Table 4.
Unanswered questions and future directions in prognostication of impending death
| Unanswered question | Explanation and example |
|---|---|
| Aims and important outcomes of prognostication | |
| 1. What are the aims of prognostication of impending death? | The ultimate goal of prognostication of impending death is to improve patient outcomes in their last days of life. However, impact studies are lacking. Investigators should start with why, and clarify the aims of prognostication. For example, do accurate prognostication and timely communication help patients and families complete unfinished business, spend the last days of life in their preferred place surrounded by their loved ones, and achieve a good death? Can prognostication help clinicians clarify patients’ goals, optimize medical treatment and care, and provide goal-concordant care in the last days of life? Or for advanced cancer patients under observation at home for COVID-19, can the timely detection of early signs of impending death or changes in vital signs by remote censoring help facilitate urgent hospitalization to prevent further deterioration? With such specific aims in mind, investigators can better develop and test prognostic tools as well as communication strategies in various settings. |
| 2. What are the clinically important outcomes? | In addition to the pursuit of accuracy of prognostic tools, clinically important outcomes of prognostication should be established. Outcome measurements may include, but are not limited to, goal-concordant care, aggressive end-of-life care, completion of unfinished business, patients’ and families’ satisfaction with care, preferred place of death, quality of dying, and a good death. These need to be clarified and validated if necessary. |
| Development of better prognostic tools | |
| 3. Validation of prognostic models | There are several prognostic models based on data from IPOD and ESAED studies. However, external validation is needed, and testing of how these models perform in a real-world setting is necessary. |
| 4. Which prognostic tools should be used in which clinical settings? | How to define and assess signs of impending death needs to be standardized. While many signs, symptoms, and tools have been suggested to inform of impending death, how to use them efficiently in settings where high sensitivity or specificity is desirable is not well-understood. Furthermore, the accuracy of clinicians’ prediction of survival in the presence of impending death needs to be compared with existing signs and tools. |
| 5. What are novel biomedical makers for the prediction of impending death? | Regardless of tumor sites, most cancer patients follow a typical, natural dying process. By investigating physiological signs in patients’ serum, electrophysiology, and imaging studies, novel biomedical markers to predict impending death may be identified. |
| 6. Is continuous monitoring of physiological signs helpful in predicting impending death? | Previous studies involved relatively infrequent assessment of clinical signs (e.g., once or twice daily), and could not identify subtle changes of vital signs and other physiologic signs sufficiently. The use of continuous monitoring may allow real-time prognostication. It may also help clinicians provide more timely symptom management and effective communication with patients and families, especially in settings with less resources such as during the night time and in home or nursing home settings. |
| 7. Can artificial intelligence help predict impending death? | Electronic medical records provide rich quantitative and qualitative data related to various signs and symptoms of impending death. Natural language processing of the medical record would help identify the point when a patient enters the dying phase before clinicians recognize it. Continuous image recognition by a bedside monitor that observes patients’ movement and facial expressions would also identify impending death without delay. This might also help clinicians remotely identify signs of impending death, and could be widely used in the era of the COVID-19 pandemic. |
| Communication of impending death | |
| 8. What are intra and inter-country differences in patients’ and families’ preferences for, and clinicians practice of, communication of impending death? | Patients’ and families’ perceptions, beliefs, and preferences regarding communication of impending death, as well as clinicians’ beliefs and practice patterns can vary widely across countries and within a country. Cross-cultural surveys involving western and eastern countries and regions may also clarify similarities and differences in communication of impending death. |
| 9. What contributes to the variable practice in communication of impending death with patients, and how can clinicians support patients without explicit prognostic disclosure? | While there are marked cross-cultural differences in communication of impending death with patients, reasons behind such differences are largely unknown. Cross-cultural qualitative studies may provide a deep understanding of different perceptions and practice patterns. In particular, better understanding of why, when, and how patients wish to be informed of their impending death may help clinicians develop individualized communication strategies. |
| 10. What is the best practice of explaining impending death to patients with various degrees of readiness? | Unlike explanation of longer life expectancy, the best way to communicate days of prognosis, either explicitly or implicitly, while tailoring to patients’ readiness and preferences has not been fully elucidated. When, how, and to whom such communication should be initiated need to be elucidated. However, there is no one-size-fits-all approach. Sequential process may be required to realize effective communication and decision-making. |
| Impact studies | |
| 11. Is a comprehensive care pathway including individualized communication strategies effective and safe, and what are the mechanisms of action? | Communication and decision-making are part of a more comprehensive care pathway for patients during the last days of life. A mixed-method, randomized controlled trial could help confirm the feasibility, effectiveness, and safety of such a care pathway and their mechanisms of action. A real-world implementation and dissemination of such care pathways may help provide quality care for imminently dying patients |
COVID-19, coronavirus disease 2019; EASED, East-Asian cross-cultural collaborative Study to Elucidate the Dying process; IPOD, Investigating the Process of Dying.