A new study has found that simple, low-cost nudges can significantly increase the number of serious illness conversations between oncologists and patients with a poor prognosis, offering a practical way to improve end-of-life care in a system where such discussions remain dangerously rare.
The research comes as a poignant reminder of the challenges doctors face. One oncologist, reflecting on a recent encounter with a long-term cancer survivor, described the delicate balance of honesty and empathy required. The patient, who has undergone an array of cancer treatments over ten years, began the consultation by asking about the doctor’s children — a gesture of goodwill that made the difficult conversation that followed even harder. The oncologist knew each new treatment offered a little less benefit, but shifting a patient’s thinking when they still feel well is extraordinarily difficult. The doctor gently raised the need to plan for a time when health worsens, for the sake of the patient and her family. At the end, the patient ducked her head back in to say: “Thank you, I needed to hear that.” The oncologist exhaled and went back to do it all over again.
The nudge experiment
The study that inspired this reflection tested two simple nudges designed to prompt serious illness conversations. Researchers assigned patients and their oncologists to one of four groups: a control group receiving no nudge; a group in which the patient received a letter encouraging them to think about these issues; a group in which the oncologist received an email before the appointment suggesting the patient might need such a conversation; and a group receiving both nudges.
The researchers then looked for evidence of a serious illness conversation in two ways. First, they checked the dedicated advance care planning section of the medical records. Second, because doctors might record the conversation elsewhere, an AI algorithm scanned the entire medical record.
The results, published in the original study, revealed a clear pattern. When only the dedicated advance care planning section was examined, nudging the patient alone made no difference: just 10% had a documented conversation, the same rate as the control group. A clinician nudge raised that figure to 16%, and a combined nudge to both clinician and patient brought it to 17%.
When the entire medical record was scanned, the numbers improved across all groups but the pattern held. Again, nudging the patient alone had no effect: 22% of both the control and patient-nudged groups had documented conversations. Nudging the clinician increased the figure to 28%, and nudging both clinician and patient produced the highest rate of 32%.
Barriers and implications
The study’s findings carry several important lessons. Serious illness conversations are not happening anywhere near as often as warranted, at an enormous cost to both the patient and society. Even when they do occur, documentation is often inadequate — a problem that becomes critical when patients are met in extremis by paramedics or emergency professionals who lack time to mine the records. As the oncologist noted, there is something to be said for the time-honoured tradition of putting an advance care directive on the fridge door.
The research also confirms an old finding: telling patients to have difficult conversations with their oncologist does not work. What does work is reminding doctors to have them, and reminding both doctors and patients at the same time is even better.
Yet significant barriers remain. For clinicians, these include inadequate communication training, time constraints, and burnout. In the UK, the Serious Illness Care Programme — a collaboration between the Palliative Care Institute Liverpool, Warrington and Halton Teaching Hospitals NHS Foundation Trust, and Ariadne Labs — aims to improve the experience of seriously ill patients through structured conversations, but conventional communication training has been criticised for not being specific enough for real-world interactions. Some studies suggest communication skills can even worsen without specific training, though learner-centred approaches based on performance assessment have shown effectiveness.
Burnout is a major issue among UK doctors, particularly younger ones, and leads to reduced empathy, poorer communication, and even defensive medicine. Compassion fatigue, a state of emotional exhaustion from prolonged exposure to suffering, is especially prevalent among GPs. The system itself also poses obstacles: the oncologist noted that considering how often appointments are mis-scheduled, matching the right doctor with the right patient on the right day for a consequential conversation about death and dying remains a formidable challenge.
The stakes are high. In the UK, an estimated 100,000 people who could benefit from palliative care die without receiving it each year. Demand is rising due to an ageing population and an increase in multiple long-term conditions. Hospitals remain the most common place of death — 42.8% in 2023 — followed by home (28.4%) and care homes (21.0%). Specialist palliative care at home costs around £145 per day, compared to £425 for hospital care, yet a substantial portion of end-of-life healthcare spending goes to emergency hospital care. Better advance care planning, including legally binding advance decisions to refuse treatment and written advance statements of wishes, could help shift care to more appropriate and less costly settings.
The oncologist’s patient nodded thoughtfully and left. Then she returned to say thank you. The doctor exhaled and went back out to do it all over again.