Palliative Symptom Control

Pain

Assessment, mechanisms and classification

Pain at the end of life is common, fluctuating and often mixed-mechanism. Assessment should identify mechanism, temporal pattern, severity, functional impact, opioid exposure, renal/hepatic reserve and patient goals. Use a numerical rating scale (NRS 0–10) or verbal rating scale; in cognitively impaired patients use behavioural tools such as PAINAD. Clinically important response is usually a reduction of ≥2/10 or ≥30% on NRS, while severe pain is conventionally ≥7/10.

Pain type	Typical features	Examples	Preferred strategy
Nociceptive somatic	Well localised, aching, movement-related	Bone metastases, soft tissue invasion	Opioid ± NSAID/paracetamol; radiotherapy for bone pain
Nociceptive visceral	Poorly localised, colicky or pressure-like, referred pain	Liver capsule stretch, bowel obstruction, pancreatic cancer	Opioid; antispasmodic if colic; corticosteroid if capsular oedema
Neuropathic	Burning, shooting, allodynia, hyperalgesia, dermatomal	Plexopathy, spinal cord compression, chemotherapy neuropathy	Opioid plus adjuvant: gabapentinoid, TCA, SNRI, steroid if compression/oedema
Incident/breakthrough	Transient exacerbation despite controlled background pain	Movement-related bone pain, dressing changes	Rapid-onset rescue opioid; pre-emptive dosing before predictable triggers

Principles and opioid pharmacology

The WHO analgesic ladder remains a pragmatic framework, but modern palliative practice emphasises mechanism-based prescribing and rapid titration rather than rigid stepwise delay. NICE and EAPC guidance support oral morphine as standard first-line strong opioid when renal function is adequate. Opioids have no analgesic ceiling, but toxicity—sedation, delirium, myoclonus, hallucinations, respiratory depression—limits dose escalation. Constipation is expected and requires regular stimulant ± osmotic laxative from initiation; tolerance develops to nausea/sedation more than constipation.

Drug	Starting dose in opioid-naïve adult	Key pharmacology and cautions
Immediate-release oral morphine	2.5–5 mg every 4 h, plus same dose PRN hourly; frail/elderly 1–2.5 mg	Onset 20–30 min; t1/2 2–4 h. Active metabolites M3G/M6G accumulate in renal impairment.
Modified-release oral morphine	Total effective 24 h dose divided 12-hourly	Use after titration; do not use alone for rapidly escalating pain.
Subcutaneous morphine	Approximately half the 24 h oral morphine dose via syringe driver	Useful when dying patient cannot swallow. Avoid or reduce in eGFR <30 mL/min/1.73 m2.
Oxycodone	1.25–2.5 mg PO every 4–6 h if opioid-naïve/frail; usual 5 mg	Oral potency ≈1.5–2 times morphine. Some renal accumulation; safer than morphine but still reduce dose in CKD.
Fentanyl patch	12 micrograms/h patch only if opioid-tolerant or low-dose conversion appropriate	Onset 12–24 h; offset prolonged after removal. Useful in stable pain and renal failure; unreliable in cachexia/fever.
Alfentanil SC	Specialist use; e.g. 0.5–1 mg/24 h SC, titrated	Very short t1/2 1–2 h; inactive metabolites; preferred opioid in severe renal failure.

Titration, breakthrough dosing and conversion

For continuous pain, prescribe regular opioid plus breakthrough rescue. Breakthrough oral morphine is typically 1/6 of the total 24-hour oral morphine dose, given every 1 h PRN; SC rescue is usually 1/6 of the 24-hour SC dose, often every 30–60 min PRN depending on local policy. If ≥3–4 rescue doses are required in 24 h, reassess pain mechanism and increase the baseline dose by the total rescue amount used or by 30–50% if pain remains uncontrolled and toxicity is absent. In frailty, renal impairment or rapidly changing physiology, titrate in smaller increments.

Equianalgesic approximation	Conversion	Exam cautions
Oral morphine to SC morphine	30 mg PO/24 h ≈ 15 mg SC/24 h	SC is about twice as potent.
Oral morphine to oral oxycodone	30 mg morphine ≈ 15–20 mg oxycodone/24 h	Reduce calculated dose by 25–50% if switching because of toxicity.
Oral morphine to fentanyl patch	60–90 mg/24 h ≈ 25 micrograms/h patch	Provide rescue opioid during first 12–24 h; patch unsuitable for acute titration.

Adjuvant analgesia and end-of-life considerations

Neuropathic pain frequently needs co-analgesia: gabapentin 100–300 mg nocte/tds titrated to 1.8–3.6 g/day if renal function permits; pregabalin 25–75 mg nocte/bd titrated to 150–300 mg bd; amitriptyline 10–25 mg nocte if anticholinergic burden acceptable; or duloxetine 30 mg daily then 60 mg daily, particularly chemotherapy-induced neuropathy. Corticosteroids can be highly effective for tumour oedema, capsular stretch and nerve compression; use dexamethasone 4–8 mg daily, rising to 16 mg daily for suspected malignant spinal cord compression while arranging urgent imaging/oncology input. NSAIDs may help inflammatory or bone pain but are limited by renal failure, thrombocytopenia, heart failure and gastrointestinal bleeding; prescribe gastroprotection where appropriate.

In the dying phase, convert essential analgesia to subcutaneous infusion if swallowing is unsafe: calculate the previous 24-hour opioid requirement, convert route, prescribe via syringe driver over 24 h, and ensure PRN rescue is available. Do not stop established opioids because the patient is unconscious; uncontrolled pain may manifest as grimacing, guarding, tachycardia or distress with movement. Conversely, opioid toxicity should prompt dose reduction, hydration review if consistent with goals, opioid rotation, and specialist advice rather than reflex naloxone; naloxone is reserved for life-threatening respiratory depression and is titrated cautiously, e.g. 20–40 micrograms IV every 2 min, to avoid precipitating severe pain and withdrawal.

Nausea

Clinical framing and assessment

Nausea in advanced disease is usually multifactorial and should be analysed mechanistically rather than treated with sequential “antiemetic trials”. In the last days of life, investigation is limited to reversible causes where treatment is proportionate to prognosis and goals of care; however, missing constipation, drug toxicity, hypercalcaemia, uraemia, raised intracranial pressure, gastritis, bowel obstruction, or vestibular disease commonly leads to poor control.

Assess severity with a reproducible patient-reported scale: a 0–10 numerical rating scale or the Edmonton Symptom Assessment System, where 0 is absent and 10 worst possible. Record frequency of vomiting, relation to meals, volume and character of vomitus, bowel habit, colic, abdominal distension, drug changes, renal function, calcium, and QT-risk. Persistent nausea without vomiting suggests chemical/metabolic or drug-mediated causes; large-volume effortless vomiting suggests gastric stasis or obstruction; early satiety and post-prandial fullness suggest gastroparesis; colic with feculent vomiting suggests distal bowel obstruction.

Pathophysiological classification

Mechanism	Dominant receptors	Typical clues	Preferred antiemetic strategy
Chemoreceptor trigger zone stimulation	D2, 5-HT3, opioid, dopamine pathways	Opioids, uraemia, hypercalcaemia, sepsis, hepatic failure, digoxin	Haloperidol; consider levomepromazine if refractory
Gastric stasis or impaired motility	D2, 5-HT4, vagal afferents	Early satiety, post-prandial nausea, large-volume vomiting	Metoclopramide if no complete obstruction
Bowel obstruction/peritoneal or visceral stimulation	5-HT3, H1, muscarinic, D2	Colic, distension, constipation, high-pitched bowel sounds or late silence	Antisecretory plus antiemetic; avoid prokinetic in complete obstruction
Vestibular	H1, muscarinic	Motion-related, vertigo, labyrinthine disease, posterior fossa disease	Cyclizine or hyoscine hydrobromide
Cortical/raised intracranial pressure	Multiple; inflammatory oedema-mediated	Morning vomiting, headache, papilloedema, focal neurology	Dexamethasone plus appropriate antiemetic

Pharmacological management

Use the oral route if reliable; otherwise use subcutaneous bolus or continuous subcutaneous infusion over 24 hours. Review after 24–48 hours. Avoid combining drugs with identical receptor targets unless deliberately using broad-spectrum rescue. Correct constipation and dehydration only where clinically meaningful; opioid-induced nausea often improves after 3–7 days, but persistent symptoms require treatment or opioid rotation.

Drug	Usual palliative dose	Key indications	Important cautions
Haloperidol	0.5–1 mg PO/SC once or twice daily; 1.5–5 mg/24 h by CSCI	Chemical/metabolic nausea, opioid-induced nausea	Extrapyramidal effects, Parkinsonism/Lewy body dementia, QT prolongation; half-life approximately 14–37 h
Metoclopramide	10 mg PO/SC three times daily; 30–60 mg/24 h by CSCI	Gastric stasis, functional ileus, partial obstruction without colic	Contraindicated in complete mechanical obstruction or perforation; extrapyramidal reactions; reduce in renal impairment; avoid with antimuscarinics if prokinetic effect desired
Cyclizine	50 mg PO/SC every 8 h; 150 mg/24 h by CSCI	Vestibular nausea, bowel obstruction, raised intracranial pressure adjunct	Antimuscarinic effects, sedation, delirium, urinary retention; may worsen gastric stasis
Levomepromazine	3.125–6.25 mg PO/SC nocte or PRN; 6.25–25 mg/24 h by CSCI	Refractory multifactorial nausea; broad D2, H1, muscarinic and 5-HT blockade	Marked sedation, hypotension, anticholinergic toxicity; start very low in frailty
Ondansetron	4–8 mg PO/SC twice daily; 8–16 mg/24 h by CSCI	5-HT3-mediated nausea, bowel obstruction, radiotherapy/chemotherapy-related nausea	Constipation, headache, QT prolongation; dose reduction in severe hepatic impairment, maximum 8 mg/day commonly used
Dexamethasone	4–8 mg PO/SC each morning; 8–16 mg/day for raised intracranial pressure or malignant bowel obstruction	Cerebral oedema, liver capsular stretch, bowel obstruction, tumour-related inflammation	Insomnia, delirium, proximal myopathy, hyperglycaemia; review benefit within 3–5 days and taper if ineffective

Specific end-of-life scenarios

In malignant bowel obstruction, distinguish partial from complete obstruction. Metoclopramide may reduce nausea in partial obstruction but can precipitate painful colic in complete obstruction. For complete obstruction, combine an antiemetic such as haloperidol, cyclizine, ondansetron or levomepromazine with antisecretory treatment where vomiting is prominent; octreotide 300–600 micrograms/24 h by CSCI reduces gastrointestinal secretions, and hyoscine butylbromide 60–120 mg/24 h may help colic and secretions. Dexamethasone 8–16 mg/day may transiently reduce peritumoural oedema and obstruction; discontinue if no response within about 5 days.

QT prolongation is examinable because several antiemetics are implicated. A QTc greater than 500 ms, or an increase greater than 60 ms from baseline, substantially increases torsades risk; risk is amplified by hypokalaemia, hypomagnesaemia, bradycardia, structural heart disease, macrolides, quinolones, methadone and antipsychotics. In the dying phase, ECG monitoring may be inappropriate, but rational drug choice remains essential.

Guideline-based practice, including UK palliative care formularies and Scottish Palliative Care Guidelines, emphasises mechanism-led prescribing, early subcutaneous administration when oral absorption fails, and daily reassessment. The evidence base is weaker than for chemotherapy-induced nausea; Cochrane reviews repeatedly note small heterogeneous trials, so clinical pharmacology and careful bedside phenotyping remain central to high-quality symptom control.

Breathlessness

Breathlessness in end-of-life care is a subjective experience of breathing discomfort, often poorly correlated with oxygen saturation, respiratory rate or radiographic burden. It is common in advanced cancer, COPD, heart failure, interstitial lung disease and neurodegenerative disease, and should be assessed as a multidimensional symptom incorporating intensity, distress, functional limitation and reversibility of contributors.

Assessment and mechanisms

Mechanistically, dyspnoea arises from mismatch between ventilatory drive and achieved ventilation, mediated by chemoreceptors, mechanoreceptors, pulmonary vagal afferents, chest wall receptors and cortical-limbic processing. In terminal illness, important drivers include hypoxaemia, hypercapnia, increased work of breathing, respiratory muscle weakness, pleural effusion, bronchospasm, pulmonary embolism, lymphangitis carcinomatosa, anaemia, acidosis, infection, heart failure, ascites and anxiety-amplified perception.

Examination should identify reversible causes only where treatment is consistent with goals of care. Useful bedside measures include respiratory rate, SpO₂, work of breathing, ability to speak, auscultation, signs of fluid overload, pleural effusion, bronchospasm and delirium. Severity may be documented using a 0–10 numerical rating scale, visual analogue scale, modified Borg scale, or functional tools such as the modified Medical Research Council dyspnoea scale. In dying patients, observer tools such as the Respiratory Distress Observation Scale may be used when communication is impaired.

Pattern	Likely mechanism	Palliative implication
Acute pleuritic dyspnoea, tachycardia	Pulmonary embolism, pneumothorax, effusion	Treat if reversible and burdens acceptable; otherwise symptomatic opioid-based approach
Wheeze, prolonged expiration	Bronchospasm/COPD	Nebulised bronchodilators if helpful; avoid futile escalation
Orthopnoea, crackles, oedema	Heart failure/fluid overload	Diuretics may give rapid symptomatic benefit
Cheyne–Stokes or irregular terminal breathing	Brainstem/metabolic terminal physiology	Usually does not require correction; treat observed distress

Non-pharmacological measures

First-line measures are low-burden and often effective: calm explanation, upright positioning, forward-leaning posture, pacing, reducing environmental heat, opening a window and directing cool airflow to the face. A handheld fan aimed at the trigeminal V2/V3 distribution may reduce dyspnoea via facial cooling and cortical modulation; it is recommended in multiple palliative guidelines due to favourable risk-benefit despite modest trial sizes. Physiotherapy-led breathing retraining, pursed-lip breathing and secretion clearance may help selected patients but are often impractical in the last days of life.

Oxygen: treat hypoxaemia, not breathlessness alone

Oxygen should be prescribed like a drug, with target saturations and route specified. It is clearly indicated for symptomatic hypoxaemia, commonly SpO₂ <90% or PaO₂ <8 kPa, unless burdens outweigh benefits. In COPD or chronic hypercapnic respiratory failure, target SpO₂ is usually 88–92%; otherwise 92–96% is typical. For non-hypoxaemic dyspnoea, oxygen is not superior to room air: the landmark double-blind randomised trial by Abernethy et al. in refractory dyspnoea with PaO₂ >7.3 kPa showed no additional symptomatic benefit of oxygen over medical air delivered by nasal cannula. Oxygen equipment can impair mobility, communication and discharge planning; therefore continue only if the individual reports meaningful benefit.

Opioids: cornerstone pharmacological treatment

Systemic opioids are the best-supported drug treatment for refractory breathlessness. They reduce ventilatory drive, modulate central perception of dyspnoea and improve the relationship between respiratory effort and distress. When titrated cautiously, clinically significant respiratory depression is uncommon in opioid-responsive breathlessness; sedation and constipation are more frequent dose-limiting effects. Morphine remains first line unless contraindicated. Laxatives and antiemetic prophylaxis should be considered when prognosis allows.

Clinical situation	Suggested regimen	Key cautions
Opioid-naïve, able to swallow	Morphine immediate-release 2.5 mg PO every 4 hours, with 2.5 mg PO hourly PRN; increase by 30–50% after 24 hours if needed	Use lower doses in frailty, renal impairment or marked CO2 retention
Opioid-naïve, unable to swallow	Morphine 1–2 mg SC every 4 hours or 1 mg SC hourly PRN; consider syringe driver 5–10 mg/24 h SC	Monitor sedation, myoclonus; reduce dose if eGFR <30 mL/min/1.73 m2
Already taking opioids for pain	Use breakthrough dose approximately one-sixth of total 24-hour opioid dose; increase background opioid by 25–50% if frequent dyspnoea doses required	Check whether dyspnoea reflects opioid toxicity, aspiration or infection
Renal failure	Consider oxycodone cautiously or alfentanil SC; specialist input preferred	Morphine-6-glucuronide accumulation increases sedation and neurotoxicity risk

Nebulised opioids are not recommended routinely: trials have not shown consistent superiority over placebo, and systemic dosing is more predictable. Diamorphine may be used where local practice supports it; approximate SC diamorphine:morphine equivalence is often taken as 1:2 to 1:3, but conversion should follow local formulary guidance.

Anxiolytics and palliative sedation interface

Benzodiazepines are not first-line for dyspnoea itself, but are appropriate when anxiety, panic, terminal agitation or existential distress amplifies breathlessness. Evidence for dyspnoea relief independent of anxiety is weak. Common regimens include lorazepam 0.5–1 mg SL/PO every 4–6 hours PRN, or midazolam 2.5 mg SC PRN, escalating to continuous SC infusion 5–10 mg/24 h if persistent anxiety or terminal distress. In the imminently dying patient with refractory severe breathlessness despite opioids and reversible measures, proportionate palliative sedation may be ethically and clinically appropriate, usually with midazolam titrated to relief rather than unconsciousness.

Targeted treatments when consistent with goals of care

• Bronchospasm: salbutamol 2.5–5 mg nebulised PRN; ipratropium 500 micrograms nebulised 4–6 hourly. Use air-driven nebulisers in CO₂ retainers unless oxygen is required.
• Pulmonary oedema: furosemide 20–40 mg PO/IV/SC, titrated to response and renal function.
• Superior vena cava obstruction or lymphangitis: dexamethasone commonly 8–16 mg/day PO/SC may palliate oedema/inflammatory tumour burden.
• Pleural effusion or ascites: drainage can provide rapid relief if expected benefit exceeds procedural burden.
• Severe anaemia: transfusion may help selected patients with longer prognosis, but benefit near death is often transient and logistically burdensome.

For examination purposes, the key principle is that breathlessness at the end of life is managed by structured assessment, reversal of proportionate causes, non-pharmacological airflow strategies, oxygen only when hypoxaemic or demonstrably beneficial, and carefully titrated systemic opioids, with benzodiazepines reserved for anxiety-associated or refractory terminal distress.

Agitation

Clinical framing and mechanisms

Agitation at the end of life is most often hyperactive or mixed delirium, frequently termed terminal restlessness. It is characterised by fluctuating disturbance in attention, awareness and cognition, with psychomotor activation, perceptual disturbance, sleep–wake reversal and distress. In the dying phase it is commonly multifactorial: systemic inflammation, hypoxia, organ failure, opioid neurotoxicity, anticholinergic burden, urinary retention, constipation, uncontrolled pain, sepsis, hypercalcaemia, hyponatraemia, uraemia, hepatic encephalopathy and corticosteroids are high-yield precipitants.

Pathophysiologically, delirium reflects failure of distributed attentional networks with cholinergic deficiency, dopaminergic excess, neuroinflammation, blood–brain barrier dysfunction, impaired oxidative metabolism and stress-axis activation. This explains the pharmacological rationale for dopamine antagonists, but also why antipsychotics may be ineffective or harmful when delirium is driven by irreversible dying physiology rather than dopaminergic psychosis.

Assessment: confirm delirium, quantify distress, identify reversible contributors

Assessment should be rapid, proportionate and goal-directed. In actively dying patients, extensive investigation is inappropriate unless a reversible cause is plausible and treatment aligns with the patient’s priorities. Key bedside checks include pain, bladder distension, faecal loading, hypoxia, medication toxicity and environmental distress.

Tool	Use in end-of-life agitation	Exam-relevant thresholds
4AT	Rapid delirium screen; no special training; useful on wards	Score ≥4 suggests delirium; 1–3 suggests cognitive impairment
CAM	Diagnostic framework: acute/fluctuating course, inattention, plus disorganised thinking or altered consciousness	High specificity in trained hands; less sensitive in hypoactive delirium
MDAS	Severity scale in cancer/palliative populations	Score ≥13/30 commonly used for delirium diagnosis/severity
RASS-PAL	Agitation–sedation monitoring in palliative sedation	Ranges from +4 combative to −5 unrousable

Non-pharmacological and reversible measures

First-line management is correction of remediable drivers and reduction of delirium load: stop non-essential anticholinergics, benzodiazepines, dopamine agonists, corticosteroids if feasible; review opioid dose and consider rotation if myoclonus, allodynia or cognitive toxicity suggest opioid-induced neurotoxicity. Treat urinary retention with catheterisation, constipation with rectal measures if appropriate, and uncontrolled pain promptly. Optimise sensory input, lighting, sleep–wake cues, familiar voices, calm explanation and minimisation of unnecessary observations. Family should be told that terminal agitation is usually a sign of brain and systemic failure, not wilful distress.

Drug treatment

Medication is indicated when agitation causes distress, poses risk, prevents essential care, or persists despite simple measures. Use the lowest effective dose, reassess frequently, and anticipate the need for a continuous subcutaneous infusion when repeated PRN dosing is required or oral intake is lost.

Drug	Typical end-of-life dosing	Key cautions and indications
Haloperidol	0.5–1.5 mg PO/SC stat or nocte; repeat every 2–4 h if needed. Common CSCI dose 1.5–5 mg/24 h; higher specialist doses sometimes used.	D2 antagonist; relatively non-sedating. Avoid or use great caution in Parkinson’s disease, Lewy body dementia, marked QT prolongation, severe extrapyramidal sensitivity. Oral half-life approximately 14–36 h.
Levomepromazine	6.25–12.5 mg SC PRN; CSCI 12.5–50 mg/24 h, titrated; refractory agitation may require 100–200 mg/24 h under specialist advice.	Broad D2, H1, muscarinic and alpha-1 blockade; useful when sedation is desired or haloperidol fails. Risks: profound sedation, hypotension, anticholinergic effects, QT prolongation.
Midazolam	2.5–5 mg SC/IV PRN; CSCI 10–30 mg/24 h, titrating commonly to 60 mg/24 h; specialist doses higher for refractory symptoms.	Short-acting benzodiazepine; onset SC 5–15 min, half-life about 1.5–3 h. Best for severe terminal agitation, anxiety/panic, or palliative sedation; may worsen delirium if used indiscriminately earlier.
Lorazepam	0.5–1 mg PO/SL/SC/IV PRN, usually every 4–6 h	Useful for anxiety, alcohol/benzodiazepine withdrawal, seizures, or as adjunct in refractory agitation. Less titratable than midazolam for continuous sedation.

The evidence base is nuanced. The palliative-care randomised trial by Agar et al. reported that oral risperidone and haloperidol were associated with worse delirium symptom scores and more extrapyramidal effects than placebo in mild–moderate delirium, with a signal toward poorer survival in the antipsychotic groups. This does not mean antipsychotics are never used; rather, they should not be reflexively prescribed for all delirium, and pharmacological treatment should target distress, danger or psychotic agitation, not merely abnormal cognition.

Refractory terminal agitation and palliative sedation

When agitation remains intolerable despite proportionate assessment and first-line treatment, it may be a refractory symptom. Current palliative-care frameworks, including EAPC principles and UK practice, support proportionate palliative sedation when the intention is relief of otherwise intractable suffering, not hastening death. Requirements are: senior review where possible, documentation of refractoriness, discussion with patient or proxy/family, review of clinically assisted hydration/nutrition, and regular monitoring of comfort and sedation depth.

A common regimen is midazolam 5 mg SC stat, then CSCI 10–20 mg/24 h with 2.5–5 mg SC PRN, titrated to comfort using RASS-PAL. If inadequate, increase stepwise or add levomepromazine 12.5–25 mg/24 h CSCI. Specialist palliative care input is appropriate for escalating midazolam beyond 60 mg/24 h, combined sedative regimens, suspected paradoxical benzodiazepine agitation, or complex family/ethical conflict.

Respiratory Secretions

Terminal respiratory secretions, often termed the death rattle, are noisy respirations caused by oscillation of pooled secretions in the hypopharynx, larynx, trachea or proximal bronchi as cough and swallowing reflexes fail. They occur in approximately 25–90% of dying patients, usually in the final 24–48 hours, and are more distressing to relatives and staff than to the unconscious or obtunded patient. Their presence is a poor prognostic sign, with median survival commonly measured in hours to a few days.

Mechanisms and classification

The fundamental mechanism is impaired clearance rather than excessive production. Reduced consciousness, bulbar dysfunction, opioid-related cough suppression, profound weakness and dehydration-related tenacious mucus all contribute. Antimuscarinics reduce new salivary and bronchial secretion formation but do not remove fluid already present; therefore early use is more effective than late treatment of established, copious secretions.

Type	Mechanism	Clinical clues	Implication
Type 1	Predominantly salivary/oropharyngeal pooling due to loss of swallowing	Gurgling upper airway noise; patient deeply drowsy; minimal lung signs	Most responsive to antimuscarinics and positioning
Type 2	Bronchial/pulmonary secretions from infection, aspiration, pulmonary oedema or tumour	Coarse crackles, fever, frothy sputum, fluid overload, aspiration history	Less responsive; consider whether reversible treatment is appropriate to goals of care

Assessment and staging

Assess whether the patient is actively dying, the level of consciousness, respiratory distress, hydration burden, recent fluid administration, aspiration risk, and whether symptoms are troubling the patient or primarily the family. A commonly used death-rattle scale grades audibility: 0 = absent; 1 = audible only close to the patient; 2 = audible at the foot of the bed; 3 = audible at the room door. This is useful for documenting response after 4–12 hours of treatment.

Do not reflexively investigate or treat presumed infection or pulmonary oedema unless consistent with the individualised care plan. In the last days of life, intravenous fluids may worsen secretions, pulmonary oedema and peripheral oedema; NICE guidance recommends reviewing clinically assisted hydration regularly and reducing or stopping it if causing harm.

Non-pharmacological management

• Explanation: tell relatives that the noise is common, usually not a sign of choking or distress, and reflects loss of swallowing in dying.
• Positioning: lateral or semi-prone positioning may allow passive drainage; avoid repeated repositioning if burdensome.
• Mouth care: frequent oral care relieves dryness and odour; avoid excessive oral fluids if swallowing is unsafe.
• Suction: avoid deep suctioning; it is distressing, stimulates more secretion and may cause bleeding. Gentle oral suction may be used only for visible secretions in the mouth.
• Medication review: stop nebulised saline, mucolytics or unnecessary enteral fluids if they worsen noise or aspiration.

Antimuscarinic therapy

NICE NG31 advises considering an antimuscarinic if noisy secretions are causing distress, monitoring at least every 4 hours, and stopping or switching if ineffective after about 12 hours or if adverse effects occur. Evidence for treatment of established death rattle is modest: Cochrane reviews have not shown consistent superiority of one agent over placebo or another agent. The SILENCE randomised trial showed prophylactic subcutaneous scopolamine butylbromide 20 mg four times daily reduced grade ≥2 death rattle compared with placebo, but prophylaxis is not routine for all dying patients.

Drug	Typical adult dose	Key properties	When favoured/avoided
Hyoscine butylbromide	20 mg SC stat, then 20 mg SC every 1–2 h PRN; CSCI 60–120 mg/24 h, occasionally up to 240 mg/24 h	Quaternary ammonium compound; poor blood–brain barrier penetration; onset ~30 min	Common first-line UK choice; less sedating/confusing than hyoscine hydrobromide
Glycopyrronium bromide	200 micrograms SC every 2–4 h PRN; CSCI 600–1200 micrograms/24 h, up to 2400 micrograms/24 h	Minimal CNS penetration; relatively long peripheral antisecretory effect; renal clearance	Useful if delirium risk; reduce dose or extend interval in significant renal impairment
Hyoscine hydrobromide	400 micrograms SC every 2–4 h PRN; CSCI 1.2–2.4 mg/24 h; transdermal patch 1.5 mg/72 h	Tertiary amine; crosses blood–brain barrier; sedative and potentially delirium-provoking	May help if agitation coexists; avoid if anticholinergic delirium is problematic. Patch onset is too slow for acute control
Atropine	1% ophthalmic solution sublingually: 1–2 drops every 4–6 h PRN; 1 drop ≈ 0.5 mg	Convenient non-parenteral option; variable absorption	Useful where injections are not feasible; less predictable in severe dry mouth or poor perfusion

Adverse effects are class effects: dry mouth, blurred vision, urinary retention, constipation, tachycardia, thickened tenacious secretions and, especially with centrally acting agents, hallucinations or delirium. Relative cautions include narrow-angle glaucoma, bladder outflow obstruction, paralytic ileus and tachyarrhythmias, though in the imminently dying patient proportionality and comfort take precedence over long-term risk.

In examinations, the key distinction is that respiratory secretions are managed by explanation, positioning and selective antimuscarinic use; they are not an indication for routine antibiotics, deep suctioning or automatic fluid administration. Treatment success should be judged by reduced audibility and family distress, not by complete abolition of airway noise.