Management of Terminal Hemorrhage in Patients With Advanced Cancer: A Systematic Literature Review

Article Outline

• Abstract
• Introduction
• Method
• Results
  • Definition of Terms
  • Incidence of Terminal Hemorrhage
  • Types of Hemorrhage
  • Management of Terminal Hemorrhage
    • Identifying Patients at Risk
    • General Supportive Measures
    • Use of Sedative Medication
    • Ethical Issues
• Discussion
• Recommendations
• Conclusion
• References
• Copyright

Abstract 

Although terminal hemorrhage is an infrequent occurrence in advanced cancer patients, it is extremely distressing for patients, their families, and health care professionals when it does occur. By definition, there is a very short time period to support and comfort the patient, and it is vital that the management approach follows the best available evidence base. A systematic literature search was carried out to retrieve relevant publications relating to the management of terminal hemorrhage in patients with advanced cancer in whom invasive or interventional procedures are no longer appropriate. From 3,564 initial citations, 18 were appropriate to include in the final review, many of which focused on patients with head and neck tumors. The reported incidence of significant bleeding in patients with advanced cancer is 6%–14% and incidence of terminal hemorrhage 3%–12%. Key areas arising from the literature were 1) identifying patients at risk, 2) general supportive measures to use, and 3) use of sedative medication. General supportive measures included use of dark towels to camouflage blood loss, use of suction, and applying external pressure. There was variation in the recommended sedative medication (drug, dose, and route). Drugs recommended included diazepam, midazolam, diamorphine, and ketamine at varying doses and routes of administration. Current guidelines are based completely on isolated case reports and expert opinion. Clinical research is needed in this area but is difficult because of practical and ethical limitations.

Key Words: Bleeding, cancer, carotid blowout, crisis medication, terminal hemorrhage

Back to Article Outline

Introduction 

A uniformly accepted definition of terminal hemorrhage does not exist, but it may be considered to be a major hemorrhage, from an artery, which is likely to result in death within a period of time that may be as short as minutes, because of the rapid internal or external loss of circulating blood volume.

It is a rare situation in palliative medicine but has a major impact on patients, their relatives and carers, and health care professionals when it does occur, particularly if the source of bleeding is externally visible, such as a carotid artery rupture. The “palliative” management of patients in this context refers to the active supportive care of patients with advanced cancer in whom active invasive and cardiopulmonary resuscitative measures are no longer appropriate.

By definition, there is a very short period of time to support and comfort the patient in this situation and it is, therefore, vital that the management approach follows the best available evidence.

Back to Article Outline

Method 

Systematic searches of the electronic databases were carried out: The Cochrane Library (2008), EMBASE (1980–2008), Ovid MEDLINE (1966–2008), CINAHL (Cumulative Index to Nursing and Allied Health Literature 1982–2008), and SIGLE (System for Information on Grey Literature in Europe 1980–2008). To complement this, we also hand-searched reference lists of the retrieved articles for other potentially appropriate articles that had not been identified by the literature search¹, ², ³, ⁴ and standard palliative care and oncology textbooks.⁵, ⁶, ⁷, ⁸, ⁹, ¹⁰, ¹¹, ¹² The search strategy used key words and medical subject headings appropriate to each database, which included combinations of “bleeding,” “carotid artery blow out,” “carotid artery rupture,” “emergencies,” “haematemesis,” “hematemesis,” “haemoptysis,” “hemoptysis,” “haemorrhage,” “hemorrhage,” “palliative,” “sedation,” and “terminal.”

This search retrieved 3,564 citations. The authors reviewed the title and abstract of all these citations, removed duplicate sources, and then applied the inclusion and exclusion criteria (Table 1) to retrieve 40 potentially relevant articles, of which 18 were appropriate to be included in the final review.¹, ¹³, ¹⁴, ¹⁵, ¹⁶, ¹⁷, ¹⁸, ¹⁹, ²⁰, ²¹, ²², ²³, ²⁴, ²⁵, ²⁶, ²⁷, ²⁸, ²⁹ A further 13 of the retrieved 40 articles², ³, ⁴, ³⁰, ³¹, ³², ³³, ³⁴, ³⁵, ³⁶, ³⁷, ³⁸, ³⁹ were used to source background information about the topic (e.g., incidence and types of hemorrhage) for the introductory sections of the review to enable the subsequent focused discussion about supportive management of terminal hemorrhage to be viewed in its wider context.

Table 1. Inclusion and Exclusion Criteria

We also searched the World Wide Web using the Google search engine for guidelines on the management of terminal hemorrhage from individual health care organizations and institutions, using combinations of the same search terms.⁴⁰, ⁴¹, ⁴², ⁴³, ⁴⁴, ⁴⁵

Back to Article Outline

Results 

Most of the published literature on this subject primarily focused on active invasive treatment, such as the advantages and disadvantages of various methods of controlling carotid artery bleeding (surgical ligation or endovascular stents). Significantly, none of the current published literature on the palliative management of terminal hemorrhage is above Level 5 of the Oxford Centre for Evidence-Based Medicine Levels of Evidence (Table 2). It was also noted that there was an overlap in the authorship of some of the key references and a large overlap in the reference lists between different sources.

Table 2. Oxford Centre for Evidence-Based Medicine Levels of Evidence

Reproduced with permission from the Oxford Centre for Evidence-Based Medicine.⁴⁹ Available at: http://www.cebm.net/index.aspx?o=1025.

SR=systematic review; RCTs=randomized controlled trials; CDR=clinical decision rule; Absolute SpPin= a diagnostic finding whose Specificity is so high that a Positive result rules in the diagnosis; Absolute SnNout=a diagnostic finding whose Sensitivity is so high that a Negative result rules out the diagnosis.

The issues and themes that were identified within the literature were broadly grouped, by the authors, into the following categories: 1) identifying patients at risk of terminal hemorrhage, 2) general supportive management of terminal hemorrhage, 3) use of sedative medication in the management of terminal hemorrhage, and 4) ethical issues in the management of terminal hemorrhage.

To place the literature review in context, an introductory appraisal of the epidemiology of terminal hemorrhage was also included to discuss the reported incidence and types of terminal hemorrhage to which the literature refers.

Definition of Terms 

Before giving a detailed account of the literature, it is important to clarify the terms, such as “terminal hemorrhage” and “crisis medication.” Many of the sources of literature do not clarify these definitions. Only one author gave a definition of terminal hemorrhage: “Major arterial hemorrhage from a patient in whom active treatment is not appropriate or possible and which will inevitably cause death in minutes. Loss of more than 1.5L in 30 seconds.”⁶ Other authors acknowledge the difficulty in defining terminal hemorrhage and suggest that the objective of management is to treat the distress of the patient and not a certain quantity of blood.²⁸

For the purposes of this review, we have defined terminal hemorrhage as a major hemorrhage, from an artery, which is likely to result in death within a period of time that may be as short as minutes, because of the rapid internal or external loss of circulating blood volume. This review will, therefore, not cover small volume, yet fatal bleeds into critical areas, such as the brain. Furthermore, it is recognized that the definitive diagnosis of terminal hemorrhage can only truly be made in retrospect, because many at-risk patients may have a sentinel bleed but do not progress to a terminal hemorrhagic event.

The term “sentinel” or “herald” bleed is used to refer to “small” prodromal bleeding occurring 24–48 hours before the rupture of an artery that either resolves spontaneously or with packaging or pressure.¹⁹, ³³, ³⁴, ⁴⁰

Finally, the term “crisis” or “emergency” medication was used to describe medication given urgently once a hemorrhage has started, with the aim of relieving the anxiety and distress of the patient, for example, midazolam by intravenous (IV) injection.

There is widespread recognition that terminal hemorrhage, although rare, is distressing for all involved when it does occur. Terms, such as “dramatic,”¹⁴, ²⁸ “feared,”²⁷, ³³, ³⁹ “horrific,”⁴¹ “dreaded,”³⁰, ⁴⁰ “traumatic,”⁴⁰ “catastrophic,”²⁴ and “distressing,”³⁶ are frequently mentioned in the literature and that, for the patients, “their distress is overwhelming because of the imminence of death.”²⁸

Incidence of Terminal Hemorrhage 

There is no consensus figure for the incidence of terminal hemorrhage in patients with advanced cancer. The current literature regarding the incidence and types of terminal hemorrhage is descriptive, based on Level 4 and 5 evidence.

Overall, the reported incidence of significant bleeding in patients with advanced cancer is 6%–14%.²⁰, ²⁴, ²⁵, ²⁶, ³⁶ The reported incidence of terminal hemorrhage in patients with advanced cancer was between 3% and 12%.¹, ¹⁴, ¹⁹, ²⁵, ²⁹, ³¹, ⁴⁰ Most figures quoted in the medical literature actually relate to patients with head and neck cancer rather than the cancer population as whole.¹, ¹⁴, ¹⁹, ²⁹, ³¹, ³⁹, ⁴⁰ Presumably, this may be because hemorrhage, such as carotid artery rupture, is most common in this group (which in turn could be explained by a higher incidence of proximity of head and neck cancers to major blood vessels), but there are no published figures of death from hemorrhage subdivided by cancer type to support this presumption.

Possible explanations for the wide variation in the aforementioned figures could relate to the often small sample size in case series and lack of clarity of definitions: terms, such as “catastrophic hemorrhage” and “terminal hemorrhage,” were used interchangeably. We would question whether some of the “catastrophic” events did not result in immediate death, that is, terminal. For practical purposes, any major (or catastrophic) bleed should be managed in the same way, as there is no way of knowing which will be terminal events until death has actually occurred.

A further explanation for the heterogeneity of results arises from the fact that most of the case series referred to hospital patients and were constructed by retrospectively analyzing hospital records, which potentially may skew the data by excluding patients in other settings (such as at home or in a hospice). For example, basing figures for the incidence and mortality from carotid artery rupture in head and neck cancer patients on those patients presenting to a hospital with carotid artery rupture excludes those who died at home before help arrived, in whom it was inappropriate to transfer from home or died from hemorrhage in a nearby hospice without being transferred.

Types of Hemorrhage 

Hemorrhage in patients with advanced cancer is reported to occur at the site of the malignancy in most cases, for example, massive hemoptysis from bronchial tumors invading intrapulmonary arteries or the aorta or carotid artery rupture from head and neck tumors.²⁰, ²³, ²⁵, ²⁸, ³⁶, ³⁷ The carotid artery, femoral artery, and intrapulmonary vessels are reported to be the most common sites,¹² but there are no specific figures comparing the incidence of terminal hemorrhage between different sites of bleeding, and most of the current medical literature refers specifically to carotid artery rupture (or “blowout”).¹, ¹⁴, ¹⁵, ¹⁶, ¹⁷, ¹⁸, ¹⁹, ²², ²⁹, ³¹, ³², ³³, ³⁸, ³⁹, ⁴²

Gagnon et al.²⁸ suggest that it is useful to consider or classify whether the bleeding is likely to be 1) anatomic, for example, massive hemoptysis secondary to a lung carcinoma; 2) generalized, for example, because of thrombocytopenia from bone marrow involvement of a tumor; or 3) mixed. This categorization may guide management.²⁸

Other authors suggest classification as “threatened,” “impending,” or “acute” hemorrhage.³³ Threatened hemorrhage refers to a condition in patients in whom hemorrhage has not yet occurred but there is evidence (clinical or radiological) of inevitable hemorrhage in the future; impending hemorrhage refers to a condition in patients in whom a sentinel bleed has occurred.³³ In practical terms, there is little clinical relevance, as both patient groups need an urgent plan of management to establish how they should be managed if an acute hemorrhage takes place. This, in turn, will depend on the patient's prognosis, quality of life, and personal preferences.

Management of Terminal Hemorrhage 

Most of the sources of literature considered the management of terminal hemorrhage by dividing discussion into 1) identifying patients at risk of terminal hemorrhage, 2) general supportive measures, and 3) use of sedative medication (often termed as “emergency” or “crisis” medication). We also have described the ethical issues that are raised. The key stages in the management of patients at risk of terminal hemorrhage are summarized in Fig. 1.

• 
  • View Large Image
  • Download to PowerPoint

• Fig. 1 

  Summary: management of terminal hemorrhage in patients with advanced cancer.

Most sources stress the importance of identifying patients at risk of terminal hemorrhage and discuss risk factors.¹⁴, ¹⁵, ¹⁶, ¹⁷, ¹⁹, ²⁰, ²¹, ²⁵, ²⁸, ⁴⁰ Most sources specifically discuss risk factors for carotid artery rupture, but one article lists advanced cancer patients who are at risk as including those with thrombocytopenia (<20,000), large head and neck carcinomas, large centrally located lung cancers, refractory acute and chronic leukemia, myelodysplasia, metastatic liver disease, and those on oral anticoagulants.³⁶No specific figures are given for the degree of risk or incidence in the advanced cancer population.

For head and neck cancers, the main risk factors are surgery (e.g., radical neck dissection), radiotherapy (the most implicated risk factor), postoperative healing problems, visible arterial pulsation, presence of a pharyngocutaneous fistula, fungating tumors with artery invasion, and other systemic factors (e.g., age above 50 years, 10%–15% loss of body weight, diabetes mellitus and immunodeficiency, generalized atherosclerosis, and malnourishment).¹³, ¹⁴, ¹⁵, ¹⁶, ¹⁷, ¹⁹, ³⁰, ³¹, ³², ³⁹, ⁴⁰ These risk factors are generally reported as percentages from case series but without any further statistical tests to examine a cause and effect relationship and demonstrate the strength of any association.

Sentinel or herald bleeds and other indicators, such as direct observation by the surgeon or imaging, for example, magnetic resonance imaging, that the tumor is infiltrating the arterial wall, are also mentioned.⁴⁰ Importantly, “even seemingly trivial bleeding may herald carotid artery rupture.”¹⁴

In patients who are identified at risk, various measures to control early signs of bleeding are suggested. These may be considered as local measures or systemic measures, and their use will depend on what is felt to be appropriate for the individual patient. Local measures include compressive dressings, hemostatic agents (e.g., gel foam or calcium alginate [Kaltostat^®; ConvaTec, Skillman, NJ, USA] or 1% sucralfate paste), vasoconstrictor agents (e.g., cocaine or adrenaline 1:1000 solution), local radiotherapy (e.g., radiotherapy is reported to stop recurrent hemoptysis in up to 75% of patients),²⁵ and surgery (vessel ligation).⁴, ²⁰, ²⁵, ²⁶, ²⁸, ³⁶ Systemic measures include vitamin K (if clotting factor abnormalities are present), antifibrinolytic agents (e.g., alpha aminocaproic acid, tranexamic acid, or ethamsylate), octreotide (in the management of gastrointestinal bleeding), and transfusion of blood products.⁴, ²⁰, ²⁵, ²⁶, ²⁸, ³⁶, ⁴⁴

The importance of identifying patients at risk is to allow proactive preparation and advanced planning of the most appropriate course of management if the event occurs.²², ³⁶, ⁴⁰, ⁴¹, ⁴² It is recommended that this process be guided by multidisciplinary team discussions.⁷, ²², ⁴⁰, ⁴¹, ⁴² Potential members of this multidisciplinary team include an oncologist or medical staff, nursing staff, palliative care staff, pharmacist, and chaplain.⁴⁰ One of the main aims of forward planning is the avoidance of a “chaotic situation involving unrealistic life-saving measures or staff unable to acknowledge the imminent death …”.¹⁹

Key information that the multidisciplinary team should consider includes an estimated prognosis, the patient's performance status, and his or her perceived quality of life and preferences.¹⁴, ²⁰, ³⁶, ³⁸, ⁴⁰, ⁴¹ Involvement of the patient in this discussion appears to be contentious and is discussed in more detail later.

Broadly, the options for management can be divided into three categories: 1) general supportive measures, for example, use of dark towels, staying with and comforting the patient; 2) general resuscitative measures, for example, volume and fluid replacement; and 3) specific measures to stop the bleeding, for example, surgical ligation of an artery.

If the patient's prognosis and overall quality of life “warrants it,” then management should consist of general resuscitative measures, such as volume and fluid replacement and specific measures to stop the bleeding.²⁰, ²², ⁴⁰, ⁴¹ However, if the patient's goals are palliative, then management may include measures to stop the bleeding without full resuscitative measures or cardiopulmonary resuscitation.²⁰, ²², ⁴⁰, ⁴¹

We would suggest that general supportive measures should be always appropriate. General resuscitative measures and specific measures to stop the bleeding may be appropriate in some palliative care patients, particularly those with a prognosis of several months, with a good performance status and quality of life.

The scope of this review is to discuss the literature for the palliative management of terminal hemorrhage, which broadly can be considered as general supportive measures and the use of sedative medication to alleviate the patient's distress.

Great emphasis is placed on appropriate planning and use of supportive measures if terminal hemorrhage occurs. The evidence to support the usefulness of these measures is limited to expert opinion/experience and a few case reports (Level 5 Oxford Centre for Evidence-Based Medicine Levels of Evidence). Suggested supportive measures in this situation are:

Part of the planning before the event should ensure that all the necessary equipment is kept in the immediate vicinity of the patient, for example, plastic aprons, face shields, dark towels.¹⁹, ²², ⁴¹ Some authors extend this to ensure that prefilled syringes of sedative medication also are kept within the vicinity of the patient.²², ²⁴, ²⁸, ⁴⁰, ⁴²

The importance of debriefing and counseling of relatives and staff is widely acknowledged.²², ²⁶, ⁴⁰, ⁴¹ Disappointingly, few sources mention children and the specific issues around discussion and support (by professionals trained to do so) at a level appropriate to their cognitive development that will aid their understanding of the situation, before and after the event.⁴⁰, ⁴⁵

The use of sedation is frequently mentioned.¹⁴, ¹⁵, ¹⁷, ¹⁹, ²¹, ²², ²⁵, ²⁶, ²⁷, ²⁸, ²⁹, ³⁶, ⁴⁰, ⁴³, ⁴⁴ However, the drug to be used, the route of administration, and the drug doses are not consistent. Specific recommendations for drugs, doses, and routes of administration are listed in Table 3.

Table 3. The Use of Crisis Medication: Summary of Published Literature

First Author	Year	Source	Country	Type of Publication	Type of Terminal Hemorrhage	Sedative Medication Recommended	Level of Evidencea
Back6	2001	Palliative Medicine Handbook	UK	Guideline	Any	Drug: midazolam Route: IV or IM; Dose: 5–10mg IM Alternatives suggested: 1. Drug: strong opioid, e.g., diamorphine Route: not stated; Dose: 10mg diamorphine if opioid naïve or twice the 4- hourly equivalent dose 2. Drug: ketamine Route: IV or IM; Dose: 250mg IV or 500mg IM	5
Cambridge and Huntington Palliative Care Group43	2006	Cambridgeshire Primary Care Trust	UK	Guideline	Any	Drug: midazolam Route: IV or buccal; Dose: 5mg, repeated as necessary to maintain drowsiness. No maximum dose stated Alternative suggested (for community use): Drug: diazepam Route: rectal; Dose: 10mg	5
Forbes29	1997	Clinical Otolaryngology	UK	Retrospective audit of 38 head and neck cancer patients in which one died from hemorrhage	Bleeding related to head and neck cancers	Drug: midazolam Route: subcutaneous; Dose: not stated	5
Frawley and Begley18, 22	2006	British Journal of Nursing	UK	Literature review and guideline	Carotid artery rupture	Drug: midazolam Route: IV, SC, or IM; Dose: 2.5–7.5mg IV or 5mg IM or 5–10mg SC, Maximum dose 10mg.	5
Fowell and Stuart 44	2007	The Foundation Years	UK	Guidance for junior doctors	All	Drug: midazolam Route: IV or IM; Dose: 10mg Alternative suggested: Drug: diamorphine (if in pain) Route: IV or IM; Dose: 5mg if opioid naïve or usual breakthrough dose	5
Frawley41	Not dated	National Council for the Professional Development of Nursing and Midwifery	Ireland	Guideline	Carotid artery rupture	Drug: midazolam Route: IV, SC, or IM injection; Dose: starting dose 2mg IV or 5mg SC/IM. Maximum dose 10mg. Alternative suggested: Drug: morphine Route: IV, SC, or IM; Dose: 4–10mg IV or 5–10mg SC or IM.	5
Gagnon et al.28	1998	Journal of Palliative Care	Canada	Case report	Massive hemoptysis and massive hematemesis (GI bleed)	Drug: midazolam Route: SC or IM; Dose: 5mg; no instructions regarding repeating doses or maximum dose	5
Kane19	1983	Oncology Nursing Forum	USA	Guideline/Protocol	Carotid artery rupture	Drug: morphine Route: IV; Dose: not stated	5
Lovel27	2000	British Journal of Oral and Maxillofacial Surgery	UK	Editorial and expert opinion based on authors' experience	Hemorrhage relating to head and neck tumors	Drug: midazolam Route: IV or SC; Dose: 5mg IV or 10mg SC	5
MacMillan and Struthers17	1987	Canadian Critical Care Nursing Journal	Canada	Guidelines	Carotid artery rupture	Drug: morphine Route: IV; Dose: not stated	5
Oneschuk24	1998	Canadian Family Physician	Canada	Expert opinion based on experience of a palliative care team	All	Drug: midazolam Dose: 5mg subcutaneously, repeated every 10–15 minutes until the patient is adequately sedated	5
Pereira and Phan20	2004	The Oncologist	Canada	Literature review of bleeding in advanced cancer	All	Drug: midazolam Route: IV or SC; Dose: 2.5–5mg repeated after 10–15 minutes. No maximum dose stated	5
Potter40	2005	British Association of Head and Neck Oncology Nurses	UK	Guideline	Carotid artery rupture	Drug: midazolam Route: IV, SC, or IM; Dose: 5mg IV or 5– 10mg IM or SC. “titrate up” until fully sedated Alternative suggested: Drug: diazepam Route: rectal; Dose: 5–10mg	5
Prommer36	2005	Hematology	USA	Clinical practice review	All	Drug: midazolam Route: SC or IM/IV; Dose: 2.5–5mg. No maximum dose stated
Regnard and Makin26	1992	Palliative Medicine	UK	Guideline and algorithm	All	Drug: diazepam Route: IV or PR. No dose given. Alternative suggested: Drug: midazolam Route: IM into deltoid	5
Society of Otorhinolaryngology and Head-Neck Nurses15	1995	ORL Head and Neck Nursing		Guideline	Carotid artery rupture	Drug: morphine Route and dose not stated	5
Scholes et al.1	1997	Progress in Palliative Care	UK	Literature review	Carotid artery rupture	Drug: benzodiazepine (specific drug not stated) Route: IV or IM; Dose: not stated	5
Smith25	1992	Annals Academy of Medicine Singapore	UK	Literature review and guidelines	All	Drug: morphine and hyoscine (scopolamine) and anxiolytic (chlorpromazine or haloperidol) Route and dose not stated. Alternative suggested: Drug: diazepam or midazolam Dose: diazepam 5–10mg pr or midazolam 5– 10mg “injected.”	5
Troke and Jeffrey42	2005	Poole Hospital NHS Trust	UK	Guideline	Carotid artery rupture	Drug: diamorphine and midazolam Route: IV or IM; Dose: diamorphine 5mg and midazolam 5mg. Repeat “as necessary” to maximum diamorphine 10mg and midazolam 10mg	5
Watson et al.7	2005	Oxford Handbook of Palliative Care	UK	Guidelines	All	Drug: midazolam Route: IV, SC, or buccal; Dose: 10mg Alternatives suggested: 1. Drug: ketamine Route: IV or IM; Dose: 150–250mg IV or 500mg IM 2. Drug: diamorphine Route and dose not stated	5

IV=intravenous; IM=intramuscular; SC=subcutaneously; PR=per rectum; GI=gastrointestinal.

The aim of sedation is uniformly cited as alleviating distress and providing comfort to the patient.¹, ¹⁹, ²², ²⁴, ²⁵, ²⁸, ³⁶, ⁴⁰, ⁴², ⁴³ However, only one author notes the importance of clarification that the intent is to relieve distress and not to hasten death (i.e., euthanasia),⁴⁰ which has particular importance in the documentation of the management plan and the explanation to the family.

Midazolam was the most frequently recommended drug.⁶, ⁷, ²⁰, ²², ²⁴, ²⁷, ²⁸, ²⁹, ³⁶, ⁴⁰, ⁴³, ⁴⁴ All sources used similar reasons to justify midazolam over other alternatives: It has a rapid onset of action, can be administered rapidly, is safe and short acting (therefore, should not cause any harm if, in fact, the event is not terminal), and it provides retrograde amnesia (hence, if the bleed was, in fact, a herald bleed, then the patient would have little memory of the event).²², ²⁴, ²⁶, ²⁷, ²⁸, ²⁹, ³⁶, ⁴⁰ One author suggests that a further advantage is its “dramatic” blood pressure–lowering effect, which in itself may help arrest the hemorrhage;²⁷ however, this may not necessarily be a therapeutic effect but more of a reflection of severe volume depletion from bleeding and vasodilatation from the benzodiazepine. The dose suggested varied (Table 3). The use of a prefilled syringe (refilled weekly) stored near the patient is also frequently mentioned, as midazolam can be stored at room temperature for up to 13 days.¹⁹, ²², ²⁸, ⁴⁰, ⁴²

Those articles that supplied evidence for the recommendation of midazolam²⁴, ²⁸ referred to a study which randomized 36 patients to receive IV midazolam 1–5mg for “conscious sedation” during minor operative procedures (e.g., abscess drainage) in an emergency department and found that it achieved quicker onset of sedation, shorter duration of action, and more amnesia than diazepam 2.5–15mg IV.⁴⁶ Patients in that study were benzodiazepine naïve, which would not necessarily be the case in palliative patients with advanced malignancy. Only a few sources suggested the need to consider background benzodiazepine use or higher dose use in patients who may be benzodiazepine tolerant, for example, heavy alcohol drinkers.⁶, ⁷, ⁴⁰

Strong opioids were suggested by some authors, usually (but not always) as second-line agents if “the course of events is prolonged and the patient's distress is not relieved by the use of midazolam” or if the patients had a bleed that was not likely to result in imminent death and complained of pain and/or breathlessness.¹⁵, ¹⁷, ²², ²⁵ Reasons against using first-line opioids included the fact that hemorrhage is usually not painful; that strong opioids are usually locked in controlled drug cupboards, leading to delays in administering them; and that diamorphine needs to be dissolved (leading to a delay).⁶, ⁷, ²⁷, ⁴⁰ The doses recommended varied between sources (Table 3), and not all authors acknowledged that consideration was required to be given to any background opioids the patient had already taken.⁶, ⁷, ⁴⁰ One set of guidelines recommended the combined use of midazolam and diamorphine from a prefilled syringe stored in the patient's locked bedside cabinet.⁴²

Some sources also suggest ensuring the availability of antidotes to the sedative medication used, for example, flumazenil to reverse midazolam. This may need to be considered, particularly for use in the event of a substantial herald bleed being mistaken for a terminal hemorrhage.²², ⁴¹, ⁴²

Finally, ketamine was suggested by two sources as an alternative and justified by stating that the effect of a standard dose is more predictable than opioids or benzodiazepines, as the patient is unlikely to have been taking it regularly.⁶, ⁷ The dose recommendations are shown in Table 3.

The suggested route of administration was variable. Although some sources solely mention subcutaneous administration of drugs,²⁴, ²⁸, ²⁹ others state that drugs given subcutaneously are poorly absorbed in circulatory shutdown,⁴⁴ and that an IV cannula should be sited in preparation so that the IV route can be used.⁴⁰, ⁴², ⁴⁵ This view may be further supported by the fact that the onset of action of midazolam given subcutaneously is 5–10 minutes compared with 2–3 minutes when administered IV as a bolus injection⁴⁷ and may even, therefore, be longer in a patient with peripheral circulatory shutdown. We were unable to find pharmacological data for the onset of action of intramuscular midazolam; the time to peak plasma concentration is reported to be 30 minutes.⁴⁷, ⁴⁸

Few authors acknowledge that, in the event of massive terminal hemorrhage, the patient may become unconscious within minutes and die very quickly before the sedation (regardless of drug given) has had a chance to work.⁴⁰, ⁴¹, ⁴², ⁴³, ⁴⁴, ⁴⁵ Furthermore, despite the breadth of medications reported, there was no clinical evidence that one route produced a more rapid or greater sedative effect than another in the context of terminal hemorrhage.

A number of ethical issues are discussed in the context of managing terminal hemorrhage. Whether the patient should be informed of the risk of terminal hemorrhage appears to be particularly topical. Some authors suggest that the patient and caregivers should uniformly be informed of the possibility of this event,²⁰, ²⁸ but most suggest that this should be considered on an individual basis.⁴⁰, ⁴¹ Most sources present a balanced argument between the need to provide this information against the potential psychological impact of doing so.

It is generally recommended that the multidisciplinary team should assess both the patient's and family's knowledge and acceptance of their prognosis and extent of disease and explore how much information they wish to receive, their desired level of participation in decisions about their care, and their coping strategies (including preexisting anxiety and depression).¹⁸, ¹⁹, ²², ²⁹, ⁴⁰ Many terminal patients will have already speculated about how the end will come.¹⁹

Timing and how this information is imparted is of crucial importance.²² Some sources suggest that these discussions should take place in the event of a herald bleed.²² The content of any discussion should include the patient's preferences with regard to aggressive attempts to prolong life or supportive care. Although active resuscitation of the patient suffering from severe hemorrhage is very rarely appropriate when a person has advanced cancer, there may be exceptions, such as in patients who want to achieve a particular event in their lives for which they would like to avoid an imminent death.²⁵, ⁴⁰

Arguments raised supporting discussion with the patient include the following: poor communication about end-of-life issues may lead to an increased sense of abandonment and isolation;⁴⁰ there may be potential comfort from the knowledge that the event is unlikely to be painful; sedation will be available and they will be supported at home and in the hospital/hospice setting;¹⁹, ⁴⁰ the potential effect of not being prepared for the event, which may be more likely to contribute to complex bereavement issues (for the family or carer);¹⁹, ⁴⁰ a “right” to know;⁴⁰ discussion allows the family to decide in advance if they would wish to stay with the patient;⁴⁰ a duty toward truth-telling and a duty to tell the patient;¹⁸ and gaining consent about decisions that affect their life is fundamentally significant to the patient's autonomy.¹⁸

Arguments raised against discussion with the patient include the following: a “right” not to know (patients and families should be able to choose how much information they are given);⁴⁰ informing a patient about a problem that may not happen, and even if it does, the person has no control over it, is questionable;²² unnecessarily frightening a patient and increasing his or her level of anxiety at a time when he or she is already compromised by the disease and its treatment;¹⁸, ²⁹, ⁴⁰ and medical decisions taken in relation to a carotid rupture that exclude the patient may be paternalistic but justified on the grounds of being beneficent.¹⁸

A detailed discussion of the ethical issues that are raised by the literature is beyond the scope of this review, but in some areas, more questions emerge than answers. Furthermore, caution is required when interpreting the current literature in this context, as all the discussion around the issues of whether to inform patients and relatives draw upon opinions of health care professionals (and what they perceive patients and carers would want), general ethical principles, and extrapolation from research in other settings on ethical decision making and patient autonomy rather than specific qualitative literature relating to the experiences and wishes of patients and carers actually in this situation.

A further ethical issue arises around the administration of sedative medication by the caregiver. Although several sources suggest that the caregiver “should” be instructed in the use of a prefilled syringe and how to administer the medication subcutaneously,²⁰, ²⁴, ²⁸, ⁴⁰ little consideration was given to the ethical implications of asking a family member to do this. On the occasions when it was mentioned, it mostly related to ensuring that there were practical alternatives (e.g., rectal diazepam) rather the whether it would actually be appropriate.²⁸, ⁴⁰ Only one author goes as far as mentioning that health care professionals should make sure the family members are “comfortable” to do so.²⁴ Again, no qualitative research has yet explored carer and family views and preferences in this situation.

Back to Article Outline

Discussion 

The evidence base that informs the management of terminal hemorrhage in patients with advanced cancer is limited to single-case reports and expert opinion. Current published recommendations are, therefore, based on Level 5 evidence alone,⁴⁹ and professionals need to be aware of this when anticipating and managing this, thankfully, uncommon occurrence.

It is widely described that the fundamentally important first step in managing patients with terminal hemorrhage is identifying patients at risk to allow proactive planning of how to manage the event if it occurs and what level of management is appropriate for each individual patient. Management may involve general supportive measures, general resuscitative measures, and/or specific invasive intervention to stop the bleeding. Figure 1 summarizes the supportive management of terminal hemorrhage in patients with advanced cancer.

There is a general message from the current medical literature that the patient and their family should be warned of the risk of terminal hemorrhage (and involved in the decision-making process) but also an acknowledgment of potentially harmful effects of such discussion. This is based upon health care professional's perceived views of the patient/family perspective rather than primary qualitative data from patients and families themselves.

General supportive measures, such as the use of dark towels and ensuring a nurse stays with the patient, are widely endorsed. The use of sedative medication is also widely recommended, although the importance of clarifying that the aim is to alleviate distress and not to hasten death (i.e., euthanasia) is not widely acknowledged. This is particularly important in the documentation of the management plan and the explanation to the relatives.

Furthermore, although the use of sedation in this situation is widely recommended, there is no consensus of choice of drug, dose, or route, and as there is little evidence on which current guidelines can be based, it is difficult to critically assess each recommendation. We have highlighted potential physiological and pharmacological reasons why subcutaneous medication is potentially the least efficacious route and acknowledge the recommendation by some authors that a preplaced IV cannula would facilitate the route of administration most likely to have the desired effect.

There are particular practical reasons against using opioids in this situation, such as delays in obtaining controlled drugs from locked medicine cabinets, difficulty in deriving doses in patients who are not opioid naïve and the wider ethical issues of using opioids in patients who are not in pain. Given that terminal hemorrhage may occur rapidly with death within minutes, it is difficult to be sure that any drug, by whatever route, would have taken effect before death occurred.

It is readily apparent that the published literature in this area is very limited. There is a large overlap in references between sources, most of which are based on expert opinion, and there is very little actual patient data (one case report and one audit series that included one patient who died from a terminal hemorrhage). The desperate need for research in this area is widely acknowledged.²², ⁴⁰

Back to Article Outline

Recommendations 

A randomized controlled trial comparing different sedative medication regimens to alleviate patient distress during terminal hemorrhage is unlikely to develop for many practical and ethical reasons, but equally, it is unethical not to engage in research of clinical practice in an area that currently has no evidence base. At the very least, we should be aiming for:

Back to Article Outline

Conclusion 

For many practical and ethical reasons, research around the management of terminal hemorrhage in cancer patients is very difficult. Presently, there is very little research evidence on which guidelines can be based, and the medical literature is limited to expert opinion and case reports. Current suggestions for managing terminal hemorrhage in this context include the use of dark towels (Level 5 evidence), use of suction when possible (Level 5 evidence), and the use of sedative “crisis” medication, which may involve midazolam, diamorphine/morphine, or ketamine by subcutaneous, intramuscular, or IV administration (Level 5 evidence). The management of terminal hemorrhage is, however, a vital area for research given the importance of being able to reduce distress for the patient if the event occurs and reduce any potential long-term impact for the family or carer.

Back to Article Outline

References 

1. Scholes C, Frankum L, Maher J. Management of advanced head and neck cancer. Prog Palliat Care. 1997;5:54–59
   • View In Article
2. Maran AGD, Amin M, Wilson JA. Radical neck dissection: a 19-year experience. J Laryngol Otol. 1989;103:760–764
   • View In Article
   • MEDLINE
3. Kvale PA, Simoff M, Prakash UB. Lung cancer. Palliative care. Chest. 2003;123:284S–311S
   • View In Article
   • MEDLINE
   • CrossRef
4. Jamjihan MC. Bleeding problems in palliative care patients. J Pharm Care Pain Symptom Control. 1999;7:37–46
   • View In Article
5. In:  Doyle D,  Hanks G,  Cherny N,  Calman K editor. Oxford textbook of palliative medicine. 3rd ed. Oxford, UK: Oxford University Press; 2005;
   • View In Article
6. Back I. Palliative medicine handbook. 3rd ed.. Cardiff, UK: BPM Books; 2001;
   • View In Article
7. Watson M, Lucas C, Hoy A, Back I. Oxford handbook of palliative care. Oxford, UK: Oxford University Press; 2005;
   • View In Article
8. Cassidy J, Bissett D, Spence RAJ, Payne M. Oxford handbook of oncology. Oxford, UK: Oxford University Press; 2006;
   • View In Article
9. In:  Abeloff MD,  Armitage JO,  Niederhuber JE,  Kastan MB,  McKenna WG editor. Clinical oncology. 3rd ed.. Philadelphia, PA: Elsevier/Churchill Livingstone; 2004;
   • View In Article
10. Otto SE. Oncology nursing. 4th ed.. St. Louis, MO: Mosby Inc.; 2001;248–257
    • View In Article
11. Peckham M, Pinedo M, Veronesi U. Oxford textbook of oncology. Oxford, UK: Oxford Medical Publications; 1995;
    • View In Article
12. Neal AJ, Hoskin PJ. Clinical oncology: Basic principles and practice. 3rd ed.. London, UK: Arnold; 2003;
    • View In Article
13. Schiech L. Carotid artery rupture. Clin J Oncol Nurs. 2000;4:93–94
    • View In Article
    • MEDLINE
14. Fortunato L, Ridge JA. Surgical palliation of head and neck cancer. Curr Probl Cancer. 1995;19:153–165
    • View In Article
    • MEDLINE
    • CrossRef
15. Society of Otorhinolaryngology and Head-Neck Nurses. Practice guidelines. Carotid artery rupture. ORL Head Neck Nurs. 1995;13:31
    • View In Article
    • MEDLINE
16. Arbogast D. What you need to know about managing a carotid artery rupture to prepare yourself to handle this sudden dramatic bleeding. Nursing. 2001;31:1–2
    • View In Article
17. MacMillan K, Struthers C. Algorithm for the emergency nursing management of spontaneous carotid artery rupture. Can Crit Care Nurs J. 1987;4:20–21
    • View In Article
    • MEDLINE
18. Frawley T, Begley CM. Ethical issues in caring for people with carotid artery rupture. Br J Nurs. 2006;15:100–104
    • View In Article
    • MEDLINE
19. Kane KK. Carotid artery rupture in advanced head and neck cancer patients. Oncol Nurs Forum. 1983;10:14–18
    • View In Article
    • MEDLINE
20. Pereira J, Phan T. Management of bleeding in patients with advanced cancer. Oncologist. 2004;9:561–570
    • View In Article
    • MEDLINE
    • CrossRef
21. Frawley T, Begley CM. Causes and prevention of carotid artery rupture. Br J Nurs. 2005;14:198–202
    • View In Article
22. Frawley T, Begley CM. Caring for people with carotid artery rupture. Br J Nurs. 2006;55:24–28
    • View In Article
23. Wrede-Seaman LD. Management of emergent conditions in palliative care. Prim Care. 2001;28:317–328
    • View In Article
    • Full Text
    • Full-Text PDF (644 KB)
    • CrossRef
24. Oneschuk D. Subcutaneous midazolam for acute hemorrhage in patients with advanced cancer. Can Fam Physician. 1998;44:1461–1462
    • View In Article
    • MEDLINE
25. Smith AM. Emergencies in palliative care. Ann Acad Med. 1992;23:186–190
    • View In Article
26. Regnard C, Makin W. Management of bleeding in advanced cancer—a flow diagram. Palliat Med. 1992;6:74–78
    • View In Article
    • CrossRef
27. Lovel T. Palliative care and head and neck cancer. Br J Oral Maxillofac Surg. 2000;38:253–254
    • View In Article
    • MEDLINE
    • CrossRef
28. Gagnon B, Mancini I, Pereira J, Bruera E. Palliative management of bleeding events in advanced cancer patients. J Palliat Care. 1998;14:50–54
    • View In Article
    • MEDLINE
29. Forbes K. Palliative care in patients with head and neck cancer. Clin Otolaryngol Allied Sci. 1997;22:117–122
    • View In Article
    • MEDLINE
30. Lesage C. Carotid artery rupture. Cancer Nurs. 1986;9:1–7
    • View In Article
    • MEDLINE
31. Heller KS, Strong EW. Carotid arterial hemorrhage after radical head and neck surgery. Am J Surg. 1979;138:607–610
    • View In Article
    • MEDLINE
    • CrossRef
32. Coleman JJ. Treatment of the ruptured or exposed carotid artery: a rational approach. South Med J. 1985;78:262–267
    • View In Article
    • MEDLINE
    • CrossRef
33. Cohen J, Rad I. Contemporary management of carotid blowout. Curr Opin Otolaryngol Head Neck Surg. 2004;12:110–115
    • View In Article
    • MEDLINE
    • CrossRef
34. Porto DP, Adams GL, Foster C. Emergency management of carotid artery rupture. Am J Otolaryngol. 1986;7:213–217
    • View In Article
    • MEDLINE
    • CrossRef
35. Shumrick DA. Carotid artery rupture. Laryngoscope. 1973;83:1051–1061
    • View In Article
    • MEDLINE
    • CrossRef
36. Prommer E. Management of bleeding in the terminally ill patient. Hematology. 2005;10:167–175
    • View In Article
    • MEDLINE
    • CrossRef
37. Patel U, Pattison CW, Raphael M. Management of massive haemoptysis. Br J Hosp Med. 1994;52:74–78
    • View In Article
    • MEDLINE
38. Desuter G, Hammer F, Gardiner Q, et al. Carotid stenting for impending carotid blowout: suitable supportive care for head and neck cancer patients?. Palliat Med. 2005;19:427–429
    • View In Article
    • MEDLINE
    • CrossRef
39. Nieto CS, Solano JME, Martinez CB, et al. The carotid artery in head and neck oncology. Clin Otolaryngol. 1980;5:404–417
    • View In Article
40. Potter E. The management of carotid artery rupture related to the terminal care of the head and neck cancer patient: information and guidelines. British Association of Head and Neck Oncology Nurses. 2005. Available from: www.bahnon.org.uk. Accessed September 17, 2008.
    • View In Article
41. Frawley T. Nursing management of carotid artery rupture. The National Council for the Professional Development of Nursing and Midwifery. Available from: www.ncnm.ieAccessed September 17, 2008
    • View In Article
42. Troke C, Jeffrey Y. Guidelines for the medical staff to prescribe and nursing and medical management of the patient with carotid artery rupture/bleed. 2005. Available from: www.poole.nhs.uk. Accessed September 17, 2008.
    • View In Article
43. Cambridge and Huntington Palliative Care Group. Emergencies inpalliative care. Cambridgeshire Primary Care Trust. 2006. Available from: www.arthurrankhouse.nhs.uk. Accessed September 17, 2008.
    • View In Article
44. Fowell A, Stuart NSA. Palliative care emergencies. Foundation Years. 2007;3:26–30
    • View In Article
45. MacKay F, Cook C. Policy, procedure and guidelines for the management of carotid artery rupture related to the terminal care of the head and neck cancer patient. Royal United Hospital Bath NHS Trust. 2008. Available from: http://www.ruh.nhs.uk/about/policies/documents/clinical_policies/blue_clinical/Blue_771_Management_of_Carotid_Artery_Rupture.pdf. Accessed September 17, 2008.
    • View In Article
46. Wright SW, Chudnofsky CR, Dronen , et al. Comparison of midazolam and diazepam for conscious sedation in the emergency department. Ann Emerg Med. 1993;22:201–205
    • View In Article
    • Abstract
    • Full-Text PDF (465 KB)
    • CrossRef
47. Twycross R, Wilcock A. Palliative care formulary. 3rd ed.. Nottingham, UK: palliativedrugs.com Ltd.; 2007;
    • View In Article
48. European Medicines Agency. Summary of product characteristics: Hypnovel and associated names. Available from: http://www.emea.europa.eu/pdfs/human/referral/midazolam/462002en.pdf. Accessed October 20, 2008.
    • View In Article
49. Oxford Centre for Evidence-based Medicine. Levels of evidence. 2009. Available from: http://www.cebm.net/index.aspx?o=1025. Accessed April 2, 2009.
    • View In Article
50. Fowell A, Johnstone R, Finlay IG, Russell D, Russell IT. Design of trials with dying patients: a feasibility study of cluster randomisation versus randomised consent. Palliat Med. 2006;20:799–804
    • View In Article
    • MEDLINE
    • CrossRef