Palliative Sedation Often Follows a Cascade of Refractory Symptoms, Not a Single One

When advanced cancer reaches its final days, some patients develop suffering that no treatment can relieve. In those moments, clinicians may turn to palliative sedation: the deliberate, proportional reduction of consciousness to relieve intolerable distress. Yet the evidence base describing exactly what patients experience in the hours before sedation begins, and how their medications are handled around that transition, has remained surprisingly thin.

The research, published in Supportive Care in Cancer, comes from the PALSED study, a prospective multicentre observational investigation embedded in a European H2020 project on palliative sedation. Between April 2021 and April 2023, researchers enrolled adults with advanced cancer admitted to seven European palliative care settings, including hospital wards, hospices, and palliative care units. All patients had limited life expectancy and were judged by their treating teams to be at risk of developing refractory suffering, meaning symptoms that could no longer be controlled, or not within an acceptable timeframe.

Of 248 enrolled patients, followed across 257 admissions, 62 died without sedation being initiated, while 80 patients underwent palliative sedation before death. The central finding is striking: refractory suffering at the threshold of sedation was almost never a single-symptom problem. Only 19 percent of sedated patients had one refractory symptom.

Forty percent had two, and more than 40 percent carried three or more simultaneously. Healthcare professionals, who documented symptoms from a predefined checklist and ranked them by perceived importance, most frequently recorded exhaustion (41 percent), dyspnoea (40 percent), and pain (39 percent) as present at initiation. But when asked which symptom was the primary reason for starting sedation, dyspnoea topped the list at 23 percent, followed by delirium at 18 percent and exhaustion at 16 percent.

The co-occurrence patterns reveal how intertwined end-of-life distress becomes. Exhaustion commonly appeared alongside pain (24 percent of patients), dyspnoea (15 percent), anxiety (13 percent), and existential suffering (11 percent). Anxiety, in turn, frequently clustered with dyspnoea and pain.

Notably, neither exhaustion nor existential suffering was ever reported as the sole reason for initiating sedation. Free-text descriptions from clinicians painted exhaustion as profound physical and mental depletion, often accompanied by an inability to cope with the burden of illness, while existential suffering encompassed loss of autonomy, increasing dependency, diminished meaning of life, and fear. To capture the patient's own perspective, the team used the Edmonton Symptom Assessment System, a nine-item scale on which symptoms are rated from 0 to 10.

Assessments were available for 66 of the 80 sedated patients within three days before initiation, with a median interval of 0.9 days before the decision. Patients themselves completed 70 percent of these assessments; the rest came from proxy ratings by familiar healthcare professionals when self-report was no longer possible. The picture that emerged was of overwhelming, near-universal burden: reduced sense of wellbeing was present in 91 percent of assessments, tiredness in 88 percent, appetite loss in 82 percent, and drowsiness in 73 percent.

Severe burden, defined as a score of 7 or higher, affected 63 to 78 percent of patients across these symptoms, and every symptom except nausea was present in more than 60 percent of cases. A sensitivity analysis restricted to the 46 patient-completed assessments largely confirmed these distributions, with one telling difference: patients themselves rated depressed mood as clinically relevant, at a score of 4 or above, more often than the combined sample did, 75 percent versus 63 percent, while nausea was reported slightly less frequently. Medication use in the final 24 hours before sedation revealed a pattern that the researchers had hypothesised in advance.

Patients received a median of six medications in that window, with analgesics dominated by opioids given to 91 percent. But the composition of treatment differed sharply by symptom count. Patients with only one or two refractory symptoms more frequently received anti-ulcer agents, corticosteroids, cardiovascular drugs such as antihypertensives, diuretics and beta-blockers, and anti-infectives.

Those with three or more symptoms were instead more often treated with anxiolytics, hypnotics, and sedatives, and less frequently received the disease-management-style drugs. The authors interpret this as evidence that mounting symptom burden signals a transition toward comfort-focused care: clinicians withdraw treatments that no longer serve a dying patient, discontinue oral agents like laxatives that can no longer be swallowed, and reach for benzodiazepines to address anxiety and sleep disturbance. Once sedation began, practice converged.

Most pre-sedation medications were discontinued, but opioids were predominantly continued or adjusted, in 93 percent of patients, in line with international guidelines that distinguish sedation from the withdrawal of pain relief. Continuous midazolam was the near-universal sedative of choice, with similar starting doses between the two symptom-count groups, 2.0 versus 2.3 milligrams per hour during the first six hours. Additional or alternative sedatives were used in 32 percent of the fewer-symptom group versus 48 percent of the multiple-symptom group, and the median duration of sedation was nearly identical at roughly 36 to 37 hours.

No significant differences emerged in concomitant medications or other sedation parameters between groups. This uniformity stands in contrast to earlier reports of wide variability in sedation practice, and the authors attribute it to the fact that all participating settings had palliative care expertise and monitored the sedation process closely. What does this mean for the concept of refractoriness itself?

Some researchers have proposed numerical symptom thresholds to define when suffering becomes refractory and sedation becomes justified. The PALSED findings argue against that approach. In these patients, a self-reported collapse in overall wellbeing, often accompanied by profound exhaustion, reflected a global, multidimensional state of suffering rather than any isolated symptom crossing a severity line.

The authors are careful to stress that such a decline should not be read as a purely subjective trigger for sedation, but rather as a clinical signal prompting careful multidisciplinary evaluation of whether symptom-focused treatments have truly been exhausted. In this framing, the cluster of worsening wellbeing, fatigue, and appetite loss functions as an early warning that the dying phase is approaching, giving teams time to withdraw futile treatments and to open conversations with patients and families about end-of-life scenarios. The study has limits that temper its conclusions.

There was no comparison group of patients following other end-of-life trajectories, so the findings cannot establish how sedated patients differ from those whose symptoms remain controllable. Nor were data collected on clinicians' underlying intentions when prescribing medications before sedation, leaving open whether some drugs were already slated for discontinuation. The authors call for longitudinal mixed-methods research combining quantitative patient-reported measures with detailed pharmacological documentation, alongside qualitative work on how clinicians and patients interpret subjective states like exhaustion and existential suffering across languages and cultures.

Even so, the study delivers a clear technical message: at the doorstep of palliative sedation, suffering is a syndrome of interacting symptoms, patient-reported wellbeing collapses before any single symptom becomes dominant, and once the decision is made, sedation practice across European settings is far more consistent than the literature has suggested.

Subject of Research: Symptom burden and medication strategies before and after initiation of palliative sedation in advanced cancer patients

Article Title: Symptom burden profiles and applied medication strategies in the days before and after initiation of palliative sedation in advanced cancer patients: an international prospective multicentre observational study (PALSED)

Article References: Rijpstra, M., Kuip, E., Belar, A., Surges, S. M., Adile, C., Rojí, R., Mercadante, S., Radbruch, L., Menten, J., Centeno, C., Hasselaar, J., & Vissers, K. (2026). Symptom burden profiles and applied medication strategies in the days before and after initiation of palliative sedation in advanced cancer patients: an international prospective multicentre observational study (PALSED). Supportive Care in Cancer, 34(11), Article 1160. https://doi.org/10.1007/s00520-026-11358-7

Image Credits: AI Generated

DOI: 10.1007/s00520-026-11358-7

Keywords: palliative sedation, advanced cancer, refractory symptoms, end-of-life care, ESAS, midazolam, dyspnoea, delirium, existential suffering, palliative care, symptom clusters, observational study