Part XV · Supportive care, survivorship, and patient experience · Chapter 88

Palliative and end-of-life care

What the palliative care trials actually showed, and why the survival claim is the weakest part of the case.

1 · Early integration of palliative care and its evidence

The argument for early palliative care is usually made with a single trial and a single number. That is a mistake, because the number is the least durable part of the evidence. The landmark trial randomised patients with metastatic non-small-cell lung cancer to early palliative care integrated with oncology or to usual care. Quality of life and mood improved, less aggressive care was given at the end of life, and median survival was longer at 11.6 against 8.9 months¹. The survival finding was secondary, was in one disease, and has not reproduced consistently.

A nurse-led psychoeducational intervention in 322 patients with advanced cancer improved quality of life and depressed mood without changing symptom intensity, hospital days or emergency visits². A cluster-randomised trial of 24 oncology clinics missed its primary endpoint at three months. The spiritual well-being difference was 3.56 points with P equal to 0.07. Several secondary endpoints were met at four months³.

The honest synthesis comes from the meta-analysis of 43 randomised trials in 12,731 patients. Quality of life and symptom burden improved at one to three months. When analysis was restricted to the five trials at low risk of bias the quality-of-life effect shrank but held, and the symptom effect lost significance. There was no survival association at all, with a hazard ratio of 0.90 and a confidence interval from 0.69 to 1.17⁴.

2 · Pain and complex symptom management

Breast cancer produces a specific and recurring symptom set. Bone metastasis dominates it. The skeletal biology that drives it is in Bone, liver, lung, and brain microenvironments compared, and its clinical management is in Bone metastases and skeletal-related events. Bone pain is usually mixed, with an inflammatory component that responds to anti-inflammatory drugs and radiation, and a mechanical component that responds to neither. A patient whose pain appears only on weight bearing has a structural problem. Escalating opioids will not fix a structural problem, and single-fraction radiotherapy or stabilisation might.

Neuropathic pain in this population usually has a name. Taxane and platinum neuropathy is covered in Acute and chronic cytotoxic toxicity. Brachial plexopathy after axillary radiation, post-mastectomy pain syndrome, and leptomeningeal involvement each present differently and each has a different management path. Treating them all as generic neuropathic pain is how patients end up on gabapentinoids that do nothing.

Aromatase inhibitor arthralgia is not cancer pain and does not respond to cancer pain treatment. It is the commonest reason for non-adherence in the adjuvant setting, which is discussed in Endocrine therapy.

Dyspnoea from lymphangitic spread, malignant effusion or treatment-related pneumonitis needs the cause separated from the symptom. Low-dose opioids reduce the sensation of breathlessness whatever the cause. Drainage, steroids or antibiotics address only some causes. Do both.

3 · Goals of care and treatment burden

Treatment burden is measurable and is almost never measured. It includes infusion hours, travel, the cost of parking, unpaid leave, childcare during appointments, and the cognitive load of managing a schedule. A regimen that adds two months of progression-free survival and requires weekly infusion has a different value in a patient who lives forty minutes from the centre than in one who lives ten.

The field's standard framing is a choice between treatment options. The real choice usually includes an option nobody named, which is less treatment with the same follow-up intensity. Naming it changes the conversation, because a patient who has never been told that stopping is a legitimate choice will not raise it.

A useful discipline is to state, for each option, what it is expected to do, what it will cost the patient in time and function, and what happens if they decline it. The third of these is the one clinicians skip. A patient who does not know what declining looks like is not making a choice.

4 · Serious illness communication

Serious illness communication is a technical skill with a literature, not a matter of temperament. Its core finding is that the conversation itself does not cause the harm clinicians fear. In a prospective cohort of 332 patient-caregiver dyads, 37.0% of patients reported an end-of-life discussion before baseline. Those discussions were not associated with higher rates of major depressive disorder, at 8.3% against 5.8%⁵.

They were associated with substantially less aggressive care. Ventilation fell from 11.0% to 1.6%, resuscitation from 6.7% to 0.8%, and intensive care admission from 12.4% to 4.1%. Hospice enrolment rose from 44.5% to 65.6%⁵. More aggressive care near death was associated with worse patient quality of life and with a higher risk of major depressive disorder in bereaved caregivers, with an adjusted odds ratio of 3.37⁵.

The structure that works is consistent across curricula. Ask what the patient already understands. Ask permission before giving information. Give the headline in one sentence and then stop talking. Respond to the emotion before returning to the facts, because information given during an emotional response is not retained. Make a recommendation rather than laying out a menu.

The sentence most clinicians get wrong is the prognosis sentence. A range attached to a reference class is honest and usable. "Some people in your situation live months, some live a couple of years, and I cannot tell you which you are" says more than a median.

5 · Hospice, end-of-life decisions, and disparities in their timing

Hospice length of stay is the metric that matters, and it is short. A referral in the last few days converts hospice into a place of death rather than a model of care. The benefit of hospice accrues over weeks, through symptom control, caregiver support and avoided hospitalisation. Three days does not deliver it.

The disparity in end-of-life care is real, and a recent analysis clarifies where it does not come from. In a propensity-matched cohort of 27,862 older adults with metastatic cancer, specialist palliative care improved outcomes in both Black and non-Hispanic White patients. Hospice use rose from 52.4% to 71.0% in Black patients and from 63.2% to 80.4% in White patients. Hospice enrolment of three days or longer rose from 44.0% to 65.5% and from 52.6% to 74.2% respectively. There was no interaction between race and the effect of palliative care on any outcome⁶.

That result is worth stating carefully. Palliative care works equally well in both groups. The gap persists because access and referral differ, not because the intervention performs differently. This separates a delivery problem from an efficacy problem, and the two have different solutions. The wider structural analysis is in Disparities across the breast cancer continuum.

6 · Bereavement and caregiver support

Caregiver outcomes are a legitimate endpoint, and the data link them to the care the patient received. Bereaved caregivers of patients who had aggressive care near death had more than three times the odds of major depressive disorder⁵. Patient quality of life and caregiver quality of life at follow-up were correlated. The two are not independent systems.

The meta-analysis found caregiver findings inconsistent across trials⁴. That inconsistency is partly a measurement problem. Caregiver burden, caregiver depression, caregiver quality of life and bereavement adjustment are four different constructs measured on different instruments at different times.

Practically, caregiver support in breast cancer has features specific to the disease. The median age at metastatic diagnosis means many caregivers are partners of working age or adult children with children of their own. A substantial minority of patients are caring for young children themselves, which inverts the usual caregiving direction and is handled poorly by services designed around an elderly patient. Anticipatory grief in a patient who has been on treatment for years is different from grief after a short illness.

7 · Clinician moral distress and sustaining this work

Moral distress is the specific experience of knowing what the right action is and being unable to take it. It is not the same as burnout, though it feeds it. In breast oncology it has recognisable sources. A drug is not covered. A patient wants treatment that will not help. A patient who would benefit cannot travel. A trial excludes the patient who most needs it.

Naming the mechanism matters because the remedies differ. Burnout responds partly to workload and control. Moral distress responds to whether the institution gives clinicians a route to act on what they judge to be right, and to whether the constraint is acknowledged rather than rationalised.

The discipline that sustains this work over a career is not detachment. It is accuracy about what is in your control. You control whether the conversation happened, whether the referral was made, whether the symptom was addressed, and whether the patient understood their options. You do not control the disease biology in Part VI, the insurance determination, or the outcome. Clinicians who hold themselves responsible for the second list burn out. Clinicians who stop caring about the first list become the reason patients feel abandoned.

References

All 6 references

1. Temel JS, Greer JA, Muzikansky A, et al. Early palliative care for patients with metastatic non-small-cell lung cancer. N Engl J Med 2010 363:733-742. PMID 20818875
2. Bakitas M, Lyons KD, Hegel MT, et al. Effects of a palliative care intervention on clinical outcomes in patients with advanced cancer: the Project ENABLE II randomized controlled trial. JAMA 2009 302:741-749. PMID 19690306
3. Zimmermann C, Swami N, Krzyzanowska M, et al. Early palliative care for patients with advanced cancer: a cluster-randomised controlled trial. Lancet 2014 383:1721-1730. PMID 24559581
4. Kavalieratos D, Corbelli J, Zhang D, et al. Association between palliative care and patient and caregiver outcomes: a systematic review and meta-analysis. JAMA 2016 316:2104-2114. PMID 27893131
5. Wright AA, Zhang B, Ray A, et al. Associations between end-of-life discussions, patient mental health, medical care near death, and caregiver bereavement adjustment. JAMA 2008 300:1665-1673. PMID 18840840
6. Guo L, Murali KP, Morrison RS, Hua M. Heterogeneity of effect of specialist palliative care for Black and non-Hispanic White older adults with metastatic cancer. J Am Geriatr Soc 2025 73:3190-3195. PMID 40566741