Abstract
Context
Hospital deaths carry a significant healthcare cost that has been confirmed to be lower when palliative care units (PCUs) are available.
Objectives
To compare the last admission hospital health care cost of dying in a first-level hospital between the PCU and the rest of the hospital services.
Methods
A retrospective, comparative, observational study evaluating costs from the payer perspective on treatments and diagnostic-therapeutic tests performed on patients who die in first-level hospital, comparing whether they were treated by the PCU or another unit (Non-PCU). Patients with a mortality risk >2 were included according to the Severity of Illness Index (SOI) and Risk of Mortality (MOR). All cost express in €, median per patient and interquartile range (IQR).
Results
From 1,833 patients who died, 1,389 were included, 442 (31.1%) treated by PCU and 928 (68.9%) Non-PCU. Statistical differences were found for the last admission total cost (€262.8 (€470.1) for PCU versus €515.3 (€980.48) in Non-PCU), daily total cost (€74.27 (€127.4) vs €115.8 (€142.4) Non-PCU). Savings were maintained when the sample was broken down by diagnosis-related group (DRG) and a multivariate analysis was performed to determine how the different patients baseline characteristics between PCU and Non-PCU patients influenced the results obtained.
Conclusions
Data from this study show that cost is significantly lower when the patients are treated by a PCU during their last hospital stay when they pass away.
Key Words
Introduction
It is well known that most of the healthcare expenditure in Western countries is generated at the end of life, within the last 6 months of life accounting for 40% of healthcare expenditure.
1
, 2
, 3
This situation, combined with an ageing population in this part of the world, has led to a rapid growth in healthcare expenditure in recent years.4
Certain studies demonstrate that palliative care (PC) teams decrease the cost of clinical attendance in patients with advanced disease without reducing the quality.
5
, 6
, - Gade G
- Venohr I
- Conner D
- et al.
Impact of an inpatient palliative care team: a randomized control trial.
J Palliat Med. 2008; 11: 180-190https://doi.org/10.1016/j.jpainsymman.2009.05.00110.1089/jpm.2007.0055
7
However, despite the demonstrated effectiveness of care by PC teams, the implementation of PC remains deficient in many countries.8
It is estimated that worldwide 56.8 million people need PC (25.7 million in the last of life), situation being observed unequally depending on country.9
Although sustained growth in palliative care is evident, it has slowed in the last decade and the situation is unequally depending on the country.8
,10
, 11
, 12
, The Economists Intelligence Unit. The 2015 quality of death index: ranking palliative care across the world. Economist Intell Unit 2021. Available at:https://impact.economist.com/perspectives/healthcare/2015-quality-death-index. Accessed March 20, 2022
13
, 14
Osman H, Rihan A, Garralda E, et al. Atlas of palliative care in the eastern Mediterranean Region. 2017. Available at: http://www.thewhpca.org/resources/global-atlas-on-end-of-life-care. Accessed May 19, 2022.
In Spain, public expenditure on health grew by 4,193.5 million euros in 2019, this has been related to increased emergency care visits and the greater number of unplanned admissions in the last weeks of life.
9
,15
Although palliative coverage is only available in this country for 50% of the population, it can be stated that, as in other countries with similar healthcare systems, it has resulted in savings and proven lower cost thanks to the optimization of the aforementioned healthcare resources.The total expenditure of the public Health System rose by 5.4% in 2019 compared to the previous year. [Government/News/Health]. Available at: https://www.lamoncloa.gob.es/serviciosdeprensa/notasprensa/sanidad14/paginas/2021/080721-gasto-sanitario.aspx. Accessed August 2, 2022.
15
, The total expenditure of the public Health System rose by 5.4% in 2019 compared to the previous year. [Government/News/Health]. Available at: https://www.lamoncloa.gob.es/serviciosdeprensa/notasprensa/sanidad14/paginas/2021/080721-gasto-sanitario.aspx. Accessed August 2, 2022.
16
, 17
, 18
, 19
Most of the PC teams cost studies measure the impact of the home PCU on the number of hospitalizations, length of hospital stays, number and length stays in the Intensive Care Unit (ICU), chemotherapy use, hospital emergency room visits and readmissions at 30 days post-discharge. Only a few analyse the hospitalization costs, but they focus on specific interventions, and do not compare hospital direct cost between PCU and the rest of the services.
20
, 21
, The effectiveness and cost-effectiveness of hospital-based specialist palliative care for adults with advanced illness and their caregivers.
Cochrane Database of Systematic Reviews. 2020; (Issue 9Art. No.: CD012780.)https://doi.org/10.1002/14651858.CD012780.pub2
22
The aim of this study, is to analyse the of end-of-life patient care cost during their last admission, comparing whether patients are hospitalized in the Palliative Care Unit (PCU) or in another Hospital Unit (Non-PCU).
Methods
A retrospective, comparative, observational, resource consumption analysis, from the hospital perspective, was conducted on patients who died of advanced disease in Infanta Elena Hospital (IEH) during a three-year period (January 2016 to December 2018). IEH is a University District Hospital, located in Valdemoro, Community of Madrid, Spain, that has 135 admission beds and a palliative care unit (PCU) that assists between 8 and 14 admitted patients daily, staffed by a physician, a nurse, and a psychologist.
Cost analysis, from the hospital perspective, was performed to compare the results between those who were attended in a PCU versus those attended in other departments (Non-PCU). Other Hospital departments with admissions available are general and digestive surgery, internal medicine, cardiology, pneumology, haematology, urology, traumatology, neurology, digestive and geriatrics and gynaecology.
The choice of service where the patient must be hospitalized in IEH is decided by a multidisciplinary team composed of all the specialties with admission beds. PCU mainly receives oncologic patients and others that have already been treated by them.
To ensure the homogeneity of the groups compared, clinical records both PCU and Non-PCU were revised by the principal investigator. Only patients with palliative care needs and high mortality risk were included. NECPAL 4.0 prognostic and palliative care needs assessment tool for inclusion in the palliative care programs was employed to confirm patients palliative care needs.
23
The Severity of Illness Index (SOI), and the Risk of Mortality (MOR) were used as mortality criterion. According to these indices, mortality risk is defined as the probability of dying from the diagnosis that led to the admission, assessed on a scale from 1 to 4, where 1 is the lowest risk, 2 moderate, 3 major, and 4 extremes.24
, 25
, 26
Patients with a severity and mortality (MOR & SOI) greater than 2 were included.Children under 18 years of age, deaths from accidents or acute complications and patients admitted in agony (signs of imminent death) were excluded.
The main cost variables analysed per patient in the last hospital admission are and include:
- Pharmacological treatment cost: enteral nutrition and drugs administered (calculated for each prescribed drug and patient through the administered drug regimen and the unitary cost).
- Other intervention cost: surgeries, transfusions, laboratory and imaging tests performed.
- Last admission cost: pharmacological treatment and other interventions.
- Total daily cost: last admission cost divided into the hospital stay length.
Healthcare personnel and services infrastructures costs were not included because they are similar (data obtained did not provide differences between the salary of healthcare personnel) among the different hospital services and the results have been adjusted by day of admission.
IEH Management and Pharmacy department provided the pharmacological treatments during cost data. All unit costs of each resource consumed between 2016-2018 expressed in 2020 € and were updated with the consumer price index variation provided by the Spanish National Statistics Institute (INE).
27
National Institute of Statistics (INE). Calculation of the Consumer Price Index percentage changes. Available at: https://www.ine.es/varipc/index.do. Accessed March 7, 2021.
The unitary costs per diagnostic test and transfusions are those referred in Ministry of Health 727/2017 Order of 7 August 2017, which establishes the provision public prices of healthcare services and activities for the centre network in the Community of Madrid public prices, also updated to 2020.
28
ORDER 727/2017, of August 7, of the Regional Minister of Health, by which public prices are set for the provision of services and activities of a healthcare nature in the network of centers of the Community of Madrid. August 21, 2017, Madrid Official State Bulletin. Available at:http://www.madrid.org/wleg_pub/secure/normativas/listadoNormativas.jsf#no-back-button. Accessed February 7, 2021.
A descriptive and univariate statistical analysis was performed initially comparing the results obtained in the group of patients who died in the PCU versus those who died in other units (Non-PCU). To analyse possible differences in results by reason for admission, the database was segmented by the groupings done according to the Diagnosis-Related Group (DRG).
29
Ginard- Alonso, MA. Ruiz-Escobar P, Ruiz Escobar J, Acevedo- Adán MJ. The GRD'S in the National Health System. Ocronos -Scientific-Technical Editorial. Available from:https://revistamedica.com/grds-sistema-nacional-salud/. Accessed February 7, 2022.
Categorical variables were expressed as relative and absolute frequencies, and quantitative variables were expressed as the median and interquartile range since most of them did not follow a normal distribution. Comparisons were subsequently made using the chi-squared test for dichotomous qualitative variables, and Fisher's test was used in the event the two variables to be compared were dichotomous. Quantitative variables were compared using a student's t test or ANOVA for independent samples in case of normal distribution, and a Wilcoxon test and a Friedman test when variables did not follow a normal distribution.
Multivariate analysis (binary logistic regression), was carried out to determinate whether the possible differences between PCU and Non-PCU patients influenced the results obtained for the patient total daily cost (categorized as a dichotomous variable using the median obtained for the total study population as reference value).
All analyses were performed using the IBM SPSS Statistics v26 and RStudio statistical package.
Results
Of the 1,833 patients who died at IEH between January 2016 and December 2018, 1,591 had palliative needs (NECPAL+), 1,370 met the inclusion criteria, 442 from PCU and 928 Non-PCU. 221 patients excluded, 202 Non-PCU because their mortality risk was ≤ 2 and 19 PCU because they were admitted with terminal sedation.
Univariate analysis to assess homogeneity of the groups of patients seen or not by the PCU showed no differences by severity index or mortality risk between the groups (Table 1). Differences by age, type of disease and length hospital stay were found.
Patients age at the time of last hospital admission. p-value significance level <0.05. DRG levels that caused last hospitalization includes: Oncological (oncology or hematology neoplasms, lymphomas and neo-formations), Sepsis (infection or sepsis), Respiratory (pneumonia, bronchitis, chronic obstructive disease…), Cardiovascular disease (acute myocardial infarction, cardiovascular events, transient ischemic attack, convulsions, nervous system vascular diseases), Other heart disease (congestive heart failure and other heart failure), Kidney and liver failure (also hepatobiliary disorders), Other gastrointestinal diseases and Other diagnosis (mainly endocrine disorders, musculoskeletal diseases and trauma).
Table 1Characteristics of the study population and their distribution according to whether they were treated or not in the Palliative Care Units (PCU) of Infanta Elena Hospital.
| Total | PCU | Non-PCU | p-value | |
|---|---|---|---|---|
| N | 1370 | 442 (32.3%) | 928 (67.7%) | |
| Age: median IQR | 85 (3) | 83 (16) | 86 (11) | p<0.001 |
| Age: N (%) | ||||
| < 75 years | 288 (21.0%) | 136 (30.8%) | 152 (16.4%) | p<0.001 |
| ≥ 75 years | 1082 (79.0%) | 306 (69.2%) | 776 (83.6%) | |
| Sex: N (%) | ||||
| Male | 666 (48.6%) | 230 (52.0%) | 436 (47.0%) | p= 0.080 |
| Female | 705 (51.4%) | 212 (48.0%) | 593 (53.0%) | |
| Year of death: N (%) | ||||
| 2016 | 430 (31.4%) | 145 (32.8%) | 285 (30.7%) | p=0.086 |
| 2017 | 473 (34.5%) | 164 (37.1%) | 308 (33.2%) | |
| 2018 | 468 (34.1%) | 133 (30.1%) | 335 (36.1%) | |
| Severity Index: N (%) | ||||
| 2 | 59 (4.3%) | 14 (3.2%) | 45 (4.8%) | p=0.307 |
| 3 | 684 (49.9%) | 228 (51.6%) | 456 (49.1%) | |
| 4 | 628 (45.8%) | 200 (45.2%) | 427 (46.0%) | |
| Mortality risk: N (%) | ||||
| 3 | 639 (46.6%) | 212 (48.0%) | 427 (46.0%) | p=0.499 |
| 4 | 732 (53.4%) | 230 (52.0%) | 501 (54.0%) | |
| Diagnosis (DRG) ⁎⁎⁎ : N (%)DRG levels that caused last hospitalization includes: Oncological (oncology or hematology neoplasms, lymphomas and neo-formations), Sepsis (infection or sepsis), Respiratory (pneumonia, bronchitis, chronic obstructive disease…), Cardiovascular disease (acute myocardial infarction, cardiovascular events, transient ischemic attack, convulsions, nervous system vascular diseases), Other heart disease (congestive heart failure and other heart failure), Kidney and liver failure (also hepatobiliary disorders), Other gastrointestinal diseases and Other diagnosis (mainly endocrine disorders, musculoskeletal diseases and trauma). | ||||
| Oncological | 177 (12.9%) | 130 (29.4%) | 47 (5.1%) | P <0.001 |
| Sepsis | 470 (34.3%) | 166 (37.6%) | 304 (32.7%) | p=0.078 |
| Respiratory disease | 357 (26.0%) | 76 (17.2%) | 281 (30.2%) | p=<0.001 |
| Cardiovascular disease | 153 (11.2%) | 38 (8.6%) | 115 (12.4%) | p=0.038 |
| Other heart disease | 67 (4.9%) | 14 (3.2%) | 53 (5.7%) | p=0.042 |
| Liver or kidney failure | 47 (3.4%) | 7 (1.6%) | 40 (4.3%) | p=0.010 |
| Other GI disease | 48 (3.5%) | 5 (1.1%) | 42 (4.5%) | p=0.001 |
| Others | 52 (3.8%) | 6 (1.4%) | 46 (5.0%) | p=0.001 |
| Length of hospital stay (days): median IQR | 5 (7) | 4 (5) | 6 (7) | p<0.001 |
DRG: Diagnosis Related Group that caused the hospital admission; GI: GastroIntestinal; IQR: Inter-Quartile Range.
Table 2 shows a lower consumption of resources when patients were attended by the PCU. The proportion of patients in whom surgery, diagnostic and laboratory tests are performed, parenteral or enteral nutrition is administered, as well as the number of drugs administered, is always lower and statistically significant when patients hospitalized in PCU.
p-value significance level <0.05. Patients who had at least one.
Table 2Patients resource consumption in the last week of life depending on whether they were treated in palliative care units (PCU) or Non-PCU.
| Total number of patients | PCU | Non-PCU | p-value |
|---|---|---|---|
| N (%) | 442 (32.3%) | 928 (67.7%) | |
| Hospital length stay (days) Median (IQR) | 4 (5) | 6 (7) | p<0.001 |
| Laboratory Tests done: Median (IQR) | 2 (3) | 4 (4) | p<0.001 |
| CT Scan N (%) | 63 (14.3%) | 241 (26.0%) | p<0.001 |
| X-ray test N (%) | 305 (69.0%) | 785 (82.9%) | p<0.001 |
| Parenteral Nutrition N (%) | 69 (15.6%) | 195 (21.0%) | p=0.018 |
| Surgeries N (%) | 4 (0.9%) | 46 (5.0%) | p<0.001 |
| Transfusions N (%) | 47 (5.0%) | 14 (3.2%) | p=0.235 |
| Total number of drugs prescribed in the last day of life: Median (IQR) | 18 (11) | 20 (12) | P<0.001 |
| 14 (8) | 17 (9) | P<0.001 |
| 3 (2) max:11 | 3 (2) max:7 | P=0.001 |
PCU: Palliative Care Unit; %: percentage; IQR: Inter-Quartile Range; CT-scan: Computerized Tomography Scan PRN: pro re nata.
The cost assessment results demonstrates that cost are significantly lower in the PCU patients compared to Non-PCU (Table 3). These median differences are €93.3 for the pharmacological treatment cost, €92.0 for other interventions cost, €252.5 for last admission total cost and 41.5€ for total daily cost.
Table 3Cost assessment of dying in the hospital comparing between PCU and Non-PCU departments.
| Total | PCU | Non-PCU | PCU vs Non-PCU | |||||
|---|---|---|---|---|---|---|---|---|
| Cost per patient (€) | Median | IQR | Median | IQR | Median | IQR | Median Difference | p-value** |
| Pharmacological treatment cost | €123.9 | €370.5 | €65.8 | €157.3 | €159.1 | €452.9 | -€93.3 | p<0.001 |
| Other interventions cost | €214.0 | €347.5 | €152.0 | €226.0 | €244.0 | €383.8 | -€92.0 | p<0.001 |
| Last admission total cost | €430.8 | €799.6 | €262.8 | €470.1 | €515.3 | €980.9 | -€252.5 | p<0.001 |
| Total daily cost | €101.17 | €142.9 | €74.3 | €127.4 | €115.8 | €127.4 | -€41.5 | p<0.001 |
| Patients (N) | 1331 | 442 | 928 | |||||
PCU: Palliative Care Unit; €: euros; IQR: Inter-Quartile Range. *p-value significance level <0.05.
The cost analysis breakdown by admission diagnosis group (Table 4) maintains the differences for the last admission total cost but not for the daily total cost. Table 5 shows the multivariate analysis results when we include DRG, department type, age group, and gender to determine which variables has a statistically significant influence on having a cost higher than the population median (€101,17) the cost per patient. Age lower than 75 years, Non-PCU and cardiovascular disease DRG shows higher risk of having a cost over the median per patient.
p-value significance level <0.05.
Table 4Expenditure breakdown by diagnosis that caused last hospitalization according to DRG* of cost assessment of dying in the hospital comparing between PCU and Non-PCU departments (expressed in € with median per patient and IQR).
| Patients diagnosed with Oncological | PCU | Non-PCU | p-value |
|---|---|---|---|
| N=173 | N=26 | N=47 | |
| Pharmacological treatment cost | €75.3 (862.2) | €307.4 (862.2) | p<0.001 |
| Other interventions cost | €183.0 (316.3) | €725.0 (1352.5) | p<0.001 |
| Last admission total cost | €386.7 (980.9) | €1219.90 (1889.7) | p<0.001 |
| Total daily cost | €80.2 (195.9) | €165.6 (201.4) | p=0.165 |
| Length of hospitalization (days) | 4.0 (6.0) | 16.0 (17.0) | p<0.001 |
| Total number of drugs prescribed in the last day of life: | 18.5 (13.25) | 21 (13) | p<0.001 |
| 14 (8.25) | 21 (12) | p<0.001 |
| 4 (2) | 2 (3) | p=0.002 |
| Patients diagnosed with Sepsis | PCU | Non-PCU | p-value* |
| N=470 | N=163 | N=294 | |
| Pharmacological treatment cost | €60.3 (111.6) | €174.9 (450.5) | p<0.001 |
| Other interventions cost | €124.0 (180.0) | €212.0 (326.0) | p<0.001 |
| Last admission total cost | €207.9 (248.9) | €462.7 (758.3) | p<0.001 |
| Total daily cost | €63.5 (93.5) | €110.1 (131.6) | p<0.001 |
| Length of hospitalization (days) | 4.0 (5.0) | 5.0 (7.0) | p=0.032 |
| Total number of drugs prescribed in the last day of life: | 17 (10) | 19 (11) | p=0.024 |
| 14 (6) | 17 (8.25) | p<0.001 |
| 3 (2) | 3 (2) | P=0.569 |
| Patients diagnosed with Respiratory Disease | PCU | Non-PCU | p-value* |
| N=357 | N=76 | N=281 | |
| Pharmacological treatment cost | €32.4 (81.6) | €115.9 (391.0) | p=0.009 |
| Other interventions cost | €203.8 (256.0) | €349.5 (672.3) | p=0.018 |
| Last admission total cost | €279.0 (431.6) | €647.9 (1000.1) | p=0.007 |
| Total daily cost | €91.7 (132.2) | €132.8 (151.7) | p=0.138 |
| Length of hospitalization (days) | 3.0 (4.0) | 6.0 (8.0) | p=0.097 |
| Total number of drugs prescribed in the last day of life: | 16 (12.3) | 19 (13) | p=0.296 |
| 12 (8) | 15 (11) | p=0.026 |
| 3 (2) | 3 (3.5) | p=0.006 |
| Patients diagnosed with Cardiovascular Disease | PCU | Non-PCU | p-value* |
| N=153 | N=38 | N=105 | |
| Pharmacological treatment cost | €32.4 (81.6) | €115.9 (391.0) | p=0.022 |
| Other interventions cost | €203.8 (256.0) | €349.5 (672.3) | p=0.001 |
| Last admission total cost | €279.0 (431.6) | €647.9 (1018.2) | p=0.001 |
| Total daily cost | €91.7 (132.2) | €132.8 (151.7) | p=0.056 |
| Length of hospitalization (days) | 3.0 (4.0) | 6.0 (8.0) | p=0.233 |
| Total number of drugs prescribed in the last day of life: | 16 (12.3) | 19 (13) | P=0.390 |
| 12 (8) | 15 (11) | p=0.027 |
| 2 (3) | 3 (3.5) | P=0.008 |
| Patients diagnosed with Other Heart Disease | PCU | Non-PCU | p-value* |
| N=67 | N=14 | N=53 | |
| Pharmacological treatment cost | €89.6 (414.5) | €328.1 (719.9) | p=0.041 |
| Other interventions cost | €168.0 (183.5) | €244.0 (276.0) | p=0.185 |
| Last admission total cost | €317.3 (497.3) | €716.0 (940.7) | p=0.019 |
| Total daily cost | €93.8 (119.9) | €109.3 (99.2) | p=0.388 |
| Length of hospitalization (days) | 5.0 (5.0) | 6.0 (6.0) | p=0.085 |
| Total number of drugs prescribed in the last day of life: | 21.5 (6.8) | 24 (12.5) | p=0.757 |
| 19 (8.8) | 20 (10) | p=0.256 |
| 3.5 (2.3) | 3 (3) | p=0.594 |
| Patients diagnosed with Liver or Kidney Failure | PCU | Non-PCU | p-value* |
| N=51 | N=7 | N=40 | |
| Pharmacological treatment cost | €60.3 (56.2) | €277.2 (879.3) | p=0.009 |
| Other interventions cost | €122.0 (257.5) | €380.5 (782.4) | p=0.005 |
| Last admission total cost | €261.1 (301.1) | €818.9 (2295.9) | p=0.003 |
| Total daily cost | €65.3 (230.2) | €129.2 (184.7) | p=0.150 |
| Length of hospitalization (days) | 4.0 (2) | 7 (17) | p=0.104 |
| Total number of drugs prescribed in the last day of life: | 19 (9) | 21 (10.5) | p=0.919 |
| 9.0 (2) | 18 (11.8) | p=0.004 |
| 3 (1) | 3 (3.8) | p=0.826 |
| Patients diagnosed with Other Gastrointestinal Disease | PCU | Non-PCU | p-value* |
| N=48 | N=5 | N=42 | |
| Pharmacological treatment cost | €125.2 (415.1) | €96.2 (422.4) | p=0.880 |
| Other interventions cost | €499.2 (849.8) | €413.3 (475.3) | p=0.880 |
| Last admission total cost | €705.1 (968.3) | €556.8 (1073.3) | p=0.336 |
| Total daily cost | €260.9 (315.2) | €138.5 (246.6) | p=0.256 |
| Length of hospitalization (days) | 4.0 (4.0) | 4.5 (7) | p=0.651 |
| Total number of drugs prescribed in the last day of life: | 19 (7.5) | 17 (11) | p=0.854 |
| 14 (6) | 16.5 (9) | p=0.287 |
| 5 (1.5) | 3 (2) | p=0.075 |
| Patients with Other Diagnosis | PCU | Non-PCU | p-value* |
| N=52 | N=6 | N=46 | |
| Pharmacological treatment cost | €24.9 (146.9) | €299.9 (523.9) | p=0.007 |
| Other interventions cost | €92.0 (63.5) | €411.0 (802.4) | p=0.002 |
| Last admission total cost | €131.4 (234.1) | €960.9 (1507.4) | p=0.001 |
| Total daily cost | €84.9 (51.3) | €119.0 (169.1) | p=0.199 |
| Length of hospitalization (days) | 2.5 (4.0) | 7.0 (8.0) | p=0.014 |
| Total number of drugs prescribed in the last day of life: | 13.5 (14.3) | 24 (12.5) | P=0.055 |
| 10 (11.3) | 21 (10) | P=0.033 |
| 2 (4) | 5 (3) | P=0.068 |
PCU: Palliative Care Unit; €: euros; IQR: Inter-Quartile Range; PRN: pro re nata.
Table 5Univariate analysis and backward multivariate analysis between the independent variables (DRG, department type, age and gender) and cost higher or lower than the median population (€107.17).
| Univariate | Multivariate | ||||||
|---|---|---|---|---|---|---|---|
| N | ≥€101,17n (%) | > €101,17n (%) | p-value* | OR | 95% CI | p-value* | |
| Diagnosis (DRG) | 0.012 | ||||||
| Oncological | 173 | 86 (49.7%) | 87 (50.3%) | 1,31 | 0.67,2.55 | 0.426 | |
| Sepsis | 457 | 251 (54.9%) | 206 (45.1%) | 1,09 | 0.60,1.97 | 0.783 | |
| Respiratory disease | 346 | 179 (51.7%) | 167 (48.3%) | 1,13 | 0.62,2.06 | 0,682 | |
| Cardiovascular disease | 143 | 54 (37.8%) | 89 (62,2%) | 2,15 | 1.11,4.16 | 0.022 | |
| Other heart disease | 67 | 32 (47.8%) | 35 (52.2%) | 1,43 | 0.68,3.01 | 0.340 | |
| Hepatic or renal failure | 47 | 19 (40.4%) | 28 (59.6%) | 1,50 | 0.66,3.41 | 0.328 | |
| Other GI disease | 47 | 18 (38.3%) | 29 (61.7%) | 1,72 | 0.76,3.89 | 0.194 | |
| Other diagnosis | 51 | 27 (52.9%) | 24 (47.1%) | — | — | ||
| Department Type | <0.0001 | ||||||
| Non-PCU | 898 | 406 (45.2%) | 492 (54.8%) | 2,07 | 1.59,2.68 | 0.000 | |
| PCU | 433 | 260 (60.0%) | 173 (40.0%) | — | — | ||
| Age | <0.0001 | ||||||
| < 75 years | 278 | 106 (38.1%) | 172 (61.9%) | 2,16 | 1.60,2.91 | 0.000 | |
| ≥ 75 years | 1,053 | 560 (53.2%) | 493 (46.8%) | — | — | ||
| Gender | 0.826 | ||||||
| Male | 643 | 324 (50.4%) | 319 (49.6%) | 0,99 | 0.79-1.23 | 0.904 | |
| Female | 688 | 342 (49.7%) | 346 (50.3%) | — | — | ||
OR: Odds Ratio; CI: Confidence Interval; GI: Gastrointestinal. ** p-value significance level <0.05.
Discussion
Although there are some different baseline characteristics per DRG and age, our study results indicate that patients who died between January 2016 and December 2018, hospitalized in the IEH PCU, had lower direct hospital costs during their last admission than patients with the same level of severity and mortality risk seen admitted in Non-PCU. This difference is clearly significant in both overall and daily costs, with a total daily cost in the PCU of €74.3 compared to €115.8 in Non-PCU. To date, quality end-of-life specialized care had been found to lower cost because of a decrease in the number of admissions, length of hospital stay, a reduction in the number of ICU stays and emergency hospital visits and the cost during last hospital admission.
7
,21
,The effectiveness and cost-effectiveness of hospital-based specialist palliative care for adults with advanced illness and their caregivers.
Cochrane Database of Systematic Reviews. 2020; (Issue 9Art. No.: CD012780.)https://doi.org/10.1002/14651858.CD012780.pub2
22
,30
,31
Increasingly, patients in the last of life with multiple or complex diseases have palliative needs. This patient profile is usually admitted to units with high scientific and technical levels, but with a lack of training in palliative care,
32
which may lead to a use of resources that provide no benefits and incur high costs for the healthcare system.22
For this reason, as it has already been initiated, we consider that knowledge of palliative care should be integrated into the training programs of these specialties.33
, 34
, 35
, - Mercadante S
- Gregoretti C
- Cortegiani A.
Palliative care in intensive care units: why, where, what, who, when, how.
BMC Anesthesiol. 2018; 18https://doi.org/10.1186/s12871-018-0574-9
36
Using all available means to prolong survival of patients under their care, not taking an initial approach that considers the obtainable risk-benefit, depending on the therapeutic effort of each intervention, could be the reason why patients not seen by the PCU have had a slightly longer survival, without achieving an objective benefit.
37
In contrast, the approach focused on the suitability of therapeutic effort by the PCU is based on a clinical approach that is, from the viewpoint of the authors of this research, more appropriate and whose clinical decision-making is based on avoiding using not proportional measures that prolong survival without improving quality of life when cure is not possible, fully applying the provisions of article 36.1 of the Code of Medical Ethics.- Normand C.
- May P.
Measuring Cost-Effectiveness in Palliative Care.
in: MacLeod R. Van den Block L Textbook of Palliative Care. Springer,
Cham2018https://doi.org/10.1007/978-3-319-31738-0_101-1
38
The savings achieved by PC teams during hospital admission are well known.
7
,21
,The effectiveness and cost-effectiveness of hospital-based specialist palliative care for adults with advanced illness and their caregivers.
Cochrane Database of Systematic Reviews. 2020; (Issue 9Art. No.: CD012780.)https://doi.org/10.1002/14651858.CD012780.pub2
31
However, they refer to the cancer population, and focus on reducing hospital stays, days in the ICU, or use of chemotherapy.39
,40
This study included both oncological and non-oncological patients, focusing on the last hospital admission of patients dying in the hospital, showing that savings with PC teams were due to the more appropriate use of hospital resources. Similarly, certain studies show economic benefits with better cost efficiency in non-cancer patients seen by PC teams, but they have again focused on overall reduction of admission costs and not on a cost analysis according to the resources used, as performed in this study.41
,- May P
- Normand C
- Del Fabbro E
- et al.
Economic Analysis of Hospital Palliative Care: Investigating Heterogeneity by Noncancer Diagnoses.
MDM Policy Pract. 2019; 42381468319866451https://doi.org/10.1177/2381468319866451
42
Early palliative care intervention gradually reduces costs in the care of cancer patients with advanced disease, but apparently, during the last stay as good palliative care needs more human resources that other specialties, costs could increase during the last hospital stay of this kind of patients.
30
,43
Our data suggest that palliative care during the last admission optimizes the resources used to care for these patients. It is necessary to check whether this optimization of resource use is associated with an increased perception of quality of life as seen in early care.In contrast to other cost analyses, mention should be made of the homogeneity of the study group, patients at high risk of morbidity and mortality in whom death was expected. This makes it possible to compare the use of hospital resources from a treatment proportionality perspective. Having performed an analysis based on the DRG shows us how resources have been used according to diagnosis and can guide us in future research on decision-making.
Results differ when we breakdown the analysis by diagnosis (DRG) that caused last hospitalization, last admission total cost stills have significant lower cost in patients attended by PCU compared to Non-PCU, but this difference does not prevail when we analyse the cost per day of admission, where it is only maintained for the 470 patients with an Infection or Sepsis DRG (34% of the sample).
The main limitation of this study is to be retrospective and that only one of the researchers revised the clinical records of the patients included, with no double check for the selection. Other limitation is that there may be screening bias in the selected sample because patients with lower expectations for survival had been referred to the PCU. To reduce this potential bias, only patients with a risk of severity and mortality (MOR & SOI) greater than 2 were included in this study, thus confirming that the two study groups had the same risk of mortality, which led to the exclusion of 202 patients from the group of deceased patients not seen by the PCU. This study is also limited because this research was conducted in a single hospital, and although the sample is large, a study could be completed using the same method in all other hospitals in the Autonomous Community to reinforce this hypothesis that PCUs save money at the end of life, without reducing quality of life, even if care is provided in a hospital setting. Such savings and good treatment suitability may be a means of promoting the much-needed development of hospital palliative care teams. Although results obtained are in line with other studies like Morrison et al 2011, we have not perform a propensity score matching between the two groups of study because it would not conform to reality and the outcome evaluated are not related to the effectiveness of the care received.
30
Conclusion
Results show that patients with advanced disease and high short-term risk of death, at the same rate of severity and mortality risk, cost is significantly lower when the patient is treated by a PCU during the last hospital stay when they pass away. Patients under 75 years old, cardiovascular disease DRG and the ones hospitalized in Non-PCU has a higher of having a cost above the median of the study population.
All authors declare that they have no conflicts of interest.
We would like to thank Dr. Ángel Jiménez Rodríguez, Dr. Miguel A. Sánchez González, Dr. Mercedes Molina Robles, Dr. Leopoldo Bárcena Goitiandía, Dr. Suarez Rueda, Dr. Alberto Alonso, Dr. Eduardo Bruera, and Dr. Herrera de la Rosa for their time, their unconditional support, and their solidarity with this project.
References
- Cost of care for elderly cancer patients in the United States.J Natl Cancer Inst. 2008; 100: 630-641https://doi.org/10.1093/jnci/djn103
- Health care costs for patients with cancer at the end of life.J Oncol Pract. 2012; 8: 75s-80shttps://doi.org/10.1200/JOP.2011.000469
- Trends in Medicare payments in the last year of life.N Engl J Med. 1993; 328: 1092-1096https://doi.org/10.1056/NEJM199304153281506
- Palliative Care Teams' Cost-Saving Effect Is Larger For Cancer Patients With Higher Numbers Of Comorbidities.Health Aff (Millwood). 2016; 35: 44-53https://doi.org/10.1377/hlthaff.2015.0752
- Cost, and non-clinical outcomes of palliative care.J Pain Symptom Manage. 2009; 38: 32-44https://doi.org/10.1016/j.jpainsymman.2009.05.001
- Impact of an inpatient palliative care team: a randomized control trial.J Palliat Med. 2008; 11: 180-190https://doi.org/10.1016/j.jpainsymman.2009.05.00110.1089/jpm.2007.0055
- Evidence on the economic value of end-of-life and palliative care interventions: a narrative review of reviews.BMC Palliat Care. 2021; 20: 89https://doi.org/10.1186/s12904-021-00782-7
- Official Position Paper on the Global Promotion of Palliative Care: Recommendations of the PAL-LIFE International Advisory Group of the Pontifical Academy of Life, Vatican City.J Palliat Med. 2018; 21: 1398-1407https://doi.org/10.1089/jpm.2018.0387
- Global atlas of palliative care.World Health Organization, Geneva, Switzerland2020 (Available at) (Accessed May 20, 2022)
- Alleviating the access abyss in palliative care and pain relief-an imperative of universal health coverage: the Lancet Commission report.Lancet. 2018; 391: 1391-1454https://doi.org/10.1016/S0140-6736(17)32513-8
- Strengthening of palliative care as a component of comprehensive care throughout the life course.J Pain Palliat Care Pharmacother. 2014; 28: 130-134https://doi.org/10.3109/15360288.2014.911801
The Economists Intelligence Unit. The 2015 quality of death index: ranking palliative care across the world. Economist Intell Unit 2021. Available at:https://impact.economist.com/perspectives/healthcare/2015-quality-death-index. Accessed March 20, 2022
- EAPC Atlas of Palliative Care in Europe.2013 (Milan. Available at:) (Accessed March 25, 2022)
Osman H, Rihan A, Garralda E, et al. Atlas of palliative care in the eastern Mediterranean Region. 2017. Available at: http://www.thewhpca.org/resources/global-atlas-on-end-of-life-care. Accessed May 19, 2022.
The total expenditure of the public Health System rose by 5.4% in 2019 compared to the previous year. [Government/News/Health]. Available at: https://www.lamoncloa.gob.es/serviciosdeprensa/notasprensa/sanidad14/paginas/2021/080721-gasto-sanitario.aspx. Accessed August 2, 2022.
- Edmonton, Canada: a regional model of palliative care development.J Pain Symptom Manage. 2007; 33: 634-639https://doi.org/10.1016/j.jpainsymman.2007.02.012
- Utilization and costs of the introduction of system-wide palliative care in Alberta, 1993-2000.Palliat Med. 2005; 19: 513-520https://doi.org/10.1191/0269216305pm1071oa
- Palliative care professionals support sedation.Policy and regulations. Diario Médico. 09/13/2017; (Available from:) (Accessed February 15, 2022)
- The last 3 months of life: care, transitions and the place of death of older people.Health Soc Care Community. 2005; 13: 420-430https://doi.org/10.1111/j.1365-2524.2005.00567.x
- The health care cost of palliative care for cancer patients: a systematic review.Support Care Cancer. 2020; 28: 4561-4573https://doi.org/10.1007/s00520-020-05512
- The effectiveness and cost-effectiveness of hospital-based specialist palliative care for adults with advanced illness and their caregivers.Cochrane Database of Systematic Reviews. 2020; (Issue 9Art. No.: CD012780.)https://doi.org/10.1002/14651858.CD012780.pub2
- The health care cost of palliative care for cancer patients: a systematic review.Support Care Cancer. 2020; 28: 4561-4573https://doi.org/10.1007/s00520-020-05512
- Utility of the NECPAL CCOMS-ICO© tool and the Surprise Question as screening tools for early palliative care and to predict mortality in patients with advanced chronic conditions: A cohort study.Palliat Med. 2017; 31: 754-763https://doi.org/10.1177/0269216316676647
- The Severity of Illness Index as a severity adjustment to diagnosis-related groups.Health Care Financ Rev. 1984; : 33-45
- Reliability and Validity of the Severity of Illness Index.Medical Care. 1986; 24: 159-178https://doi.org/10.1097/00005650-198602000-00007
- Validation of the All Patient Refined Diagnosis Related Group (APR-DRG) Risk of Mortality and Severity of Illness Modifiers as a Measure of Perioperative Risk.J Med Syst. 2018; 42: 81https://doi.org/10.1007/s10916-018-0936-3
National Institute of Statistics (INE). Calculation of the Consumer Price Index percentage changes. Available at: https://www.ine.es/varipc/index.do. Accessed March 7, 2021.
ORDER 727/2017, of August 7, of the Regional Minister of Health, by which public prices are set for the provision of services and activities of a healthcare nature in the network of centers of the Community of Madrid. August 21, 2017, Madrid Official State Bulletin. Available at:http://www.madrid.org/wleg_pub/secure/normativas/listadoNormativas.jsf#no-back-button. Accessed February 7, 2021.
Ginard- Alonso, MA. Ruiz-Escobar P, Ruiz Escobar J, Acevedo- Adán MJ. The GRD'S in the National Health System. Ocronos -Scientific-Technical Editorial. Available from:https://revistamedica.com/grds-sistema-nacional-salud/. Accessed February 7, 2022.
- Prospective Cohort Study of Hospital Palliative Care Teams for Inpatients With Advanced Cancer: Earlier Consultation Is Associated With Larger Cost-Saving Effect.J Clin Oncol. 2015; 33: 2745-2752https://doi.org/10.1200/JCO.2014.60.2334
- Cost savings associated with US hospital palliative care consultation programs.Arch Intern Med. 2008; 168: 1783-1790https://doi.org/10.1001/archinte.168.16.1783
- Twelve tips for developing palliative care teaching in an undergraduate curriculum for medical students.Med Teach. 2019; 41: 1359-1365https://doi.org/10.1080/0142159X.2018.1533243
- Palliative Care in Neurology.Mayo Clin Proc. 2017 Oct; 92: 1592-1601https://doi.org/10.1016/j.mayocp.2017.08.003
- Palliative Care Training in Cardiology Fellowship: A National Survey of the Fellows.Am J Hosp Palliat Care. 2018 Feb; 35: 284-292https://doi.org/10.1177/1049909117703728
- Palliative care in intensive care units: why, where, what, who, when, how.BMC Anesthesiol. 2018; 18https://doi.org/10.1186/s12871-018-0574-9
- Palliative care for respiratory disease: An education model of care.Chron Respir Dis. 2018 Feb; 15: 36-40https://doi.org/10.1177/1479972317721562
- Measuring Cost-Effectiveness in Palliative Care.in: MacLeod R. Van den Block L Textbook of Palliative Care. Springer, Cham2018https://doi.org/10.1007/978-3-319-31738-0_101-1
- Central Deontology Commission - Collegiate Medical Organization (Spain).Medical Deontology code. 2018; (version: 10/15/Available at:) (Accessed January 23, 2022)
- End-of-Life Care Interventions: An Economic Analysis.Ont Health Technol Assess Ser. 2014; 14: 1-70
- Cost Savings Associated With Palliative Care Among Older Adults With Advanced Cancer.Am J Hosp Palliat Care. 2021; 38: 1250-1257https://doi.org/10.1177/1049909120986800
- Economic Analysis of Hospital Palliative Care: Investigating Heterogeneity by Noncancer Diagnoses.MDM Policy Pract. 2019; 42381468319866451https://doi.org/10.1177/2381468319866451
- Community-based specialist palliative care is associated with reduced hospital costs for people with non-cancer conditions during the last year of life.BMC Palliat Care. 2017; 16: 68https://doi.org/10.1186/s12904-017-0256-2
- Cost-effectiveness of early palliative care intervention in recurrent platinum-resistant ovarian cancer.Gynecol Oncol. 2013; 130: 426-430https://doi.org/10.1016/j.ygyno.2013.06.011
Article info
Publication history
Published online: July 13, 2022
Accepted:
June 29,
2022
Identification
Copyright
© 2022 The Authors. Published by Elsevier Inc. on behalf of American Academy of Hospice and Palliative Medicine.
User license
Creative Commons Attribution – NonCommercial – NoDerivs (CC BY-NC-ND 4.0) | How you can reuse 
Elsevier's open access license policy
Creative Commons Attribution – NonCommercial – NoDerivs (CC BY-NC-ND 4.0)
Permitted
For non-commercial purposes:
- Read, print & download
- Redistribute or republish the final article
- Text & data mine
- Translate the article (private use only, not for distribution)
- Reuse portions or extracts from the article in other works
Not Permitted
- Sell or re-use for commercial purposes
- Distribute translations or adaptations of the article
Elsevier's open access license policy
