Background: For patients with kidney failure, hemodialysis is lifeprolonging, but not curative. The treatment can relieve symptoms but may also cause other symptoms and complications. The annual mortality is high: 15-20%. Consequently, there is a need for an integrated palliative care approach where conversations in serious illness are crucial. However, these conversations tend to be avoided or come too late. One reason is difficulties in identifying optimal time points to initiate these conversations.

The surprise question (SQ) – “Would I be surprised if this patient died within xx months?” is the most common way to identify patients with a potential need of conversation in serious illness. The evaluation of the SQ is mostly based on physicians’ responses to the SQ, with variations in sensitivity and accuracy. Moreover, little is known about what nurses and physicians base their judgment on when they respond to the SQ and how responses to the SQ are associated with patients’ self-reported health-related quality of life (HRQoL). In addition, there is a need to better understand why conversations in serious illness are not initiated in a timely fashion. A better understanding of nurses’ and physicians’ experiences in how to conduct conversations in serious illness may help to increase the understanding of why these conversations are not conducted as frequently as they should.

Aim: The overall aim was to explore how to identify patients on hemodialysis who are approaching the end of life. An additional aim was to study nurses’ and physicians’ experiences of conversations in serious illness.

Method: Studies I-III are all based on data from the same project involving nine hemodialysis units in three regions in Sweden. The data was collected between January 2020 and December 2023. Study I included 361 patients for whom nurses and physicians responded to the SQ. Patient clinical characteristics, performance status, and comorbidities were obtained. Study II included 282 patients who had responded to a HRQoL measure and for whom nurses and physicians responded to the SQ. Study III included 442 patients for whom nurses and physicians had responded to the SQ and nurses had also assessed patient performance status. Study IV included interviews with 11 nurses and seven physicians.

Results: The results showed that the proportions of patients for whom nurses and physicians responded “No, not surprised” to the SQ were similar. However, there was a substantial difference concerning which specific patient the nurses and physicians responded “No, not surprised” for.

Results showed an association between nurses’ responses to the SQ and the patient’s age, albumin, performance status, self-reported worsened health compared to one year ago, and lower perceived overall health. For physicians, results showed an association between their responses to the SQ and patient age, albumin, performance status, comorbidities, Kt/V (dialysis efficacy), hemoglobin, parathyroid hormone, overall health, and physical functioning. No association was found between patient self-reported pain, general health, fatigue, and emotional and social aspects and responses to the SQ.

Furthermore, results showed that nurses and physicians identified a similar number of patients who died within 12 months. Combining responses to the SQ from nurses and physicians regarding the 12-month timeframe identified most patients who died within 12 months. Results also showed that ECOG performance status offered the possibility to identify patients who would die within 12 months. Finally, the overall theme of nurses’ and physicians’ experiences of conversations involving end-of-life issues was: “balancing between the sense of responsibility for communication involving end-of-life issues and not harming the patient”.

Conclusions: To identify patients approaching the end of life to initiate conversations in serious illness is complex. It involves various aspects such as timing and type of measures, as well as differences in perspectives from nurses and physicians. Combining nurses’ and physicians’ responses to the SQ with awareness of time frames, considering patients’ age and performance status, but also comorbidity and albumin seem to strengthen the identification of patients approaching the end of life. Acknowledging the need for different types of conversations, at different time points, by different professions is important to reach a shared understanding with the patients and their significant others. Altogether, this aligns with the need for a comprehensive view of the patient’s state and the need for team collaboration anchored in the palliative care approach.