Anatomopathological Causes of Death in Patients with Advanced Cancer: Association with the Use of Anticoagulation and Antibiotics at the End of Life

Abstract

Background:

Anatomopathological studies that described the immediate causes of death of patients with advanced cancer were first published approximately 20 years ago.

Objective:

Our objective was to analyze if causes of death changed with a wider use of broad spectrum antibiotics and prophylactic anticoagulation.

Methods:

We conducted a retrospective study of all patients with an advanced cancer hospitalized in the Division of Palliative Medicine at the University Hospital Geneva from 2004 to 2010 who had an autopsy.

Results:

Two hundred forty patients were included (130 men, mean age: 74±13). Main causes of death discovered at the autopsy were pulmonary infection (n=131; 55%), advanced cancer (n=39; 16%), pulmonary infection together with pulmonary embolism (PE) (n=27; 12%), PE alone (n=22; 9%), cardiac complications (n=19; 5%) and others (n=2; 1%). In a logistic regression model, with adjusting for age, gender, main diagnosis, comorbidities, blood count, corticosteroids, and antibiotics, there were no independent factors associated with pulmonary infection at autopsy. In a similar model, with adjusting for age, gender, main diagnosis, comorbidities, and anticoagulation, the only independent factor associated with PE at autopsy was the history of thrombo-embolic disease and therapeutic anticoagulation.

Conclusion:

The results of this retrospective study demonstrate that causes of death did not change with the modification of our practice. The high rate of pulmonary infection and embolism in this population, including in patients who received broad spectrum and prophylactic anticoagulation should encourage us to pursue other prospective studies to actually demonstrate the benefit of these treatments in this population.

Introduction

Anatomopathological studies that described the immediate causes of death of patients with advanced cancer were first published approximately 20 years ago. Main causes of death in these studies were infections (33%–54%) and cardiovascular (19%–21%) and thrombo-embolic events (8%–20%).^1–4 However, since then, treatment practice and recommendations have changed in palliative medicine. Broad spectrum antibiotics and in particular parenteral agents as well as low-molecular-weight heparins are more widely prescribed in hospitalized patients with advanced stage cancer. The use of these broad spectrum antibiotics is not harmless. Not only are they responsible for drug toxicities and a heavy burden on health care costs, but patients receiving antimicrobials are also exposed to additional stress due to the way of administration of the treatment (intravenous or subcutaneous injection, oral supplementary drug). Furthermore, they have an increased risk of subsequent antimicrobial resistance that not only affects the patients who are receiving treatment, but also presents a general public health concern. As to the benefit of preventive anticoagulation, in particular the use of low-molecular-weight heparin, recommendation is not clearly defined in this population especially at risk of complications, in particular bleeding.⁵

Our aims are therefore to describe the anatomical-pathological main causes of death in patients who died with an advanced cancer and to measure the association between pulmonary infection and the prescription of antibiotics and between pulmonary embolism and the prescription of prophylactic anticoagulation.

Methods

We conducted a retrospective chart review of all patients with an advanced cancer hospitalized in the Division of Palliative Medicine at the University Hospital Geneva from 2004 to 2010 who had an autopsy without any exclusion criteria.

Collected data included demographic data, type of cancer, comorbidities, and history of thrombo-embolic disease. Episodes of infections and venous thrombo-embolism (radiologically confirmed or not) that occurred during the last 2 weeks of life were collected. Bleeding complications were also noted.

The last blood tests (complete blood count, C-reactive protein, renal function and, microbiological records) made during the last 2 weeks of life of the patients were collected. Specific medications—antibiotics, corticosteroids, and prophylactic and therapeutic anticoagulation (fundaparinux, enoxaparin, acenocoumarol)—and their indications were noted.

The autopsy procedure authorized by the patient's family included an extended autopsy (brain included), with a detailed macro- and microscopic examination of major organs. In each case, the immediate cause of death, for example, the disease, injury, or complication that directly preceded death, was determined. The tumor burden as well as incidental findings were noted. The presence of PE and pulmonary infection were systematically noted.

The association between antibiotics and infection on one side, and venous thrombo-embolism and anticoagulation on the other were compared using either the χ² test or Fisher's exact test when appropriate. Logistic regression was used to predict the presence of an infection or a PE after adjustment for age, gender, type of cancer, and main comorbidities. All analyses were performed with the Stata 12 statistical package (StataCorp., College Station, TX).

Study protocol was approved by the Central Ethic Committee of the University Hospital of Geneva.

Results

Two hundred forty patients were included. Main causes of death at autopsy were pulmonary infection (n=131; 55%), advanced cancer (n=39; 16%), pulmonary infection together with PE (n=27; 12%), PE alone (n=22; 9%), cardiac complications (n=19; 5%), and others (n=2; 1%). One patient had acute appendicitis, one acute pancreatitis, one mesenteric ischemia, and one acute pyelonephritis. Three patients had pulmonary aspergillosis that was not diagnosed before death.

Main characteristics of the patients are shown in Table 1. Characteristics of the patients included in this study were very similar to the characteristics of the patients included in two studies conducted in our palliative care units.^6,7 Patients had mean±standard deviation (SD) 2.3±2.6 years of history of their oncological disease. During their last 2 weeks of life, 136 (57%) patients presented a clinical suspicion of infections such as pulmonary (n=97; 71%), urinary (n=14; 10%), digestive (n=7; 6%), bacteremia (n=6; 4%), and of unknown origin (n=6; 4%). Two hundred one patients had a blood sample during this period. Results demonstrated a C-reactive protein higher than 10 in 186 patients; a white blood cell count >10 g/L in 75 patients, and granulocytes >55% of white blood cells in 81 patients. Seventy-six patients received antibiotics (Table 2). Fifty-five patients received parenteral antibiotics. Eleven patients received two different antibiotics during their last 2 weeks of life; one hundred thirty-eight received corticosteroids during this period.

Table 1.

Characteristics of Patients

	Total N=240	Pulmonary infection and embolism at autopsy N=27	Pulmonary embolism at autopsy N=22	Pulmonary infection at autopsy N=131	Other causes of death N=60
Age (mean±SD)	74±13	74±10	74±10.	73±11	76±9
Gender (male) (n)	130	15 (56)	9 (41)	77 (59)	29 (48)
Main cancer (n; %)
Gastrointestinal tract	75 (31)	6 (22)	12 (55)	42 (32)	15 (25)
Respiratory system	57 (24)	7 (26)	3 (14)	38 (29)	9 (15)
Genitourinary tract	45 (20)	10 (37)	2 (9)	27 (21)	6 (10)
Breast	21 (9)	1 (4)	1 (5)	5 (4)	14 (23)
Hematological	10 (4)	2 (7)	1 (5)	4 (3)	3 (5)
Other	32 (12)	2 (7)	3 (14)	15 (12)	12 (20)
Main comorbidities
High blood pressure	86 (36)	9 (33)	9 (41)	41 (31)	27 (45)
Cardiac disease	74 (31)	9 (33)	5 (23)	38 (29)	22 (37)
COPD	26 (11)	0	1 (5)	16 (12)	9 (15)
Diabetes	23 (10)	3 (11)	2 (9)	9 (7)	9 (15)

COPD, chronic obstructive pulmonary disease; SD, standard deviation.

Table 2.

Anticoagulation and Antibiotics during the Two Last Weeks

	Total N=240	Pulmonary infection and embolism at autopsy N=27	Pulmonary embolism at autopsy N=22	Pulmonary infection at autopsy N=131	Other causes of death N=60
Anticoagulation (n; %)
Therapeutic	30 (13)	7 (26)	5 (23)	12 (9)	6 (10)
Low-molecular-weight heparin	15 (50)
Acenocoumarol	8 (27)
Heparin	7 (23)
Prophylactic	103 (43)	12 (44)	8 (36)	59 (45)	24 (40)
Low-molecular-weight heparin	95 (94)
Heparin	6 (6)
Antibiotics (n; %)
Co-amoxicilline	10 (4)	2 (7)	2 (9)	6 (5)	0
Piperacilline-tazobactam	20 (8)	1 (4)	1 (5)	15 (12)	3 (5)
Cephalosporin	24 (10)	4 (15)	2 (9)	12 (9)	6 (10)
Fluoroquinolone	17 (7)	2 (7)	3 (14)	9 (7)	3 (5)
Carbapenem	5 (2)	0	0	5 (4)	0
Metronidazole	5 (2)	1 (4)	1 (5)	3 (3)	0
Others	6 (3)	0	0	4 (3)	0

During the same period 30 (13%) patients had therapeutic anticoagulation: 13 for cerebrovascular disease, 11 for thrombo-embolic disease, 6 for recent PE (4 radiologically confirmed). One hundred three patients had prophylactic anticoagulation prescribed. Concerning bleeding complications, 2 patients presented hemoptysis.

Results concerning risk factors showed that 18 (8%) patients had hormone therapy prescribed for the treatment of cancer, and 3 (1%) patients had recent surgery. Forty-nine (20%) had a history of thrombo-embolism since their diagnosis of cancer. Nineteen (38%) of the 49 patients with PE at autopsy had received prophylactic anticoagulation. There was no significant statistical association between the two (p>0.05).

Nine patients (47%) who died of a cardiac complication were known to have cardiac comorbidities, eight (42%) had a history of high blood pressure.

In a logistic regression model, with adjusting for age, gender, main diagnosis, comorbidities, blood count, corticosteroids, and antibiotics, there was no independent factor associated with pulmonary infection as diagnosed at autopsy (see Table 3). There was as well a trend for high blood pressure, diabetes, and leucocytosis. In a similar model, with adjusting for age, gender, main diagnosis, comorbidities, and anticoagulation, the only independent factor associated with PE at autopsy was the history of thrombo-embolic disease and being treated with therapeutic anticoagulation (see Table 4).

Table 3.

Predictor of Pulmonary Infection at Autopsy

	Crude OR (95% CI)	P	Adjusted OR (95% CI)	P
Age	0.99 (0.97–1.01)	0.49	1.01 (0.98–1.04)	0.55
Gender	1.61 (0.94–2.76)	0.08	2.04 (1.00–4.16)	0.05
Main diagnosis
Respiratory tract	0.66 (0.30–1.46)	0.30	1.04 (0.40–2.71)	0.94
Gastrointestinal system	0.44 (0.11–1.81)	0.26	1.63 (0.22–11.92)	0.63
Genitourinary tract	0.50 (0.19–1.36)	0.18	0.68 (0.20–2.34)	0.54
Other	0.88 (0.33–2.33)	0.80	1.67 (0.49–5.71)	0.42
Main comorbidities
High blood pressure	0.86 (0.49–1.53)	0.61	0.78 (0.37–1.67)	0.52
Cardiac disease	0.55 (0.32–0.95)	0.03	0.49 (0.24–1.01)	0.05
COPD	0.81 (0.35–1.88)	0.63	0.92 (0.30–2.79)	0.89
Diabetes	0.53 (0.22–1.26)	0.15	0.38 (0.14–1.05)	0.06
CRP	1.00 (0.99–1.00)	0.06	1.00 (0.99–1.00)	0.17
WBC	0.59 (0.34–1.02)	0.06	0.94 (0.88–1.01)	0.08
Granulocytosis	0.95 (0.91–0.99)	0.02	0.72 (0.32–1.59)	0.42
Corticsteroids	1.59 (0.93–2.72)	0.49	1.26 (0.62–2.55)	0.53
Antibiotics	1.09 (0.61–1.93)	0.08	0.89 (0.42–1.91)	0.77

CI, confidence interval; COPD, chronic obstructive pulmonary disease; CRP, C-reative protein; OR, odds ratio; WBC, white blood cell.

Table 4.

Predictor of Pulmonary Embolism at Autopsy

	Crude OR (95% CI)	P	Adjusted OR(95% CI)	P
Age	1.00 (0.98–1.03)	0.96	1.01 (0.98–1.03)	0.65
Gender	0.77 (0.41–1.44)	0.42	0.81 (0.39–1.68)	0.58
Previous pulmonary embolism	2.05 (1.06–3.99)	0.03	4.55 (1.51–13.66)	0.01
Main diagnosis
Respiratory tract	1.71 (0.73–4.01)	0.22	1.29 (0.52–3.23)	0.58
Gastrointestinal system	0.86 (0.34–2.20)	0.76	0.73 (0.23–2.26)	0.58
Genitourinary tract	2.01 (0.44–9.16)	0.37	2.26 (0.45–11.34)	0.32
Others	1.07 (0.33–3.50)	0.91	0.74 (0.20–2.68)	0.64
Main comorbidities
High blood pressure	0.87 (0.44–1.749)	0.70	0.92 (0.40–2.09)	0.84
Cardiac disease	1.05 (0.55–2.02)	0.88	1.03 (0.48–2.20)	0.94
COPD	0.14 (0.02–1.05)	0.06	0.12 (0.01–1.09)	0.06
Diabetes	1.09 (3.10–0.87)	0.38	0.99 (0.31–3.13)	0.87
Therapeutic anticoagulation	3.37 (1.39–8.19)	0.01	2.88 (1.02–8.15)	0.05
Prophylactic anticoagulation	1.17 (0.58–2.38)	0.66	1.26 (0.58–2.74)	0.56

CI, confidence interval; COPD, chronic obstructive pulmonary disease; OR, odds ratio.

Discussion

In our study, infection, especially pulmonary infection, remains a major complication in patients with advanced malignant disease, especially bronchopneumonia. There are many factors predisposing to infection in this patient population, including local factors due to a tumor, deficiencies in host defense mechanisms secondary to cancer and/or antitumor treatment, and the presence of invasive devices.⁸ Antimicrobial therapy is frequently viewed as usual care and not as an ‘‘aggressive’’ treatment.⁹ For clinicians, it is often easier and handier to prescribe antibiotic therapy than to withhold the treatment, and they may feel compelled to offer antimicrobial treatment for possible infection even among patients in whom other aggressive therapies would generally not be appropriate.

But whether patients do really benefit from this therapy still remains an open question. On the one hand, infection during the course of an advanced malignant disease has an incidence on mortality. A prospective single cohort study demonstrated that antibiotic therapy increases survival in two-thirds of patients in which sepsis or organ-related infection occurred.¹⁰ On the other hand, improvement of symptoms due to infection is not guaranteed by antibiotic therapy. The other third of the 66 patients in the study receiving antibiotics were considered as poor responders and their survival was not influenced by antibiotic therapy. Furthermore, the empiric use of broad spectrum antibiotics among hospitalized patients close to death is discussed in different retrospective, but also prospective studies.^11–14 One of them evaluated the effect of antibiotic treatment on infection-related symptoms in 36 patients with advanced cancer. Patients described a significant improvement in dysuria and physicians in cough of the patients; however, the global improvement of symptoms was only 48%. One quarter of the patients died within one week of antibiotic administration.¹⁵ In another study, all patients receiving parenteral antibiotics in a palliative care unit were prospectively monitored over a 13-months period. Of 913 consecutive admissions, 41 patients received 43 courses of parenteral antibiotics. On 27 of 43 occasions, the use of parenteral antibiotics was considered helpful (62%); in eight cases it was considered unhelpful (19%) and in a further eight cases the outcome could not be assessed (19%).¹⁶ Taking into account the results of our study, we found that among the 87 patients who had received antibiotics in their last 2 weeks of life, 64 of them still died of pneumonia, which means that probably the treatment didn't prolong the survival of the patients. We should probably initially try to control symptoms with other means than through administration of antibiotics. This is compounded by the numerous adverse drug events and the microbial resistances that we create by administrating a futile treatment. However, the difficulty does not lie in the decision to administer or not administer an antibiotic, but to estimate when will be the “point of no return,” for example the last 2 weeks of life have been reached and such a treatment seems to become useless. Personal experience, team work, and further research may lighten up this zone of transition. But in summary, we should consider more often that pneumonia remains, despite our endless efforts, “an old man's best friend.” The empirical indication to prescribe antibiotics in patients with advanced cancer for an infection is complicated by the fact that in many cases fever is not the result of an infection.¹⁷

The association between cancer and venous thrombo-embolic disease (VTE) was first described by Trousseau 140 years ago.¹⁸ Cancer type and extension can facilitate VTE together with other risk factors such as bed confinement, anticancer treatments, central venous devices, and advanced age.^19–25 Meta-analyses have demonstrated that the relative risk of VTE is increased approximately sixfold in patients with cancer compared with age- and sex-matched controls without cancer.²⁶ Johnson and colleagues screened 287 patients on admission to a specialist palliative care unit using light reflection rheography, demonstrating evidence of likely deep venous thrombosis (DVT) in 52% of patients.²⁷ Moreover, in 17% of these cases, bilateral DVTs were observed. Previous studies have shown that fatal PE results in abrupt death in only a minority of cases (<10%).²⁸ However, both DVT and PE are often associated with important symptomatology in patients with cancer in the palliative care setting.²⁷ In the study of Johnson and colleagues, only 9% of likely venous thromboses were symptomatic at time of original diagnosis, but another 32% of patients with likely DVT subsequently developed symptoms that included significant lower limb pain and swelling.²⁷ However in patients with cancer, abnormal renal function, metastatic disease, recent major bleeding, and recent immobility for more than 4 days are associated with an increased risk for both fatal PE and fatal bleeding.²⁹ In palliative care for cancer patients, primary prophylaxis of VTE has not been proven useful despite the presence of many risk factors in most patients. Decision to administer VTE prophylaxis may thus depend on health professional experience, place of care, and patient preferences.³⁰ Widely prescribed in acute hospitals, VTE prophylaxis is rarely administered in nursing homes and even less frequently at home, although risk factors are not different for these patients.³¹ In a retrospective review of 75 patients admitted to a hospice in Scotland for symptom control, respite, or rehabilitation, all patients had at least one risk factor for VTE, and 81% had three risk factors.³² Nevertheless, only one was prescribed prophylactic low molecular weight hepanin (LMWH). Reasons for avoidance of prophylaxis were documented in one-third of cases and included bleeding (13 patients), anemia (4 patients), and a dying patient (1 patient). A second retrospective analysis of 71 patients admitted to a palliative care unit and diagnosed with VTE revealed that only 10 patients (23.2%) had received thromboprophylaxis prior to admission. All patients had at least one risk factor in addition to cancer, including surgery and immobilization. VTE related death occurred in 11 patients (15.5%).³³

One open randomized study (prophylactic nadroparin versus placebo) conducted in our palliative medicine division included 20 patients with cancer and the aim was to determine utility of prophylactic anticoagulation. One venous thrombo-embolism and one major bleeding occurred in the group receiving nadroparin, whereas two minor bleedings occurred in the control group. At 3 months, 9 of 10 participants had died in the control group versus 5 of 10 in the group receiving nadroparin (p=0.141).³⁴

Comparing our anatomopathological results with the above numbers we find a VTE-related death in 49 patients (20 %), which is higher than described in the literature. Taking also into account those who were treated therapeutically and still died of VTE, the percentage is even higher at 24.1%. This population therefore could benefit from anticoagulation to reduce the symptoms that are associated with VTE. These results must be confirmed by studies conducted in other settings. Whether the remaining patients need a prophylactic anticoagulation close to the end of life remains open. Guidelines suggest a prophylaxis in patients with advanced cancer who still have a good performance status; however, a consensus also exists that in the terminal phase, or when the Karnofsky Performance Scale score is less than 10, VTE prophylaxis can be withdrawn.³¹ The availability of new treatments such as rivaroxaban could also change the recommendations during these next years.^13,35

Footnotes

Author Disclosure Statement

No competing financial interests exist.

References

Ambrus

, Ambrus

, Mink

, Pickren

. Causes of death in cancer patients. J Med, 1975; 6:61–64.

Gross

, Neufeld

, Libow

, Gerber

, Rodstein

. Autopsy study of the elderly institutionalized patient. Review of 234 autopsies. Arch Intern Med, 1988; 148:173–176.

Svendsen

, Karwinski

. Prevalence of pulmonary embolism at necropsy in patients with cancer. J Clin Pathol, 1989; 42:805–809.

Gee

. Causes of death in a hospitalized geriatric population: An autopsy study of 3000 patients. Virchows Arch A Pathol Anat Histopathol, 1993; 423:343–349.

Merminod

, Zulian

. Diagnosis of venous thromboembolism in cancer patients receiving palliative care. J Pain Symptom Manage, 2000; 19:238–239.

Pautex

, Herrmann

, Zulian

. Role of advance directives in palliative care units: A prospective study. Palliat Med, 2008; 22:835–841.

Pautex

, Moynier-Vantieghern

, Herrmann

, Zulian

. State of consciousness during the last days of life in patients receiving palliative care. J Pain Symptom Manage, 2009; 38:e1–e3.

Bodey

. Infection in cancer patients. A continuing association. Am J Med, 1986; 81:11–26.

Lam

, Chan

, Tse

, Leung

. Retrospective analysis of antibiotic use and survival in advanced cancer patients with infections. J Pain Symptom Manage, 2005; 30:536–543.

10.

Thai

, Lau

, Wolch

, Yang

, Quan

, Fassbender

. Impact of infections on the survival of hospitalized advanced cancer patients. J Pain Symptom Manage, 2012; 43:549–557.

11.

Chun

, Rodgers

, Vitale

, Collins

, Malani

. Antimicrobial use among patients receiving palliative care consultation. Am J Hosp Palliat Care, 2010; 27:261–265.

12.

, Kim

, Im

, Kim

, Heo

, Bang

, Kim

. Antibiotic use during the last days of life in cancer patients. Eur J Cancer Care (Engl), 2006; 15:74–79.

13.

Niessner

. Oral treatment of acute pulmonary embolism with a fixed dose of rivaroxaban is non-inferior to standard treatment. Evid Based Med, 2013; 18:29–30.

14.

Claret

, Bobbia

, de la Coussaye

. Oral rivaroxaban for pulmonary embolism. N Engl J Med, 2012; 366:2526author reply 2526–2527.

15.

Mirhosseini

, Oneschuk

, Hunter

, Hanson

, Quan

, Amigo

. The role of antibiotics in the management of infection-related symptoms in advanced cancer patients. J Palliat Care, 2006; 22:69–74.

16.

Clayton

, Fardell

, Hutton-Potts

, Webb

, Chye

. Parenteral antibiotics in a palliative care unit: Prospective analysis of current practice. Palliat Med, 2003; 17:44–48.

17.

Homsi

, Walsh

, Panta

, Lagman

, Nelson

, Longworth

. Infectious complications of advanced cancer. Support Care Cancer, 2000; 8:487–492.

18.

Trousseau

. Phlegmasia Alba Dolens. Society

TNS

. Clinique médicale de l'Hôtel Dieu de Paris, 3. London186594.

19.

Baron

, Gridley

, Weiderpass

, Nyren

, Linet

. Venous thromboembolism and cancer. Lancet, 1998; 351:1077–1080.

20.

Deitcher

, Gomes

. The risk of venous thromboembolic disease associated with adjuvant hormone therapy for breast carcinoma: A systematic review. Cancer, 2004; 101:439–449.

21.

Goodnough

, Saito

, Manni

, Jones

, Pearson

. Increased incidence of thromboembolism in stage IV breast cancer patients treated with a five-drug chemotherapy regimen. A study of 159 patients. Cancer, 1984; 54:1264–1268.

22.

Gouin-Thibaut

, Samama

. [Venous thrombosis and cancer] Ann Biol Clin (Paris), 2000; 58:675–682.

23.

Jacobs

. Prophylactic anticoagulation for venous thromboembolic disease in geriatric patients. J Am Geriatr Soc, 2003; 51:1472–1478.

24.

Lee

. Epidemiology and management of venous thromboembolism in patients with cancer. Thromb Res, 2003; 110:167–172.

25.

Saphner

, Tormey

, Gray

. Venous and arterial thrombosis in patients who received adjuvant therapy for breast cancer. J Clin Oncol, 1991; 9:286–294.

26.

Cunningham

, White

, O'Donnell

. Prevention and management of venous thromboembolism in people with cancer: A review of the evidence. Clin Oncol (R Coll Radiol), 2006; 18:145–151.

27.

Johnson

, Sproule

, Paul

. The prevalence and associated variables of deep venous thrombosis in patients with advanced cancer. Clin Oncol (R Coll Radiol), 1999; 11:105–110.

28.

Havig

. Deep vein thrombosis and pulmonary embolism. An autopsy study with multiple regression analysis of possible risk factors. Acta Chir Scand Suppl, 1977; 478:1–120.

29.

Monreal

, Falga

, Valdes

, Suarez

, Gabriel

, Tolosa

, Montes

. Fatal pulmonary embolism and fatal bleeding in cancer patients with venous thromboembolism: Findings from the RIETE registry. J Thromb Haemost, 2006; 4:1950–1956.

30.

Johnson

, Sherry

. How do palliative physicians manage venous thromboembolism? Palliat Med, 1997; 11:462–468.

31.

McLean

, Ryan

, O'Donnell

. Primary thromboprophylaxis in the palliative care setting: a qualitative systematic review. Palliat Med, 2010; 24:386–395.

32.

Brabin

, Douglas

, Mackay

. The role of primary thromboprophylaxis in hospice patients: A retrospective review of current practice. 22 Palliat Med, 2008; 569–603.

33.

Soto-Cardenas

, Pelayo-Garcia

, Rodriguez-Camacho

, Segura-Fernandez

, Mogollo-Galvan

, Giron-Gonzalez

. Venous thromboembolism in patients with advanced cancer under palliative care: Additional risk factors, primary/secondary prophylaxis and complications observed under normal clinical practice. Palliat Med, 2008; 22:965–968.

34.

Weber

, Merminod

, Herrmann

, Zulian

. Prophylactic anti-coagulation in cancer palliative care: A prospective randomised study. Support Care Cancer, 2008; 16:847–852.

35.

Cios

, Fanikos

. Rivaroxaban to prevent pulmonary embolism after hip or knee replacement. Circulation, 2012; 125:e542–e544.