Is the 1-day surprise question a useful screening tool for predicting prognosis in patients with advanced cancer?—a multicenter prospective observational study
Introduction
Accurate survival prediction of cancer patients is critical (1). It is essential for determining the goals of care among physicians, patients, and their families because many important decisions, such as those determining care plans and advance care planning depend on expected survival time (2,3). Therefore, if the responsible physicians’ prognosis is inaccurate or wrong, patients may receive unwanted and harmful treatments (4). It is even more important to predict the prognosis of cancer patients with impending death because their conditions change rapidly, and individualized care that incorporates predicting prognosis is key for providing better care to patients and their families (5). Several prognostic tools such as prognosis in palliative care study predictor models (PiPS), palliative prognostic index (PPI), and palliative prognosis (PaP) Score have been utilized to predict the prognosis of cancer patients on a weekly or monthly basis (6-10). However, these tools have not been validated as tools for predicting prognosis on daily basis.
The surprise question (SQ), “Would I be surprised if this patient died in the next 12 months?” is widely known as a practicable, simple, and helpful tool for identifying cancer patients who are at increased risk of one year mortality and would respond well to hospice and palliative care (11,12). Previous studies have shown that the 7-day SQ (7DSQ), “Would I be surprised if this patient died in the next 7 days?” is a highly sensitive and feasible way to predict the prognosis (13). Our study group had previously reported that the 3-Day SQ (3DSQ), “Would I be surprised if this patient died in the next 3 days?” was also highly sensitive in cancer patients with impending death (14).
There are few screening tools to identify cancer patients who die within one day, and we examined the possibility that the 1-Day Surprise Question (1DSQ), “Would I be surprised if this patient died in the next 1 day?” may be as helpful as the 3DSQ to predict the prognosis of cancer patients with impending death. It has been previously reported that family members were most stressed during the unpredictable death of the patient (15). If the sensitivity of the 1DSQ was as high as that of the 3DSQ, physicians and family members would be able to better prepare for rapidly changing conditions at the time of ending, and allow patients and their family members to be together at the time of death.
Therefore, our aims were to elucidate the usefulness of the 1DSQ in cancer patients with impending death, and to identify the characteristics of patients who ultimately died among those whose physicians answered “not surprised” to the 1DSQ.
We present the following article in accordance with the STROBE reporting checklist (available at https://dx.doi.org/10.21037/apm-21-1718).
Methods
Participants
The current study was a sub-analysis using Japanese domestic data as a part of the East-Asian collaborative cross-cultural Study to Elucidate the Dying process (EASED) data. The EASED was a multicenter, prospective, observational study conducted to better understand the process of death and terminal care in patients with advanced cancer admitted to palliative care units (PCUs) in Japan, Korea, and Taiwan.
Adult patients over 18 years old who were diagnosed with metastatic or locally extensive cancer and were newly admitted to PCUs were included. Patients who refused to be enrolled in this study or were scheduled to be discharged within 1 week were excluded. Consecutive patients were registered if they had been referred to the participating PCUs during the study duration.
The observations were conducted in daily clinical practice. The responsible physicians prospectively reported data on a data-collecting sheet created for this study, which was piloted prior to study initiation.
Data collection
We collected data regarding the characteristics of the patients and medical care received during the PCUs admission. The characteristics of the patients were sex, primary cancer site, metastasis, and past medical treatment as examples. The medical care received during the PCUs admission were oxygen and opioid administration, corticosteroid medication, infusion therapy, sedative therapy, administration of benzodiazepines, and administration of airway secretion inhibitors. Furthermore, we also collected data on patients’ vital signs, physical signs, and clinical symptoms on the first day, when the patient had palliative performance scale (PPS) ≤20. The vital signs were respiratory rate, oxygen saturation of peripheral artery, urine output, and body temperature. The physical signs were modified Richmond Agitation Sedation Scale score (modified RASS), pulse of radial artery, peripheral cyanosis, bronchial secretions, respiration with mandibular movement, Eastern Cooperative Oncology Group Performance Status, dysphagia of liquids, response to visual stimuli, hyperextension of neck, response to verbal stimuli, grunting of vocal cords, inability to close eyelids, apnea periods, and Cheyne-Stokes breathing. The modified RASS has been validated as a tool for assessing the severity of agitation and sedation levels in cancer patients (16). The clinical symptoms were dyspnea, pain, fatigue, delirium, and edema. The clinical symptoms were assessed using the integrated palliative care outcome scale (IPOS). The IPOS of Japanese version has been demonstrated to be valid for assessing the physical and psychological status of patients with cancer (17).
All factors were chosen as representative factors, which we considered to be prognostic in daily clinical practice.
Measurements
We defined “day 0” as the first day on which each patient had PPS ≤20; physicians’ answers to the 1DSQ, “Would I be surprised if this patient died in the next 1 day?” were collected on this day. The same responsible physician, who collected the patient information, including physical signs and clinical symptoms, answered the 1DSQ. The response was categorized as “surprised” or “not surprised”. We followed up the patients until death, and defined “died in the next 1 day” as death having occurred between day 0 and day 1.
Data analysis and statistics
First, to summarize the patients’ baseline characteristics, we performed descriptive analyses.
Second, the patients were placed in the “surprised” and “not surprised” groups according to the 1DSQ answers of responsible physicians. We also added the patients’ state: alive or dead, on day 1 and created a 2×2 contingency table based on these results. We used simple statistical analysis to compute the sensitivity, specificity, positive predictive value and negative predictive value based on this table.
Third, to clarify factors of patients who died within the next 1 day as those whom the physicians answered “not surprised” to the 1DSQ, we categorized the patients into two groups. Patients whose physicians answered “not surprised” and died within the next 1 day were categorized group A and defined as the “predictable death group”. Patients whose physicians answered “not surprised” but did not die within the next 1 day were categorized as group B.
Fourth, we performed a Fisher’s exact test for categorical variables and a Cochran-Armitage trend test for ordinal variables to clarify patients’ factors related to the “Predictable death group”.
Fifth, we put variables with P˂0.05 derived from univariate analyses into a multivariate logistic regression analysis to identify the patients’ factors related to the “predictable death group”.
We performed statistical analyses using JMP Pro version 14 (SAS, Cary, NC, USA).
Ethical statement
The present study was conducted in accordance with the ethical standards of the Declaration of Helsinki (as revised in 2013) and the ethical guidelines for medical and health research involving human subjects presented by the Ministry of Health, Labour, and Welfare in Japan, and was approved by the local institutional review boards of all participating institutions. All procedures were in accordance with the ethical standards of the independent ethics committee of Tohoku University School of Medicine (approval No. 2016-1-689). Japanese law does not require individual informed consent from participants in a non-invasive observational trial, such as in the present study. Therefore, we used an opt-out method rather than acquiring written or oral informed consent.
Results
A total of 1,896 patients were enrolled from 22 PCUs in Japan between January and December 2017. Among these, 485 patients were excluded: 245 patients were discharged from the hospital alive and 240 patients had missing data on day 0. The remaining 1,411 patients were analyzed (Figure 1).
We summarized the baseline characteristics of 1,411 patients in Table 1. The mean (standard deviation) age was 72.6 (12.2) years, and the most common primary cancer site was the lungs (17.1%).
Table 1
Characteristics (n=1,411) | No. (%) |
---|---|
Age (years), mean (SD) [range] | 72.6 (12.2) [25–100] |
Sex (male) | 716 (50.7) |
Primary cancer site | |
Lung | 242 (17.1) |
Stomach/esophagus | 207 (14.7) |
Colon/rectum | 186 (13.1) |
Pancreas | 146 (10.3) |
Liver/biliary system | 121 (8.5) |
Prostate/bladder/kidney/testis | 102 (7.2) |
Ovary/uterus | 82 (5.8) |
Others | 325 (23.0) |
Metastatic site | |
Liver | 565 (40.0) |
Lung | 530 (37.5) |
Bone | 380 (26.9) |
Cancer treatment | |
Chemotherapy | 866 (61.3) |
Surgery | 594 (42.0) |
Hormonal therapy | 14 (0.9) |
Radiation therapy | 11 (0.7) |
Eastern cooperative oncology group performance status | |
0–1 | 8 (0.6) |
2 | 79 (5.6) |
3 | 549 (38.9) |
4 | 775 (54.9) |
Palliative performance scale | |
20 or less | 326 (23,1) |
30 | 296 (21.0) |
40 | 410 (29.0) |
50 | 281 (20.0) |
60 and above | 98 (6.9) |
Median survival time (days), mean (SD) [range] | 16.0 (29.9) [0–375] |
Based on (14). SD, standard deviation.
We indicate a 2×2 contingency table (Table 2). The responsible physicians answered “not surprised” for 847 (60.0%) patients. This prediction showed a sensitivity of 82.0% [95% confidence interval (CI): 77.5–85.8%] and specificity of 45.5% (95% CI: 44.4–46.4%). The positive predictive value was 27.4% (95% CI: 25.9–28.7%), and the negative predictive value was 91.0% (95% CI: 88.7–92.9%).
Table 2
Group | Death within 1 day | Not death within 1 day | Sensitivity | Specificity | Positive predictive value | Negative predictive value |
---|---|---|---|---|---|---|
Not surprised | 232 (group A) | 615 (group B) | 82.0% (95% CI: 77.5–85.8%) | 45.5% (95% CI: 44.4–46.4%) | 27.4% (95% CI: 25.9–28.7%) | 91.0% (95% CI: 88.7–92.9%) |
Surprised | 51 | 513 |
Group A: patients whose physicians answered “not surprised” and actually died in the next 1 day. We defined group A as “predictable death group”; group B: patients whose physicians answered “not surprised” and did not actually die in the next 1 day. CI, confidence interval.
Table S1 showed the results of all variables for which univariate analysis was performed to clarify factors related to the “predictable death group”, and Table 3 summarizes 17 variables with P˂0.05. The 17 variables were urine output over last 12 h <100 mL (P<0.01), decreased response to verbal stimuli (P<0.01), oxygen administration (P<0.01), decreased response to visual stimuli (P<0.01), radial artery (P<0.01), dysphagia of liquids (P<0.01), peripheral cyanosis (P<0.01), respiration with mandibular movement (P<0.01), saturation of percutaneous oxygen (SpO2) (P<0.01), hyperextension of neck (P=0.02), grunting of vocal cords (P=0.02), dyspnea (P=0.02), infusion therapy (P=0.02), age (P=0.03), sex (P=0.04), inability to close eyelids (P=0.04), and opioid administration (P=0.04).
Table 3
Variables | Subgroup | Total (n=847), n | Predictable death group (group A) | Group B | P value | |||
---|---|---|---|---|---|---|---|---|
n | % | n | % | |||||
Age | 20–69 year | 328 | 104 | 31.7 | 224 | 68.3 | 0.03 | |
70 and above year | 519 | 128 | 24.7 | 391 | 75.3 | |||
Sex | Male | 429 | 104 | 24.2 | 325 | 75.8 | 0.04 | |
Female | 418 | 128 | 30.6 | 290 | 69.4 | |||
Decreased response to verbal stimuli | No | 648 | 150 | 23.1 | 498 | 76.9 | <0.01 | |
Yes | 186 | 69 | 37.1 | 117 | 62.9 | |||
Decreased response to visual stimuli | No | 549 | 116 | 21.1 | 433 | 78.9 | <0.01 | |
Yes | 285 | 103 | 36.1 | 182 | 63.9 | |||
Dysphagia of liquid | Absent | 444 | 169 | 38.1 | 275 | 61.9 | <0.01 | |
Present | 952 | 446 | 46.9 | 506 | 53.1 | |||
Peripheral cyanosis | Absent | 638 | 142 | 22.3 | 496 | 77.7 | <0.01 | |
Present | 196 | 77 | 39.3 | 119 | 60.7 | |||
Pulselessness of radial artery | Absent | 769 | 184 | 23.9 | 585 | 76.0 | <0.01 | |
Present | 65 | 35 | 53.9 | 30 | 46.1 | |||
Respiration with mandibular movement | Absent | 784 | 186 | 23.7 | 598 | 76.3 | <0.01 | |
Present | 50 | 33 | 66.0 | 17 | 34.0 | |||
Hyperextension of neck | Absent | 789 | 200 | 25.4 | 589 | 74.6 | 0.02 | |
Present | 45 | 19 | 42.2 | 26 | 57.8 | |||
Inability to close eyelids | Absent | 773 | 196 | 25.4 | 577 | 74.6 | 0.04 | |
Present | 61 | 23 | 37.7 | 38 | 62.3 | |||
Grunting of vocal cords | Absent | 793 | 201 | 25.4 | 592 | 74.6 | 0.02 | |
Present | 41 | 18 | 43.9 | 23 | 56.1 | |||
Urine output over last 12 hour | 200 mL and above | 630 | 132 | 21.0 | 498 | 79.0 | <0.01 | |
200 mL or less | 204 | 87 | 42.7 | 117 | 57.3 | |||
Dyspnea (IPOS) | 0–1 | 493 | 115 | 23.3 | 378 | 76.7 | 0.02 | |
2–4 | 341 | 104 | 30.5 | 237 | 69.5 | |||
SpO2 | 90% and above | 92 | 48 | 52.2 | 44 | 47.8 | <0.01 | |
89% or less | 707 | 158 | 22.4 | 549 | 77.6 | |||
Oxygen administration | Absent | 327 | 63 | 19.3 | 264 | 80.7 | <0.01 | |
Present | 507 | 156 | 30.8 | 351 | 69.2 | |||
Opioid administration | Absent | 188 | 38 | 20.2 | 150 | 79.8 | 0.04 | |
Present | 646 | 181 | 28.0 | 465 | 72.0 | |||
Infusion therapy | Absent | 331 | 102 | 30.8 | 229 | 69.2 | 0.02 | |
Present | 503 | 117 | 23.3 | 386 | 76.7 |
Group A: patients whose physicians answered “not surprised” and actually died in the next 1 day. We defined group A as “predictable death group”; group B: patients whose physicians answered “not surprised” and did not actually die in the next 1 day. IPOS, integrated palliative care outcome scale; SpO2, saturation of percutaneous oxygen.
Table 4 lists the results of the multivariate analysis. We found that five factors associated with the “predictable death group”. The five factors were urine output over the last 12 hours [<100 vs. ≥100 mL; odds ratio (OR) 2.11; 95% CI: 1.41–3.12; P<0.01], decreased response to visual stimuli (present vs. absent; OR 1.71; 95% CI: 1.03–2.82; P=0.04), respiration with mandibular movement (present vs. absent; OR 2.69; 95% CI: 1.28–5.63; P=0.01), radial artery (pulselessness vs. palpable; OR 2.32; 95% CI: 1.21–4.46; P=0.01), and SpO2 (<89% vs. ≥90%; OR 3.20; 95% CI: 1.95–5.26; P<0.01).
Table 4
Variables | Subgroup | OR (95% CI) | P value |
---|---|---|---|
Decreased response to visual stimuli | Yes | 1.710 (1.03–2.82) | 0.04 |
No | Ref | ||
Radial artery | Pulselessness | 2.325 (1.21–4.46) | 0.01 |
Palpable | Ref | ||
Respiration with mandibular movement | Present | 2.695 (1.28–5.63) | 0.01 |
Absent | Ref | ||
Urine output over last 12 hour | 200 mL or less | 2.107 (1.41–3.12) | <0.01 |
200 mL and above | Ref | ||
SpO2 | 89% or less | 3.204 (1.95–5.26) | <0.01 |
90% and above | Ref |
OR, odds ratio; CI, confidence interval; SpO2, saturation of percutaneous oxygen.
Discussion
This is the first study to verify the usefulness of the 1DSQ as a tool for estimating the prognosis of cancer patients with impending death using a large group.
Our analyses revealed that the 1DSQ has a high sensitivity in predicting death in cancer patients within 1 day and may be useful as a screening tool in this context. White et al. reported the mean sensitivity of SQs to predict the prognosis cancer patients within 12 months was 77.1% (18). In comparison, the 1DSQ has a higher sensitivity (82.0%). This may be attributed to the standardization of the question when a patient had PPS ≤20. Most of palliative physicians recognize that patients with PPS ≤20 have a poorer prognosis, which may have caused a physician to respond with “not surprised”. When physicians answered “not surprised” to the 1DSQ, the physicians should carefully explain to the family member the possibility of the patient dying within 1 day. We believe that using the 1DSQ to carefully explain the prognosis to the patient’s family will significantly improve psychological care for those affected.
However, compared to other studies that used SQs for short-term prognosis, the 1DSQ had a slightly lower sensitivity; Hamano et al. reported that the sensitivity and specificity of the 30-day surprise question were 95.6% and 37.0%, respectively, and 7-day surprise question were 84.7% and 68.0%, respectively (13). In addition, our previous findings revealed a sensitivity and specificity of 94.3% and 26.3% for the 3DSQ, respectively (14). One possible explanation for the discrepancy in sensitivity may be that physicians could not confidently answer “not surprised” to cancer patients who died within 1 day. It is known that physicians refer to the physical signs and clinical symptoms of cancer patients with impending death when predicting their prognosis (5). Hui et al. reported on 16 clinical signs that often occur between 1.5 and 5.5 days prior to death in cancer patients (19). Among them, the clinical sign that occurred most often within 1 day before death was only “pulselessness of radial artery (19-21)”. The lack of clarity in the physical signs and clinical symptoms of cancer patients who die within 1 day may cause physicians to hesitate in making their decisions. However, we believe that the 1DSQ is a valuable screening tools to identify cancer patients who die within one day, as it has a high sensitivity while there are few accurate physical signs and clinical symptoms that can predict death within one day. In addition, the 1DSQ can be conveniently performed at a patient’s bedside. The usefulness of the 1DSQ may be further enhanced as more research is conducted on the clinical signs and symptoms that often occur immediately before death.
We identified five variables associated with the factors of patients who were likely to die within 1 day, as the physicians predicted. These were urine output over last 12 hours <100 mL, decreased response to visual stimuli, respiration with mandibular movement, pulselessness of radial artery, and SpO2 <90%. These signs have been previously reported to occur in cancer patients 3 days before death (5,19-22). In particular, urine output over last 12 hours <100 mL, respiration with mandibular movement, and pulselessness of radial artery have been shown to be the signs that most likely occur 1-1.5 days before death (20). If physicians answer the 1DSQ with “not surprised” and notice these clinical signs, the patients’ condition should be observed more carefully.
In our previous study investigating the 3DSQ, adjunctive medical treatment, such as opioid administration and continuous deep sedation, was identified as a factor to assist in predicting prognosis (14). However, this study only presented the patients’ physical signs. The result of this study suggests that the physical signs may be more important for predicting which patients will die within 1 day.
This study has some limitations. First, physical signs and clinical symptoms of patients might influence the physician’s answer to the 1DSQ. However, as the SQ is a tool that relies on the physician’s intuition, we believe that it is unavoidable in this study and in clinical practice. We consider that the 1DSQ is not a screening tool to be used alone, but rather should be used in conjunction with physical signs and symptoms. Among these, we revealed the five signs that may be more helpful when combined in this study. Second, the 1DSQ was answered by palliative care physicians; non-palliative care physicians may have provided different answers and thus yielded different results. Further studies involving non-palliative care physicians should be conducted to clarify this possibility. Third, the physicians may have been influenced by other prognostic tools such as PiPS models, PPI, or PaP score when answering the 1DSQ. However, we considered this effect to be unclear as these tools have not been optimized for predicting the prognosis of cancer patients with impending death (6-8). Fourth, the medical care provided in PCUs may differ from that provided in a general ward. However, the medical care provided to patients with PPS ≤20 was expected to be almost identical, and thus, it might not be a determining factor for this study. Fifth, we lack information regarding the attending physicians, such as age, gender, and years of experience, which might have influenced our results.
Conclusions
Based on our findings, the 1DSQ has proven to be a helpful screening tool for identifying cancer patients with impending death.
Acknowledgments
We would like to thank Editage (
Funding: This work was supported in part by a Grant-in-Aid from the Japanese Hospice Palliative Care Foundation (Grant Numbers 16H05212 and 16KT0007).
Footnote
Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://dx.doi.org/10.21037/apm-21-1718
Data Sharing Statement: Available at https://dx.doi.org/10.21037/apm-21-1718
Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://dx.doi.org/10.21037/apm-21-1718). The authors have no conflicts of interest to declare.
Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The present study was conducted in accordance with the ethical standards of the Declaration of Helsinki (as revised in 2013) and the ethical guidelines for medical and health research involving human subjects presented by the Ministry of Health, Labour, and Welfare in Japan, and was approved by the local institutional review boards of all participating institutions. All procedures were in accordance with the ethical standards of the independent ethics committee of Tohoku University School of Medicine (approval No. 2016-1-689). Japanese law does not require individual informed consent from participants in a non-invasive observational trial, such as in the present study. Therefore, we used an opt-out method rather than acquiring written or oral informed consent.
Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.
References
- White N, Reid F, Vickerstaff V, et al. Specialist palliative medicine physicians and nurses accuracy at predicting imminent death (within 72 hours): a short report. BMJ Support Palliat Care 2020;10:209-12. [Crossref] [PubMed]
- Hui D, Nooruddin Z, Didwaniya N, et al. Concepts and definitions for "actively dying," "end of life," "terminally ill," "terminal care," and "transition of care": a systematic review. J Pain Symptom Manage 2014;47:77-89. [Crossref] [PubMed]
- Hwang IC, Ahn HY, Park SM, et al. Clinical changes in terminally ill cancer patients and death within 48 h: when should we refer patients to a separate room? Support Care Cancer 2013;21:835-40. [Crossref] [PubMed]
- Ellershaw J, Ward C. Care of the dying patient: the last hours or days of life. BMJ 2003;326:30-4. [Crossref] [PubMed]
- Domeisen Benedetti F, Ostgathe C, Clark J, et al. International palliative care experts’ view on phenomena indicating the last hours and days of life. Support Care Cancer 2013;21:1509-17. [Crossref] [PubMed]
- Maltoni M, Nanni O, Pirovano M, et al. Successful validation of the palliative prognostic score in terminally ill cancer patients. Italian Multicenter Study Group on Palliative Care. J Pain Symptom Manage 1999;17:240-7. [Crossref] [PubMed]
- Morita T, Tsunoda J, Inoue S, et al. The Palliative Prognostic Index: a scoring system for survival prediction of terminally ill cancer patients. Support Care Cancer 1999;7:128-33. [Crossref] [PubMed]
- Gwilliam B, Keeley V, Todd C, et al. Development of prognosis in palliative care study (PiPS) predictor models to improve prognostication in advanced cancer: prospective cohort study. BMJ 2011;343:d4920. [Crossref] [PubMed]
- Billings JA, Bernacki R. Strategic targeting of advance care planning interventions: the Goldilocks phenomenon. JAMA Intern Med 2014;174:620-4. [Crossref] [PubMed]
- Baba M, Maeda I, Morita T, et al. Survival prediction for advanced cancer patients in the real world: A comparison of the Palliative Prognostic Score, Delirium-Palliative Prognostic Score, Palliative Prognostic Index and modified Prognosis in Palliative Care Study predictor model. Eur J Cancer 2015;51:1618-29. [Crossref] [PubMed]
- Moss AH, Lunney JR, Culp S, et al. Prognostic significance of the "surprise" question in cancer patients. J Palliat Med 2010;13:837-40. [Crossref] [PubMed]
- Moroni M, Zocchi D, Bolognesi D, et al. The ‘surprise’ question in advanced cancer patients: A prospective study among general practitioners. Palliat Med 2014;28:959-64. [Crossref] [PubMed]
- Hamano J, Morita T, Inoue S, et al. Surprise Questions for Survival Prediction in Patients With Advanced Cancer: A Multicenter Prospective Cohort Study. Oncologist 2015;20:839-44. [Crossref] [PubMed]
- Ikari T, Hiratsuka Y, Yamaguchi T, et al. "3-Day Surprise Question" to predict prognosis of advanced cancer patients with impending death: Multicenter prospective observational study. Cancer Med 2021;10:1018-26. [Crossref] [PubMed]
- Barry LC, Kasl SV, Prigerson HG. Psychiatric disorders among bereaved persons: the role of perceived circumstances of death and preparedness for death. Am J Geriatr Psychiatry 2002;10:447-57. [Crossref] [PubMed]
- Benítez-Rosario MA, Castillo-Padrós M, Garrido-Bernet B, et al. Appropriateness and reliability testing of the modified Richmond Agitation-Sedation Scale in Spanish patients with advanced cancer. J Pain Symptom Manage 2013;45:1112-9. [Crossref] [PubMed]
- Sakurai H, Miyashita M, Imai K, et al. Validation of the Integrated Palliative care Outcome Scale (IPOS) - Japanese Version. Jpn J Clin Oncol 2019;49:257-62. [Crossref] [PubMed]
- White N, Kupeli N, Vickerstaff V, et al. How accurate is the ‘Surprise Question’ at identifying patients at the end of life? A systematic review and meta-analysis. BMC Med 2017;15:139. [Crossref] [PubMed]
- Hui D, Hess K, dos Santos R, et al. A diagnostic model for impending death in cancer patients: Preliminary report. Cancer 2015;121:3914-21. [Crossref] [PubMed]
- Hui D, dos Santos R, Chisholm G, et al. Clinical signs of impending death in cancer patients. Oncologist 2014;19:681-7. [Crossref] [PubMed]
- Hui D, Dos Santos R, Chisholm G, et al. Bedside clinical signs associated with impending death in patients with advanced cancer: preliminary findings of a prospective, longitudinal cohort study. Cancer 2015;121:960-7. [Crossref] [PubMed]
- Bruera S, Chisholm G, Dos Santos R, et al. Variations in vital signs in the last days of life in patients with advanced cancer. J Pain Symptom Manage 2014;48:510-7. [Crossref] [PubMed]