This site is intended for health professionals only
Christine Hirsch BPharm, PhD
Lecturer in Clinical Pharmacy
School of Clinical and Experimental Medicine
University of Birmingham
This article examines the difficult subject of the management of respiratory secretions that occur near to death, commonly called death rattle. Although death rattle is anticipated in palliative care, this phenomenon has also been noted as a problem in dying patients who have been withdrawn from ventilator treatment in intensive care units. With the implementation of the end-of-life care pathways, many acute and secondary care units have adopted policies that advocate anticipatory therapeutic strategies for the routine treatment of secretions with anticholinergic drugs. This article explores the postulated causes and the current evidence base supporting the treatment of death rattle.
Causes and definitions
The treatment of death rattle is based on the theory that the ‘rattling’ noise is caused by an accumulation of secretions produced by the salivary glands and bronchial mucosa in the oropharynx and bronchi. When the patient loses their swallowing and cough reflexes because of fatigue and weakness, turbulent airflow through or over these secretions in the hypopharynx is thought to produce the rattling or gurgling noise. This process is likely to be accentuated if the patient is in a supine or semi-recumbent position. The true underlying cause, however, remains uncertain.
Based on the above premise, pharmacological intervention has been aimed mainly at reducing salivary and bronchial secretions by administering anticholinergic agents. As air flow turbulence may be affected both by ventilatory rate and airways resistance, sedative drugs that can reduce respiratory rate, such as opioids or benzodiazepines that are widely used in the terminal stages of palliative care, may also have an impact on noisy breathing.
Ellershaw and colleagues defined death rattle as, “the sound audible at the bedside produced by movement of secretions in the hypopharynx or the bronchial tree in association with respiration.”
This definition has been employed in subsequent studies providing a degree of reference for standardisation when measuring incidence and therapeutic drug effect. Studies also use the terms ‘respiratory tract secretions’ or ‘noisy breathing’, which are more sensitive descriptions to use when talking to patients and relatives.
Salivation is under muscarinic control via acetylcholine in the parasympathetic nervous system (M1, M2, M3 receptors). Dry mouth is a well-recognised side-effect of anticholinergic agents and there is evidence to show that anticholinergics can be helpful in the treatment of drooling. However, the effectiveness of anticholinergic treatment in reducing the production of excess saliva is difficult to establish and may cause patients to have a dry mouth and difficulty in coughing up phlegm from the back of the throat.
Respiratory passages are kept moist by a layer of mucus, which is a dilute aqueous solution containing electrolytes, mucins (glycoproteins) and enzymes. The mucus is secreted partly by surface epithelial goblet cells lining the passages and partly by the submucosal glands. The mucus exists as two layers: an upper layer that traps inhaled particles and a lower layer in which the cilia beat. The efficiency of mucociliary clearance – upwards towards the pharynx propelled by the cilia – is determined by the physical properties of the mucus layer, which can be impaired when the volume or viscosity of the mucus increase or change. The secretion of mucus is under neuronal control, mainly cholinergic although adrenergic stimulation, vasoactive intestinal peptide (VIP), substance P, neurokinin A and nitric oxide may also be involved. The contribution of various transmitters is, however, species specific. Opioid drugs depress neronal activity by interacting with opioid receptors at terminals of cholinergic and sensory afferent nerve fibres.
It is likely that the underlying process resulting in death rattle is multifactorial and may be influenced by patients’ concomitant conditions, dehydration, infection or inflammatory processes and involve different pharmacological pathways.
Rationale in palliative care
Recognising the probable multifactorial pathologies, Bennett postulated that there was a Type 1 death rattle due to a loss of swallowing reflex near to death, in which predominantly salivary secretions accumulated unpredictably during the last few hours of life. The cause of Type 2 death rattle was postulated to be mainly an accumulation of bronchial secretions over a period of several days, as the patient became too weak to cough effectively.
While anticholinergic treatment may be effective for patients with Type 1, those with Type 2 responded less well to anticholinergic treatment because of additional factors. This hypothesis was recently supported by Morita who found that increased secretions as a result of tumour, oedema or bleeding caused death rattle resistant to treatment with anticholinergics. Macleod also suggested that neurogenic pulmonary oedema (NPE) was an under-recognised cause of death rattle (resistant to traditional antisecretory medication or postural change).
In the UK, the three main anticholinergics used to treat death rattle are hyoscine hydrobromide, hyoscine butylbromide and glycopyrronium bromide. Although atropine is used to treat death rattle in Europe, in the UK it is usually used to alleviate symptoms of excessive saliva causing drooling, for example in motor neurone disease in palliative care.
This drug has an onset of antisialagogue action of 30–60 minutes with a duration of action reported between one and nine hours. It is considered more potent than atropine in terms of its action on secretory glands, but is less potent in its effects on heart, intestinal and bronchial smooth muscle. In contrast to atropine, hyoscine hydrobromide produces bradycardia and central nervous system depression.
Normal doses of hyoscine hydrobromide cause drowsiness, but higher doses can cause central stimulation restlessness and irritability. Doses for 24-hour subcutaneous infusion generally range from 600µg to 1200µg in 24 hours with stat doses of 200–400µg.
Hyoscine hydrobromide injection is licensed for subcutaneous administration in the UK. Hyoscine hydrobromide hydrolyses below pH 3, but is compatible with most drugs given by subcutaneous infusion used in palliative care. It is also available as a transdermal preparation applied to the postauricular area and has been used to treat death rattle, particularly in community care.
A quaternary derivative of hyoscine, hyoscine butylbromide has poor oral absorption and does not readily cross the blood–brain barrier. Central effects are therefore uncommon. Pharmacokinetic data for subcutaneous administration comes from healthy volunteers who demonstrate an anti-secretary duration of action of less than two hours. Hyoscine butylbromide injection is not licensed for subcutaneous infusion but good compatibility data is available to allow the use of hyoscine butylbromide injection in combination with most drugs commonly administered by subcutaneous infusion in palliative care. The usual dose range is 20–120mg over 24 hours for subcutaneous infusion with stat doses of
This drug is a synthetic quaternary ammonium anticholinergic agent that does not cross the blood–brain barrier, making central effects rare. The elimination of glycopyrronium bromide is significantly prolonged in uraemic patients. Glycopyrronium bromide is not licensed for subcutaneous administration but data is available for compatibility in a syringe with diamorphine. The injection has a pH of 2.3–4.3, and care must be taken when combining drugs for subcutaneous infusion because the rate of hydrolysis increases significantly above pH 6.0.
In addition to the desired antisialogogue effect, antichoinergics may also, to varying degrees, cause blurred vision, cardiovascular effects, constipation, urine retention, hesitancy of micturition and reduced sweating. However, in patients with reduced consciousness these side effects become difficult to monitor and may go unnoticed. The anticholinergic effects of other drugs that the patient may be taking should also be considered. The usual 24-hour dose range for subcutaneous infusion is 600–1200 µg, with stat doses of 200–400µg.
The most straightforward, non-drug treatment is repositioning of the patient. If this occurs at the same time as pharmacological intervention, objective evaluation of the outcome can be difficult. Some units also employ suction as a method to reduce the noisy breathing.
Death rattle is reported to occur in between 23% and 92% of dying patients. This data mainly represents patients in palliative care settings. The variation in incidence may reflect the general difficulties in conducting studies on death rattle in this group of patients.
A Cochrane review concluded that there was no evidence to show that any intervention, pharmacological or non-pharmacological was superior to placebo in the treatment of ‘noisy breathing’. The review, updated in 2010, identified 32 studies of which only four met the inclusion criteria, which were randomised controlled trials, controlled before and after studies, interrupted time series with more than 10 participants and an outcome of the intensity of noisy breathing. One large study (333 participants) that compared the use of atropine, hyoscine hydrobromide and hyoscine butylbromide showed no difference between the groups. The effectiveness after one hour was between 37% and 42%. Most of the studies included relatively small cohorts ranging from 10 to 31 participants and usually involved the nurse looking after the patient performing the outcome assessment. Insufficient data was available to carry out detailed data analysis.
Studies of this type in this group of patients are difficult to conduct for a variety of reasons. Ethical difficulties in recruiting to this patient group, lack of objective outcome measures and methods of monitoring that are acceptable to ethics committees, result in subjective evaluation and observer bias. Repositioning of the patient may occur at the same time as treatment is given making it difficult to assess the benefit of one intervention over another.
Explanation and reassurance
The World Health Organization (WHO) definition of palliative care is a holistic definition and addresses the quality of life of patients and their families.
Treatment for death rattle in a patient is often initiated by healthcare staff concerned about the distress caused to relatives by the noise, rather than any evidence of harm or distress to the patient. However when interviewed, only about half of relatives supported this view.
There is little evidence to show that death rattle is distressing to the patient, although side-effects caused by administration of anticholinergic agents may go unnoticed if the patient is not fully conscious and able to communicate.
There is still no evidence that determines whether pharmacological or non-pharmacological intervention is better than placebo in treating death rattle or that one anticholinergic is more effective than another.
Good communication and explanation of the symptom, together with reassurance to relatives that the patient is not distressed by the noise along with repositioning of the patient where appropriate should be encouraged as a first-line intervention. Once anticholinergics are started there is a tendency to continue treatment but this should be reviewed if there is no therapeutic benefit.
1. Kompanje EJ. ‘Death rattle’ after withdrawal of mechanical ventilation: practical and ethical considerations. Intensive Crit Care Nurs 2006; 22: 214–19.
2. Ellershaw JE, Sutcliffe JM, Saunders CM. Dehydration and the dying patient. J Pain Symptom Manage 1995; 10: 192–7.
3. Jongerius PH, van Tiel P, van Limbeek J et al. A systematic review for evidence of efficacy of anticholinergic drugs to treat drooling. Arch Dis Child 2003; 88: 911–14.
4. Lopez-Vidriero MT, Colstello J, Clark THJ et al. Effect of atrophine on sputum production. Thorax 1975; 30: 543–47.
5. Rogers DF. Pharmacological regulations of the neuronal control of airway’s mucus secretion. Curr Opin Pharmacol 2002;2:249–55.
6. Bennett MI. Death rattle: An audit of hyoscine (scopolamine) use and review of management. J Pain Symptom Manage 1996;12:
7. Morita T, Hyodo I, Yoshimi T et al. Incidence and underlying etiologies of bronchial secretion in terminally ill cancer patients: a multicentre, prospective, observational study. J Pain Symptom Manage 2004; 27: 533–39.
8. Macleod AD. Neurogenic pulmonary edema in palliative care. J Pain Symptom Manage 2002; 23: 154–56.
9. Ali-Melkkila T, Kante J, Iisalo E. Pharmacokinetics and related pharmacodynamics of anticholinergic drugs. Acta Anaesthesiol Scand 1993; 37: 633–42.
10. Clissold SP and Heel RC. Transdermal hyoscine (scopolamine): A preliminary review of its pharmacodynamic properties and therapeutic efficacy. Drugs 1985;29:189–207.
11. Herxheimer A and Haefeli L. Human pharmacology of hyoscine butylbromide. Lancet 1966; 418: 418–21.
12. Smith J, Hirsch C, Marriott J et al. The stability and compatibility of diamorphine and glycopyrrolate in PCA syringes. Pharm J 2000; 265: R69.
13. Wee B and Hillier R. Interventions for noisy breathing in patients near to death. Cochrane Database of Systematic Reviews 2008, Issue 1, Art, No,:CD005177.DOI:10.1002/14651858.CD005177.pub2.
14. Wee BL, Coleman PG, Hillier R et al. The sound of death rattle I: are relatives distressed by hearing the sound? Palliat Med 2006; 20: 177–81.