice, and home or community-based palliative and end of life care (Bjørnelv et al., 2020; Brick et al., 2017; Rolden et al., 2014; Yi et al., 2020); n=2 included cost outcomes for hospital and hospice (Hoverman et al., 2020; Saygili & Çelik, 2019); n=13 included cost outcomes related to hospital and home (Comans et al., 2021; Duncan et al., 2019; Emmert et al., 2013; Kalluri et al., 2020; Kato & Fukuda, 2017; Lustbader et al., 2017; McBride et al., 2011; Seow et al., 2022; Spilsbury & Rosenwax, 2017; Tanuseputro et al., 2015; Terada et al., 2018; Urban et al., 2018; Yu et al., 2015), n=1 related to hospice and home costs (Kim et al., 2022), and n=2 related to enhanced supported care services such as ESC and ACP (Monnery et al., 2023; Nguyen et al., 2017). In terms of quality, all (n=22) were of high quality.
3.4.1 Combination models of care: Hospital, Hospice and Home
Four studies focused on combination models of palliative care encompassing hospital, hospice, and home/community costs. These studies were published between 2014 and 2020 and were from Ireland, Norway, the Netherlands (Bjørnelv et al., 2020; Brick et al., 2017; Rolden et al., 2014) and England, Ireland and the USA (Yi et al., 2020). Three of the studies were from a healthcare system perspective, and one was from a health insurer perspective (Rolden et al., 2014). See Section 6 of this report for more details.
In Ireland, the mean total formal palliative care costs (calculated over the total sample of decedents, n=215) in the last year of life did not vary significantly across three areas of Ireland (P > 0.136). The healthcare end of life costs ranged from €50,071 in the Midlands, to €50,036 in the Mid-West, to €40,137 (Table 3) in the South East (2011 prices). Informal care was valued as the replacement cost of care (Brick et al., 2017). In 2014, in the Netherlands, the average cost of dying was €25,919 in 2014. The authors of this study included all deceased subjects for whom healthcare expenses were known for 26 months prior to death. Costs of dying were defined as healthcare expenses made in the last six months before death (Rolden et al., 2014).
End of life costs in the last three months of life was calculated in a 2020 study comparing costs in England, Ireland, and the USA. Mean care costs per person with cancer/non-cancer were US$37,250/US$37,376 (the United States), US$29,065/US$29,411 (Ireland), US$15,347/ US$16,631 (England) and differed significantly (F = 25.79/14.27, p < 0.000). In all countries, hospital care accounted for > 80% of total care costs; community care 6%– 16%, palliative care 1%–15%; 10% of decedents used ∼30% of total care costs. Being a high-cost user was associated with older age (>80 years), facing financial difficulties and poor experiences of home care, but not with having cancer or multimorbidity. Results were similar in the sensitivity analyses using the same unit costs for all countries. Hospital costs were 79%–88% of total care costs. (Yi et al., 2020). Also, in 2020, a study from Norway investigated the healthcare costs of persons now deceased. They estimated that the costs for the last six months of life were NOK 400,000 (approximately £31,734 in 2023) (Bjørnelv et al., 2020).
3.4.2 Combination models of care: Hospital and Hospice
Two papers were included that described combination models of care, including hospital and hospice care between 2019 and 2020. One study was from the USA (Hoverman et al., 2020), and the other was from Turkey (Saygili & Çelik, 2019). One study was from a healthcare system perspective (Hoverman et al., 2020), and the other was from a societal perspective (Saygili & Çelik, 2019).
A cost-effectiveness study from Turkey looked at the cost of palliative care services from a societal perspective. From a patient perspective, home healthcare services (HHC) were found to be more cost-effective compared to the other two models. The average indirect cost ($164.10) for the patients receiving care from Hospital Inpatient Services (HISs) was found to be the lowest compared with the indirect costs of HHCs ($344.62) and comprehensive palliative care centre (CPCCs) ($778.43). Estimated incremental cost-effectiveness ratios (ICERs) indicated that HHC was more likely to produce a better quality of life at the cost of an additional $33.43 per additional one quality of life (QoL) score when it was compared with HIS. However, HHC has the capability of producing a better QoL score and even reduces indirect costs ($18.30) for patients with an additional QoL score compared with CPCC. The hospital inpatient service model was found to be more cost-effective than the CPCC model (Saygili & Çelik, 2019).
The aim of the Medicare cost analysis study from the USA was to measure and characterise the total cost of care for those who received less than three days of hospice care (HC) at the end of life compared with those who received three days or more. It was found that dying in hospital was twice the cost of dying at home ($20,113 vs. $10,803). The average final 30-day spend was $22,410 if the death was at hospital, but more if a person died in the intensive care unit (ICU) at $28,301. Dying at a Skilled Nursing Facility (SNF) cost $19,400, dying in a medical hospice cost $17,418, and dying with hospice at home support cost $10,098 in 2020 (Hoverman et al., 2020).
3.4.3 Combination models of care: Hospital and Home
Thirteen of the included papers estimated the cost of hospital and home palliative care. All of the studies were cost analyses, and they were from Australia (Comans et al., 2021; Spilsbury & Rosenwax, 2017), Canada (Kalluri et al., 2020; Seow et al., 2022; Tanuseputro et al., 2015; Yu et al., 2015), England (McBride et al., 2011), Germany (Emmert et al., 2013), Japan (Kato & Fukuda, 2017; Terada et al., 2018), and the USA (Duncan et al., 2019; Lustbader et al., 2017; Urban et al., 2018). Nine of the studies were from a healthcare system perspective, one was from a payer perspective (Kalluri et al., 2020), one was from an insurer perspective (Lustbader et al., 2017), and one was from a societal perspective (McBride et al., 2011). These studies will be described below.
3.4.3.1 Australian setting
In 2017, an Australian cost analysis study found that community-based specialist palliative care was associated with a reduction of inpatient averaged hospital costs. The cohort included 12,764 decedents who, combined, spent 451,236 (9.7%) days of the last year of life in hospital. Overall, periods of time receiving community-based specialist palliative care were associated with a 27% decrease from A$112 (A$110-A$114) per decedent per day to $A82 (A$78-A$85) per decedent per day of CA hospital costs. Community-based specialist palliative care was also associated with a reduction of inpatient averaged hospital costs of 9% (7%-10%) to A$1030 per hospitalised decedent per day. Hospital cost reductions were observed for decedents with organ failures, chronic obstructive pulmonary disease, Alzheimer’s disease, Parkinson’s disease, and cancer, but not for motor neurone disease.
Cost reductions associated with community-based specialist palliative care were evident four months before death for decedents with cancer and by one to two months before death for decedents dying from other conditions.(Spilsbury & Rosenwax, 2017).
In contrast, a more recent Australian cost analysis study found that a palliative care bed in hospital cost less than a bed in an end of life palliative care facility in a modified unit in a Residential Aged Care Facility (RACF). An additional $120 per day is required to provide the higher level of care required by people with complex palliative care needs in a modified RACF unit. QoL and utility of the participants were measured at baseline, end of programme, and three and six months post baseline using the EQ-5D and ICECAP-O (Comans et al., 2021).
3.4.3.2 USA setting
Three studies which focused on hospital and home palliative care were from the USA (Duncan et al., 2019; Lustbader et al., 2017; Urban et al., 2018). In 2017, a retrospective study found that the cost per patient of end of life care in the last three months of life was $12,000 lower with home based palliative care (HBPC) than with usual care ($20,420 vs. $32,420; p = 0.0002); largely driven by a 35% reduction in Medicare Part A ($16,892 vs. $26,171; p = 0.0037). HBPC also resulted in a 37% reduction in Medicare Part B in the final three months of life compared to usual care ($3,114 vs. $4,913; p = 0.0008). Hospital admissions were reduced by 34% in the final month of life for patients enrolled in HBPC. The number of admissions per 1,000 beneficiaries per year was 3,073 with HBPC and 4,640 with usual care (p = 0.0221). HBPC resulted in a 35% increased hospice enrolment rate (p = 0.0005) and a 240% increased median hospice length of stay compared to usual care (34 days vs. 10 days; p < 0.0001) (Lustbader et al., 2017). Another Medicare cost analysis study from 2019 also presented evidence that more effective use of palliative care and hospices offers a lower cost, higher quality alternative for patients at end of life (Duncan et al., 2019). A further study from the USA utilised the linked Surveillance, Epidemiology and End Results (SEER)-Medicare database and identified a cohort of women with stage III/IV epithelial ovarian cancer diagnosed between 1995 and 2007. The authors defined the end of life as the last 90 days prior to death. They concluded that reducing the prescription of chemotherapy and increasing the use of hospice services for ovarian cancer patients is a way of reducing costs at the end of life (Urban et al., 2018).
3.4.3.3 German setting
In Germany, a cost analysis study was conducted to estimate the costs of palliative care for colorectal cancer (CRC) from the perspective of German statutory health insurance and to measure the patients’ quality of life (QoL) for a 2-year period. The mean costs per patient during the first and second years were calculated to be €42,361 and €32,023, respectively. Highest mean costs were calculated for the second quarter, which reached an amount of €12,900 (95 % CI: €11,127–€14,673)(Emmert et al., 2013).
3.4.3.4 Canadian setting
Four of the included papers were from Canada (Kalluri et al., 2020; Seow et al., 2022; Tanuseputro et al., 2015; Yu et al., 2015). The aim of one of the 2015 Canadian cost analysis studies was to examine healthcare use and cost in the last year of life. Among 264,755 decedents, the average healthcare cost in the last year of life was $53,661 (Quartile 1–Quartile 3: $19,568–$66,875). The total captured annual cost of $4.7 billion represents approximately 10% of all government-funded healthcare. Inpatient care, incurred by 75% of decedents, contributed 42.9% of total costs ($30,872 per user). Physician services, medications/devices, laboratories, and emergency departments combined to less than 20% of total cost. About one quarter used long-term care, and 60% used home care ($34,381 and $7,347 per user, respectively). Costs rose in the last 120 days prior to death, predominantly for inpatient care (Tanuseputro et al., 2015).
Another Canadian cost analysis study published in 2015 found that the estimated total societal cost of end of life care was $34,197.73 per patient over the entire palliative trajectory (four months on average). Results showed no significant difference (P > 0.05) in total societal costs between home and hospital deaths. Higher hospitalisation costs for hospital deaths were replaced by higher unpaid caregiver time and outpatient service costs for home death patients. Thus, from a societal cost perspective, alternative sites of death, while not associated with a significant change in the total societal cost of end of life care, resulted in changes in the distribution of costs borne by different stakeholders such as carers (Yu et al., 2015).
The aim of a Canadian administrative health data study published in 2020 was to evaluate the differences in resource use and associated costs of end of life care between patients with idiopathic pulmonary fibrosis (IPF) who received early integrated palliative care and patients with IPF who received conventional treatment. Multidisciplinary collaborative (MDC) patients were less likely to die in the hospital (44.9% MDC vs. 64.9% Specialised Care (SC) vs. 66.8% NSC; P,0.001) and had the highest rates of no hospitalisation in the last year of life. The median total healthcare costs in the last three months of life were approximately $CAN7,700 lower for MDC patients than for those receiving SC, driven primarily by fewer hospitalisations and emergency department visits. MDC patients were also less likely to die in the hospital (44.9% MDC vs. 64.9% SC vs. 66.8% Non-Specialist Care (NSC); P,0.001) and had the highest rates of no hospitalisation in the last year of life. (Kalluri et al., 2020).
A 2022 Canadian publication aimed to investigate the impact of early versus not-early palliative care among cancer decedents on end of life healthcare costs. In the early-palliative care group, 56.3% used inpatient care in the last month compared with 66.7% of control group (P < .001), which resulted in a statistically different average inpatient hospital costs: $7,105 (CAN$10,710) in the early group versus $9,370 (CAN$13,685) in the hard matched (on age, sex, cancer type, and stage at diagnosis) control group (P < .001). The average overall health system costs per patient in the early-palliative care group versus control group was $12,753 (CAN$10,868) versus $14,147 (CAN$14,288; P, .001) in the last month of life. The sensitivity analyses looked at early versus late paired groups and showed the same statistically significant trends as the main and sub-analysis. However, in the early versus never paired groups, the never users had lower overall costs, although this was not statistically significant (Seow et al., 2022).
3.4.3.5 Japanese setting
Two cost analysis papers were from the Japanese setting (Kato & Fukuda, 2017; Terada et al., 2018). The aim of the 2017 Japanese study was to quantify the difference between adjusted costs for home-based palliative care and hospital-based palliative care in terminally ill cancer patients. Home care was significantly associated with a reduction of $7,523 (95% CI $7,093–$7,991, P = 0.015) in treatment costs. The cost data was collected through insurance claims and medical records (Kato & Fukuda, 2017). The aim of the 2018 Japanese study was to evaluate the costs associated with healthcare and long-term care costs according to public insurance schemes during the last 24 months before death among all decedents older than 75 years, according to major disease groups. For the 2,149 decedents studied, the average healthcare costs per capita in the last 24 months of life for moderately old (75 to 84 years) and extremely old (85 years and older) decedents was 4,135,467 Japanese Yen (JPY) and 2,493,001 JPY, respectively, while the average long-term care costs per capita for 24 months was 1,300,710 JPY and 2,723,239 JPY, respectively. The total costs (healthcare and long-term care combined) ranged from 9,169,547 JPY for chronic kidney disease to 5,023,762 JPY for ischemic heart disease. In all the diseases studied, the moderately old decedents incurred higher healthcare costs, while the extremely old decedents incurred higher long-term care costs (Terada et al., 2018).
3.4.3.6 UK setting
One Markov modelling study from England, UK found that the cost to the taxpayer of providing care in the last year of life, based on 127,000 patients who died of cancer in 2006, was approximately £1.8 billion or £14,236 per patient in 2006 currency. Costs were based on daily costs of care in hospital, costs of community care, and ambulance journey costs (McBride et al., 2011).
3.4.4 Combination models of care: Hospice and Home
One Markov modelling study from Korea focussed on hospice and home palliative care from a healthcare system perspective. The incremental cost-effectiveness ratio (ICER) of the home-start group was 796,476 Korean won/quality-adjusted life week (KRW/QALW). Based on one-way sensitivity analyses, the ICER was predicted to increase to 1,626,988 KRW/QALW if the weekly cost of home-based hospice doubled, but it was estimated to decrease to −2,898,361 KRW/QALW if death rates at home doubled (Kim et al., 2022).
3.4.5 Combination models of care: Enhanced supported care services (ESC services)
Two studies were regarding enhanced supported palliative care services (Monnery et al., 2023; Nguyen et al., 2017). One was study was from a healthcare system perspective in Australia (Nguyen et al., 2017) and the other was from a societal perspective in England (Monnery et al., 2023).
An Australian Markov modelling study indicated that if the cost per individual of advanced care planning (ACP) reached $850 (equivalent to seven visits), then the programme is no longer cost-effective. Nonetheless, this scenario is unlikely because individual ACP has been provided successfully in Australia through group information sessions followed by 1–2 visits by trained ACP facilitators, which costs less than $250. At this cost, the programme is more likely to be cost-effective than the base case scenario. Extensive sensitivity analyses, including threshold analyses, were conducted on the key parameters to assess the likelihood of ACP remaining cost-effective. The result was highly sensitive to several key parameters: ACP completion and compliance rates and dying choice (hospital versus non-hospital settings). The probability sensitivity analysis of 5000 Monte Carlo replications highlighted that there was a 50–50 chance that a nationwide ACP programme would be cost-effective (see Fig. 2) due to high uncertainty around the key parameters (Nguyen et al., 2017).
A cost analysis study from England published in 2023 investigated enhanced supported care (ESC) across eight cancer centres in England. ESC service design and costs were recorded. Data relating to patients’ symptom burden were collected using the Integrated Palliative Care Outcome Scale (IPOS). For patients in the last year of life, secondary care use was compared against an NHS England published benchmark. In total, £1,676,044 was spent delivering ESC across the eight centres. Reductions in secondary care usage for the 1,061 patients who died saved a total of £8,490,581 for the UK health system. ESC services appear to be effective at supporting people dying of cancer and significantly reduce the costs of their care (Monnery et al., 2023). Quality appraisal combined models of care studies can be viewed in Appendix 9.2.
See Table 6 for the summary of the cost-effectiveness study relating to combined models of palliative care.
See Table 7 for the table of inflated and converted community-based palliative care costs.
3.4.6 Bottom line results for models of care with combined costs
There are many models of palliative care defined as any structured care model involving multiple components, including setting. Hospital care costs tend to be higher than the costs of hospice end of life care, and home-based palliative care is the least costly choice in many studies (Duncan et al., 2019; Saygili & Çelik, 2019; Yi et al., 2020).
4. DISCUSSION
This rapid review aimed to determine the costs and outcomes of different service models of palliative care or end of life care in OECD countries. In this review, the term ‘model of palliative care’ was defined as any structured care model involving multiple components, including ‘who delivers (e.g., professionals, paid carers) the intervention (specialist or generalist palliative care), where (setting e.g., hospital), to whom (care recipients), when (i.e. timing and duration), how (e.g., face to face) and for what purpose (i.e. expected outcomes) (Brereton et al., 2017; Davidson et al., 2006). Much of the focus of this RR was on the setting (hospital, hospice, home/community setting or a combination of settings).
Hospital care costs account for most of the costs of dying, and this has been reported to be 80% or more of the costs of death for some people (Duncan et al., 2019; Saygili & Çelik, 2019; Yi et al., 2020). Other authors have also reported on the high cost of dying in hospital, especially for cancer patients (Kerr et al., 2017; Tan & Jatoi, 2011). Most of the hospital-based palliative care studies found that hospital based palliative care costs increased significantly in the last thirty days of life. This is due to increased hospitalisation costs as the condition of the patients deteriorates (Pollock et al., 2022; Tan & Jatoi, 2011). However, there was some evidence to indicate that if a palliative consultation was done prior to death, the hospitalisation costs were less as palliative care consultation is followed by decisions to forego costly treatment, resulting in greater cost-savings (Hanson et al., 2008; Isenberg et al., 2017; McCarthy et al., 2015). A cost-effectiveness study conducted in Australia found that Advanced Care Planning (ACP) is more costly but more effective in facilitating adherence to patient preferences for end of life care (Sellars et al., 2019). A cost-effectiveness study from the USA found that a hospital based palliative care unit (PCU) produced cost savings and was profitable for the hospital compared to usual hospital end of life care at the hospital (Isenberg et al., 2017).
There is limited evidence for hospice care costs, and the data is not comparable across different countries and health systems currently as the times of measurements and what is measured differs a great deal (Duncan et al., 2019; Huskamp et al., 2008). This rapid review presents evidence that palliative care is more costly on the first and last days of end of life care in hospices (Comans et al., 2021).
Findings from eleven cost analyses of home-based models of palliative and end of life care reported either positive (n=6) or neutral findings (n=5) in relation to costs. Informal care makes up significant financial contribution to models of home-based palliative care (Butler et al., 2022; Chai et al., 2014). Additionally, a residential care home model of end of life care reduced healthcare costs (Amador et al., 2014).
There is a dearth of evidence on the cost-effectiveness of community-based models of palliative and end of life care. However, one cost-effectiveness analysis indicated that in-home palliative care in Canada was cost-effective and reduced costs by approximately $4,400 per patient. It also increased the chance of dying at home by 10%(Pham & Krahn, 2014). A home-based palliative care approach in Italy for people with severe MS was similar to usual care in terms of costs. The slight reduction of symptom burden produced by home-based palliative care was not associated with increased costs. The National Healthcare System and people with severe MS almost equally sustained these costs (Rosato et al., 2021).
Most of the included papers were cost analyses based on observational evidence as there are additional challenges when conducting traditional economic evaluation methods, such as cost-effectiveness studies that use quality-adjusted life years (QALY’s) as an outcome measure. Improving QALYs may not be the intended aim of palliative care or end-of life care interventions (Pham & Krahn, 2014).
There were only two studies relating to the costs of paediatric palliative care models (Gans et al., 2016; Mitchell et al., 2020). None of the studies included in our review provided evidence on the costs or cost-effectiveness of neonatal palliative care. This is a notable gap in the evidence. However, the authors are aware that a large NIHR funded study to establish children’s palliative care models is underway but has yet to be completed. This work covers Wales (Bedendo et al., 2023; Papworth et al., 2023).
The quality and applicability of the evidence we found in our rapid review were variable, and therefore, uncertainty remains, especially when the perspective of analysis was not stated clearly (Perea-Bello et al., 2023). Therefore, it was difficult to ascertain whether all relevant costs were considered. Assumptions on costs were not varied in many studies, and most studies had different time horizons.
Models of palliative care for patients with cancer, CKD, and COPD were well represented in the evidence base identified in our rapid review. However, other conditions were not well represented in the RR, indicating a gap in the evidence base.
4.1 Summary of the findings
Cost analyses were found for hospital, hospice, home-based and community-based palliative care models as well as primary care focussed models of palliative care and mixed models of care.
Generally, hospital palliative care is the most costly with a range of costs between £10,000 and £64,000 per hospital death in 2023 prices (Kerr et al., 2017; Sellars et al., 2019).
Hospice end of life care costs ranged between £2,000 and £16,000 in 2023 prices, according to the evidence presented in our review (Huskamp et al., 2008).
Only one study looked at the cost of paediatric palliative care, which ranged from £13,000 to £16,000 per death in 2023 prices (Gans et al., 2016).
In terms of home and community based palliative care costs for end of life care ranged from £900 - £21,000 in 2023 prices (Bentur et al., 2014; Spiro et al., 2020).
In terms of the combined models of palliative care studies, the costs of end of life care ranged from £5,000 to £41,000 in 2023 prices (Kalluri et al., 2020; Tanuseputro et al., 2015).
4.2 Strengths and limitations of the available evidence
The available economic evidence of palliative care is focused on costs of hospital care and less so on hospice care. Combination models of care are commonplace, and there is emerging evidence of Enhanced Supported Care (ESC), which seems to be cost saving to the NHS in the UK (Monnery et al., 2023). More studies on this type of palliative care service are needed. Assessing the cost-effectiveness of palliative and end of life care models is challenging as they often do not meet the criteria required for conducting economic evaluations. There was a lack of information about the economic perspective adopted in many of the included studies. This lack of clarity makes it difficult to determine whether all relevant costs (and outcomes in the included economic evaluations) were included in the analysis. Many of the studies were retrospective studies of decedents and not interventional studies or RCTs.
4.3 Strengths and limitations of this Rapid Review
The main strength of this rapid review is that it identified 40 relevant cost studies, and five cost-effectiveness studies, three Markov modelling studies as well as eight relevant systematic reviews (See Table 22). The main limitation was that the evidence focused on different time periods as well as various aspects of costs of services, making interpretation difficult. In this review, we focused mainly on cost analyses, as there were few cost-effectiveness papers comparing one type of palliative care model with another. We did not consider the quality of death or the burden imposed on caregivers as outcomes.
This RR used the JBI critical appraisal checklist for economic evaluations to appraise the evidence in this review. However, most of the included studies were not full economic evaluations, and therefore, some of the checklist questions were not applicable to the cost analysis study design. The lack of a standardised cost analysis quality appraisal checklist/tool limits the ability to quality appraise such studies (Xu et al., 2021). Due to the nature of such studies, they fail to meet some components of the JBI Critical Appraisal Checklist for Economic Evaluations. Notably, questions surrounding discounting, incremental analyses, and the comprehensive description of alternatives. The authors chose to extend the application of the JBI Critical Appraisal Checklist for Economic Evaluations checklist to cost analyses by awarding an equal point score to any element marked with an ‘NA’ to not penalise such studies. The scoring algorithm employed by the authors awarded a single point to any element marked Y or NA, while awarding no point to any element marked U or N. These points were totalled out of 11, and quality cut offs created to categorise the evidence into quality levels. For the costing studies, only the ‘cost’ aspect of questions 3, 5 and 6 was considered, and the outcome aspect was disregarded due to irrelevancy. Cut off scores are defined in this review as; 11 to 9 out of 11 – high quality, 6 to 8 out of 11 – moderate quality, 0 to 5 out of 11 – low quality.
Another limitation is that a specialist database of children’s palliative care studies was not searched. Future searches should include the Together for Short Lives website for children and young people’s palliative care (www.togetherforshortlives.org.uk).
4.4 Implications for policy and practice
This rapid review has shown that hospital-based palliative care costs are higher than hospice or home-based palliative care. However, hospice care can be less costly because of the number of volunteers helping to deliver the service and fundraise for the hospices, which are, in most cases, charities (at least in the UK) funded by the NHS and private donations.
Reducing hospital utilisation at the end of life should be a goal for healthcare planners only if access to quality home care at the end of life is guaranteed. Patients should be given a choice with regard to where they would prefer to die without moving the costs from the healthcare system to the home caregivers, rendering the costs invisible.
4.5 Implications for future research
More research is needed from the UK to determine the impact of new services such as Enhanced Supported Care (ESC) to examine which palliative care costs can be reduced with the implementation of such a programme.
Future research should consider which methods would be most appropriate to evaluate palliative care models. Standard economic evaluation methodology, such as the calculation of QALYs, may not be the most appropriate methodology in this end of life population (Wichmann et al., 2020). Prolonging death may be inconsistent with patient preferences and wishes. It is also important to consider appropriate patient reported outcome measures (PROMs) in this population. Some PROMs will be more suited to a particular condition e.g., dementia. PROMs are used in palliative care to evaluate the quality of care by quantifying various aspects of potential suffering, such as sleeplessness, loss of appetite, and pain. If PROM data is routinely collected, this data could be made available to researchers to evaluate end-of life care services.
Data Availability
All data produced in the present study are available upon reasonable request to the authors
6. RAPID REVIEW METHODS
6.1 Eligibility criteria
The eligibility criteria are described in Table 8.
6.2 Literature search
The search strategy conducted in Medline through OVID is presented in Appendix 1. Dates of the searches were from January 2003 to October 2023. We limited the dates of the searches to only include recent evidence from the past twenty years due to the anticipated substantial number of database hits. Previous high calibre Cochrane Reviews have previously identified literature from earlier dates. The following databases were used for the searches:
Medline
EMBASE
Cochrane Library
CINAHL
6.3 Study selection process
Two reviewers screened 100% of titles and abstracts independently using the Covidence review management software. After this, the level of agreement was assessed with disagreements settled by discussion and consensus. During independent screening, the review lead (LHS) consulted with the two reviewers to come to an agreement on the final inclusions if there was ongoing disagreement.
6.4 Data extraction
Data extraction was based on the outlined eligibility criteria. We extracted details/characteristics on study country, study design, type of intervention/model, type of economic evaluation, perspective of analysis, number of participants, relevant costs, and outcomes (see eligibility criteria) and study settings. All four members of the core BIHMR team were involved with the data extraction with the review lead (LHS) checking 25% of the data extraction tables and the other reviewers checking the remaining 25%.
6.5 Quality appraisal
Full economic evaluations and cost studies were assessed with the JBI economic evaluations checklist (Joanna Briggs Institute, 2022).
6.6 Synthesis
Due to the heterogeneity of the costs and outcomes in the included studies, a narrative synthesis of the results was reported.
7. EVIDENCE
7.1 Search results and study selection
The title and abstract searches yielded 101 inclusions. Full texts (n=77) were reviewed, and n=8 Systematic Reviews (SRs) and n=48 primary studies were included in this Rapid Review (RR). The PRISMA flow diagram is shown in Figure 1.
7.2 Data extraction
The data extraction are shown below in Tables 9-–20.
7.3 Quality appraisal
The summary tables for the quality appraisals are presented in Appendix 9.2.
7.4 Information available on request
The data that supports the findings of this study are available in the data extraction tables of this report.
8. ADDITIONAL INFORMATION
8.1 Conflicts of interest
The authors declare they have no conflicts of interest to report.
8.2 Acknowledgements
Ms Mala Mann, Information Specialist, Cardiff University is thanked for allowing us to use part of a search strategy that she has developed for another project for the Palliative Care Evidence Review Service (PaCERS). Dr Sofie Angharad Roberts is thanked for proofreading the rapid review and providing valuable feedback. We wish to thank the Health and Care Research Wales Evidence Centre Public Partnership Group - Praveena Pemmasani and Rashmi Kumar - and stakeholders - Idris Baker of Swansea Bay University Health Board, and Melanie Lewis of Cwm Taf Morgannwg University Health Board - for their time, expertise and input.
9. APPENDIX
List of Appendices
APPENDIX 1: Palliative care Rapid Review search strategy for MEDLINE (via OVID) database
APPENDIX 2: Quality Appraisal Tables
APPENDIX 3: List of inflation calculators used for inflating and converting costs from original currency into GBP
9.1 APPENDIX 1: Palliative care Rapid Review search strategy for MEDLINE (via OVID) database
Palliative care/
Terminal Care/
Terminally ill/
Hospice care/
(“palliative care” or “hospice care” or “end of life care“).tw.
((hospice or terminal*) adj3 (care or caring or ill*)).tw.
(“last year of life” or LYOL or “end of life” or “end of their lives“).tw.
(end-stage disease* or end stage disease* or end-stage ill* or end stage ill*).tw.
1 or 2 or 3 or 4 or 5 or 6 or 7 or 8
“Evidence-Based Practice“/
*“Models, Organisational“/
“Organisational Innovation“/
Diffusion of Innovation/
Patient-Centered Care/
Health Priorities/
“Delivery of Healthcare“/
((integrat* or combined or joined) adj3 (care or service or delivery or strategy* or programme* or management)).tw.
“models of healthcare“.tw.
(service adj (delivery or innovation or programme* or model*1 or restructur* or chang*)).tw.
10 or 11 or 12 or 13 or 14 or 15 or 16 or 17 or 18 or 19
Cost-benefit analysis/
Cost-effective*.tw.
Cost-benefit.tw.
Cost-utility.tw.
Cost-consequence.tw.
Cost-minimisation.tw.
Cost-minimization.tw.
Social Return on Investment.tw.
SROI.tw.
Return on Investment.tw.
Economic evaluation.tw.
Budget impact.tw.
exp “Costs and Cost Analysis“/
21 or 22 or 23 or 24 or 25 or 26 or 27 or 28 or 29 or 30 or 31 or 32 or 33
9 and 20 and 34
9.2 APPENDIX 2: Quality Appraisal Tables
List of quality appraisal tables
Table 21. Quality appraisals – hospital based palliative care. Table 22. Quality appraisals – hospice based palliative care.
Table 23. Quality appraisals – home/community-based palliative care Table 24. Quality appraisals – combined models of palliative care Table 25. Quality appraisals – systematic reviews The lack of a standardised cost analysis quality appraisal checklist/tool limits the ability to quality appraise such studies (Xu et al., 2021). Due to the nature of such studies, they fail to meet some components of the JBI Critical Appraisal Checklist for Economic Evaluations. Notably, questions surrounding discounting, incremental analyses, and the comprehensive description of alternatives. The authors chose to extend the application of the JBI Critical Appraisal Checklist for Economic Evaluations checklist to cost analyses, through awarding an equal point score to any element marked with an ‘NA’ to not penalise such studies. The scoring algorithm employed by the authors awarded a single point to any element marked Y or NA, while awarding no point to any element marked U or N. These points were totalled out of 11 and quality cut offs created to categorise the evidence into quality levels. For the costing studies, only the ‘cost’ aspect of questions 3, 5 and 6 were considered and the outcome aspect was disregarded due to irrelevancy. Cut off scores are defined in this review as; 11 to 9 out of 11 – high quality, 6 to 8 out of 11 – moderate quality, 0 to 5 out of 11 – low quality.
9.3 APPENDIX 3: List of inflation calculators used for inflating and converting costs from original currency into GBP
Currency, calculator, and web link:
AUS$ - Reserve Bank of Australia Inflation Calculator - https://www.rba.gov.au/calculator/
CAN$ - Bank of Canada Inflation Calculator - https://www.bankofcanada.ca/rates/related/inflation-calculator/
Euro – CPI Inflation Calculator: https://www.in2013dollars.com/Euro-inflation
GPB – Bank of England Inflation Calculator: https://www.bankofengland.co.uk/monetary-policy/inflation/inflation-calculator
Japanese Yen - Inflation Tool - https://www.inflationtool.com/japanese-yen
Korean Won – CPI Inflation Calculator: https://www.in2013dollars.com/south-korea/inflation
Norwegian Krone – Norges Bank Price Calculator: https://www.norges-bank.no/en/topics/Statistics/Price-calculator-/
US$ - US Inflation Calculator: https://www.usinflationcalculator.com/
Currency converter – XE Currency Converter: https://www.xe.com/currencyconverter/convert/?Amount=1&From=USD&To=GBP
Footnotes
A watermark was incorrectly left from page 43 of the manuscript onwards, stating 'Draft: Not for sharing'. I have removed this in the updated version.
1.3 List of Abbreviations
- Acronym
- Full Description
- ACP
- Advance Care Planning - Advanced care planning is a voluntary process of person-centred discussion between an individual and their care providers about their preferences and priorities for their future care.
- CAN$
- The currency of Canada (dollars)
- CBA
- Cost-Benefit analysis
- CEA
- Cost-Effectiveness Analysis
- CCT
- Controlled Clinical Trial
- CHF
- Congestive Heart Failure
- CKD
- Chronic Kidney Disease
- CNS
- Clinical Nurse Specialist-led
- COPD
- Chronic Obstructive Pulmonary Disease
- CPCC
- Comprehensive Palliative Care Centre
- CRC
- Colorectal Cancer
- CV
- Cardiovascular
- Decedents
- Also understood as ‘deceased.’
- End-of-life care (EoL)
- End-of-life care is for people who are thought to be in the last year of life. This time frame can be difficult to predict, so some people might only receive end of life care in their last weeks or days. Others may have end of life care for longer.
- Enhanced Supported Care (ESC)
- Enhanced Supportive Care (ESC) is the prevention and management of the adverse effects of cancer and its treatment.
- Euro €
- The currency of the Eurozone
- GBP £
- The currency of the United Kingdom (Great British Pounds)
- GP
- General Practitioner
- HBPC
- Home Based Palliative Care
- HSPC
- Hospital Based Palliative Care
- HatH
- Hospice at home Service
- HHC
- Home healthcare Service
- HIS
- Hospital Inpatient Service
- HC
- Hospice Care
- HRQoL
- Health Related Quality of Life
- ICER
- Incremental Cost-Effectiveness Ratio
- ICU
- Intensive Care Unit
- IPF
- Idiopathic Pulmonary Fibrosis
- JPY
- Japanese Yen
- KRW
- Korean Won
- MDC
- Multidisciplinary collaborative
- NHS
- National Health Service
- OECD
- The Organisation for Economic Cooperation and Development (OECD)
- Hospice
- A home providing care for the sick or terminally ill.
- NOK
- The currency of Norway
- NSC
- Non-Specialist Care
- ONS
- Office for National Statistics
- Palliative care (PC)
- Palliative care is an interdisciplinary medical caregiving approach aimed at optimising quality of life and mitigating suffering among people with serious, complex, and often terminal illnesses
- PCDS
- Palliative Care Day Services
- PCS
- Palliative Care Services
- PCCS
- Palliative Care Consultation Services
- PCU
- Palliative Care Units
- PFC
- Partners for Children (PFC), a paediatric palliative care pilot programme offering hospice-like services for children eligible for full- scope Medicaid delivered concurrently with curative care, regardless of the child’s life-expectancy.
- QALD
- Quality-Adjusted Life Day
- QALW
- Quality-Adjusted Life Week
- QALY
- Quality-Adjusted Life Year
- QoL
- Quality of Life
- RR
- Rapid Review
- RCT
- Randomised Controlled Trial
- SC
- Specialist Care
- SNF
- Skilled Nursing Facility
- SR
- Systematic Review
- SROI
- Social Return on Investment
- UK
- United Kingdom
- US$
- The currency of the USA (dollars)
- USA
- United States of America
- WEC
- Wales Evidence Centre (Health and Care Research Wales Evidence Centre)
- WHO
- World Health Organization
- ZBI
- Zarit Carer Burden Inventory
5 REFERENCES
