Ageing has been recognized as one of the most critically important health-care issues worldwide. It is relevant to Asia, where the increasing number of older populations has drawn attention to the paramount need for health-care investment, particularly in end-of-life care. The advocacy of advance care planning is a mean to honor patient autonomy. Since most East Asian countries are influenced by Confucianism and the concept of 'filial piety,' patient autonomy is consequently subordinate to family values and physician authority. The dominance from family members and physicians during a patient's end-of-life decision-making is recognized as a cultural feature in Asia. Physicians often disclose the patient's poor prognosis and corresponding treatment options to the male, family member rather to the patient him/herself. In order to address this ethical and practical dilemma, the concept of 'relational autonomy' and the collectivism paradigm might be ideally used to assist Asian people, especially older adults, to share their preferences on future care and decision-making on certain clinical situations with their families and important others. In this review article, we invited experts in end-of-life care from Hong Kong, Indonesia, Japan, South Korea, Singapore and Taiwan to briefly report the current status of advance care planning in each country from policy, legal and clinical perspectives. According to the Asian experiences, we have seen different models of advance care planning implementation. The Asian Delphi Taskforce for advance care planning is currently undertaken by six Asian countries and a more detailed, culturally sensitive whitepaper will be published in the near future.
Background: The characteristics of physician communication with patients at the end of life (EOL) in East Asia have not been well studied. We investigated physicians' communications with imminently dying patients with cancer and their families in palliative care units (PCUs) in Japan, South Korea, and Taiwan.
Methods: This observational study included patients with cancer newly admitted and deceased during their first admission to 39 PCUs in three countries. We evaluated 1) the prevalence and timing of informing patients and families of patients' impending death and 2) the prevalence of communication to assure the families of the patient's comfort.
Results: We analyzed 2138 patients (Japan: 1633, South Korea: 256, Taiwan: 249). Fewer Japanese (4.8%: 95% confidence interval [95% CI], 3.8%–5.9%) and South Korean (19.6%: 95% CI, 15.2%–25.0%) patients were informed of their impending death, whereas 66.4% (95% CI, 60.2%–72.1%) of Taiwanese were informed; among all three countries, =90% of families were informed. Although most patients in all three countries and the families in South Korea and Taiwan were informed of the impending death greater than or equal to four days before death, 62.1% (95% CI, 59.6%–64.6%) of Japanese families were informed less than or equal to three days prior. Most families in all three countries received assurance that the patient would remain comfortable (could hear until death, no distress with death rattle or respiration with mandibular movement).
Conclusions: Physicians in Taiwan communicated about patient's impending death most frequently, and physicians in all three countries generally provided assurance to families that the patients would remain comfortable. Further studies should explore the reasons for these differences and the effects of such communications in East Asia.
BACKGROUND: Antimicrobials are frequently prescribed to terminally-ill patients with cancer; however, physicians' practice patterns regarding antimicrobial use vary widely. This study aimed to systematically identify factors that determine physicians' attitudes toward the management of infectious diseases in terminally-ill patients with cancer.
METHODS: A nationwide cross-sectional survey involving 600 oncologists, 600 infectious diseases physicians, 600 palliative care physicians, and 220 home care physicians was conducted between November 2017 and January 2018. The primary endpoint was physicians' attitudes toward the management of infectious diseases in terminally-ill patients with cancer with a few weeks of prognosis. Physicians' beliefs regarding management of infectious diseases as well as physician-perceived 'good death' were also assessed (1=strongly disagree-6=strongly agree).
RESULTS: There were 895 (44.3%) analyzable response, and average scores of physicians' attitudes ranged between 2.69 and 4.32 In total, 241 (27%; 95%CI=24-30) to 691 (78%; 95%CI=75-81) respondents showed proactive attitudes toward various infectious diseases management. In linear regression analysis, determinants of proactive attitudes included: physicians' belief that examination and treatment will improve quality of life and prognosis and reduce suffering (ß=0.32, t=9.99, p=0.00); greater physician-perceived importance on receiving enough treatment (ß=0.09, t=2.88, p=0.00) and less importance on dying a natural death (ß=-0.07, t=-2.14, p=0.03) for a 'good death'; working at a tertiary care hospital (ß=0.16, t=4.40, p=0.00); and not being a home care physician (ß=-0.20, t=-5.51, p=0.00) or palliative care physician (ß=-0.12, t=-3.64, p=0.00).
CONCLUSIONS: Physicians have divergent attitudes toward the management of infectious diseases in terminally-ill patients with cancer. Reflection by physicians on their own beliefs and perceptions regarding infectious disease management and a 'good death' may help provide the best end-of-life care.
Death rattle occurs during the last days of life, and relatives of those afflicted frequently report that it is very distressful. However, there is no effective treatment for it. The purpose of this study was to investigate the perceptions of Japanese palliative care physicians in clinical practice in Japan. We conducted a nationwide survey of 268 physicians via an anonymous, self-report questionnaire. We assessed pharmacological and non-pharmacological management and anticholinergic agent choice. One hundred eighty-nine physicians (70.5%) returned the questionnaires. Fifty-five participants (29.1%) treating patients with Type-1 (real death rattle) and 36 participants (19%) treating patients with Type-2 (pseudo-death rattle) death rattle reported that they would frequently administer an anticholinergic agent. One-fourth would administer scopolamine butylbromide or scopolamine hydrobromide. In conclusion, more Japanese palliative care physicians thought that anticholinergic agents might be effective for treating Type-1 death rattle rather than Type-2. Further clinical trials of these agents are needed.
OBJECTIVES: It is unclear whether patients with non-specific dyspnoea are suitable candidates for studies investigating the effectiveness of benzodiazepines against dyspnoea. The objective of this survey was to investigate suitable subjects for studies of benzodiazepines for cancer dyspnoea.
METHODS: A nationwide questionnaire survey was conducted among 536 Japanese-certified palliative care physicians. We randomly selected 268 physicians and inquired about their approach to dyspnoea management in patients with cancer, with and without anxiety, as follows: (1) Administration of a benzodiazepine. (2) Administration or titration of an opioid. We also asked them to consider their approach in the following situations: (1) Opioid-naïve. (2) Low-to-moderate baseline opioid dose. (3) High baseline opioid dose. We assessed the use of specific benzodiazepines separately.
RESULTS: Overall, 192 physicians responded to the questionnaire (71.6%). For patients without anxiety, the proportion of participants reporting that they frequently or very frequently 'administer a benzodiazepine' increased with baseline opioid dose (opioid-naïve: 5.2%, low-to-moderate: 11.5%, high: 26.0%). The proportion of participants reporting that they frequently or very frequently 'administer or titrate an opioid' decreased with baseline opioid dose (opioid-naïve: 83.3%, low-to-moderate: 73.4%, high: 41.1%). The pattern was similar for patients with anxiety, although more respondents said they prescribe benzodiazepine for these patients (naïve: 22.4%, low-to-moderate: 34.4%, high: 45.8%) and fewer prescribed an opioid. Alprazolam and lorazepam are frequently used.
CONCLUSION: Patients with anxiety or receiving a high baseline opioid dose could be potential candidates for future studies investigating the effectiveness of benzodiazepines against cancer dyspnoea.
Purpose: Parenteral morphine is widely used for dyspnea of imminently dying cancer patients, but the outcomes to expect over time remain largely unknown. We examined outcomes after the administration of parenteral morphine infusion over 48 h in cancer patients with a poor performance status.
Methods: This was a multicenter prospective observational study. Inclusion criteria were metastatic/locally advanced cancer, ECOG performance status = 3–4, a dyspnea intensity = 2 on a Support Team Assessment Schedule, Japanese version (STAS-J), and receiving specialized palliative care. After initiating parenteral morphine infusion, we measured dyspnea STAS-J as well as Memorial Delirium Assessment Scale (MDAS), item 9, and Communication Capacity Scale (CCS), item 4, every 6 h over 48 h.
Results: We enrolled 167 patients (median survival = 4 days). The mean age was 70 years, 80 patients (48%) had lung cancer, and 109 (65%) had lung metastases. The mean STAS-J scores decreased from 3.1 (95% confidence interval (CI) = 3.0–3.2) at the baseline to 2.1 (95%CI = 1.9–2.2) at 6 h, and remained 1.6–1.8 over 12–48 h. The proportion of patients with dyspnea relief (STAS-J = 1) increased to 39% at 6 h, and ranged between 49 and 61% over 12–48 h. In contrast, up to 6.6 and 20% of patients showed hyperactive delirium (MDAS item 9 = 2) and an inability to communicate (CCS item 4 = 3), respectively, over 48 h.
Conclusions: Overall, terminal dyspnea was relatively well controlled with parenteral morphine, though a significant number of patients continued to suffer from dyspnea. Future efforts are needed to improve outcomes following standardized dyspnea treatment using patient-reported outcomes for imminently dying patients.
CONTEXT: Morphine is recommended as the first-line pharmacological therapy for cancer dyspnea. However, the detailed practice of morphine has not been evaluated, and consensus about other opioids for cancer dyspnea has not been established.
OBJECTIVES: To explore the physician-reported practice of opioid for cancer dyspnea.
METHODS: Nationwide mail-questionnaire survey was conducted among 536 Japanese certified palliative care physicians. We randomly selected 268 and asked the following: 1) how the physicians themselves initiate and use morphine for cancer dyspnea, 2) opioid choice for dyspnea in patients who have already used opioid other than morphine regularly, and 3) opioid choice for dyspnea in patients with various degrees of renal impairment in their daily practice.
RESULTS: Overall, 192 physicians responded (response rate, 71.6%). The major (58.3%) practice of initiating morphine was "immediate-release morphine as needed" in opioid-naïve patients and the mean % increase when they titrate morphine for cancer dyspnea was 29.4±11.3% of the baseline dose. Although "titrate baseline oxycodone" was the most frequent (42.3%) for low-to-moderate-dose regular oxycodone cases, "stepwise switch to morphine" (30.0%) and "add morphine on baseline oxycodone" (27.1%) were the more frequent practices for high-dose regular oxycodone. Regardless of the baseline dose, "add morphine on baseline fentanyl" was the most frequent practice for regular transdermal fentanyl cases. Oxycodone was the most frequent choice in renal insufficiency cases, regardless of its degree.
CONCLUSIONS: Among Japanese palliative care physicians, using oxycodone for cancer dyspnea was relatively popular practice, while fentanyl was not. Oxycodone was the most preferred opioid for cancer dyspnea in the setting of renal insufficiency among Japanese palliative care physicians. We should conduct studies to confirm the safety and effectiveness of these opioid practices for cancer dyspnea.
CONTEXT: When discussing end-of-life issues with cancer patients, the addition of reassurance statements is considered helpful. However, patients' preferences for such statements have not been systematically demonstrated.
OBJECTIVES: The objectives of this study were to clarify if phrases with additional reassurance statements would be more preferable to phrases without them and explore variables associated with patients' preferences.
METHODS: In a cross-sectional survey, 412 cancer patients assessed their own preferences for phrases with/without additional statements using a six-point scale (1 = not at all preferable; 6 = very preferable). These included the statements of "hope for the best and prepare for the worst" ("hope/prepare") when discussing prognosis; symptom palliation when discussing code status; and specific goals, continuity of care, and nonabandonment when discussing hospice referral. We evaluated demographic data and the coping style and conducted multivariate regression analysis.
RESULTS: Compared with the phrase of life expectancy (i.e., median + typical range) alone [mean (SD), 3.5 (1.2); 95% CI, 3.4-3.6], the phrase with the additional "hope/prepare" statement was more preferable [3.8 (1.4); 3.7-3.9]. Compared with the phrase of do-not-resuscitate alone (3.1(1.3); 3.0-3.3), the phrase with the additional statement of symptom palliation was more preferable [3.9 (1.3); 3.7-4.0]. Compared with the phrase of hospice referral alone [3.4 (1.2); 3.3-3.5], phrases with the addition of a specific goal [3.9 (1.0); 3.8-4.0], specific goal and continuity (4.4(1.0); 4.3-4.5), and specific goal, continuity, and nonabandonment [4.8 (1.2); 4.7-4.9] were more preferable. In multivariate analyses, task-oriented coping was significantly correlated with preferences for phrases including additional reassurance statements.
CONCLUSION: Cancer patients systematically preferred reassurance statements. In end-of-life discussions, especially with patients with task-oriented coping, clinicians may provide additional reassurance statements.
Trousseau syndrome was first described by Armand Trousseau in 1865 and is characterized by hypercoagulation resulting from malignant tumors. This complication can markedly impact quality of life (QOL). This is the first report of a terminally ill patient who developed large-vessel occlusion stroke from Trousseau syndrome and underwent mechanical thrombectomy. A 75-year-old woman presented with stage IV ovarian cancer. Goals of care were transitioned to palliative care. The patient was hospitalized with vertebral compression fracture, and suddenly developed right hemiparesis and total aphasia during admission. Magnetic resonance imaging and angiography showed occlusion of segment 1 of the left middle cerebral artery. We administered tissue-plasminogen activator, but symptoms remained unimproved. We performed mechanical thrombectomy based on medical indications and with the consent of her family. Thrombectomy improved symptoms dramatically. She was able to walk and talk with her family at discharge. She eventually died of respiratory failure on postoperative day 98, but QOL remained high for those 98 days. Mechanical thrombectomy has the potential to markedly improve QOL in terminally ill patients with large-vessel occlusion associated with Trousseau syndrome.
PURPOSE: Although little improvement has been made in the survival rate among young cancer patients over recent decades, whether they have achieved a good death has never been systematically explored. We aimed to clarify whether young cancer patients (aged 20-39 years) have achieved a good death, and compare their achievement with that of middle-aged patients (aged 40-64 years).
METHODS: We analyzed combined data of three nationwide, cross-sectional surveys of families of cancer patients who died at inpatient hospices in Japan (2007-2014). We measured 10 core items of the Good Death Inventory (GDI) short-version on a 7-point scale, and calculated rates of "agree/absolutely agree" and the mean scores.
RESULTS: We analyzed 245 and 5140 responses of families of young and middle-aged patients, respectively. Less than 60% of families of young patients reported “agree/absolutely agree” regarding 9 items, which included “feeling that one’s life was completed” in 44 (18%; 95% confidence interval (CI) = 14–23%), “being independent in daily life” in 48 (20%; 95% CI = 15–25%), and “being free from physical distress” in 103 (42%; 95% CI = 36–48%) young patients. Young patients were significantly less likely to feel “one’s life was completed” (mean = 3.3 (standard deviation = 2.0) vs. 3.8 (1.9), respectively; effect size (ES) = 0.29; adjusted p value = 0.000) and “not being a burden to others” (3.1 (1.5) vs. 3.5 (1.6), respectively; ES = 0.24; adjusted p value = 0.010) than the middle-aged.
CONCLUSIONS: Overall, young cancer patients did not achieve a good death. Future efforts are needed to improve the quality of palliative care for young patients, focusing on psychosocial/spiritual suffering.
OBJECTIVES: Family conflict has several adverse impacts on caregivers. Thus, there is significant value in determining the prevalence and predictors of family conflict, which can enable the health care provider to intervene if family conflict arises during end-of-life care. Accordingly, we aimed to explore the prevalence and predictors of conflict among the families of patients with advanced cancer who died in palliative care units.
METHODS: This study was a nationwide multicenter questionnaire survey of bereaved family members of cancer patients who died in Japanese palliative care units participating in evaluation of the quality of end-of-life care.
RESULTS: We sent out 764 questionnaires, and 529 questionnaires (69.2%) were returned. As 70 family members refused to participate and we could not identify the answers in one questionnaire, we analyzed a total of 458 responses. The average Outcome-Family Conflict score was 13.5 ± 4.9 (maximum score: 39.5), and 42.2% of family members reported at least one family conflict during end-of-life care. Greater family conflict was significantly associated with younger family age, with family members asserting control over decision making for patient care and with communication constraints among family members, although absent family members "coming out of the woodwork" reduced family conflict.
CONCLUSIONS: Many families of patients with advanced cancer experienced conflict during end-of-life care. Family members asserting control over decision making and communication constraints among family members after diagnosis of cancer can predict the occurrence of family conflict. Absent family members "coming out of the woodwork" might reduce family conflict in particular cultures.
OBJECTIVES: Explanation about the impending death of imminently dying patients with cancer is important for their families. However, little is known about how clinicians explain impending death and how families perceive the explanation. We aimed to clarify bereaved families' perception of the need for improvements in the explanation about impending death and to explore the factors contributing to the need.
METHODS: In a nationwide survey of 818 bereaved families of patients with cancer admitted to inpatient hospices in Japan, we evaluated family-perceived need for improvements in the explanation about impending death and families' experiences of the explanation.
RESULTS: Among all the participants (n=516, 63%), 35 (6.8%), 123 (24%) and 297 (58%) families felt that much/considerable, some and no improvements were needed, respectively. Independent determinants of the need were a younger patient age (OR=0.97; 95% CI 0.95 to 0.99; P=0.009); not receiving an 'explicit explanation about physical signs of impending death' (OR=0.67; 95% CI 0.51 to 0.88; P=0.004); not receiving an 'explanation of how long the patient and family could talk' (OR=0.67; 95% CI 0.51 to 0.88; P<0.001); receiving an 'excessive warning of impending death' (OR=1.45; 95% CI 1.03 to 2.03; P=0.033) and having a feeling of 'uncertainty caused by vague explanations about future changes' (OR=1.77; 95% CI 1.38 to 2.27; P<0.001).
CONCLUSIONS: Nearly a third of the bereaved families perceived some need to improve the explanation about impending death. To better help patients/families prepare for their end-of-life, clinicians should recognise and explain various impending death signs; find a balance between detailed explanation and excessive warning and address how long they could talk in the remaining time.
PURPOSE: This study investigated the effect of two types of palliative sedation defined using intervention protocols: proportional and deep sedation.
METHODS: We retrospectively analyzed prospectively recorded data of consecutive cancer patients who received the continuous infusion of midazolam in a palliative care unit. Attending physicians chose the sedation protocol based on each patient’s wish, symptom severity, prognosis, and refractoriness of suffering. The primary endpoint was a treatment goal achievement at 4 h: in proportional sedation, the achievement of symptom relief (Support Team Assessment Schedule (STAS) = 1) and absence of agitation (modified Richmond Agitation-Sedation Scale (RASS) = 0) and in deep sedation, the achievement of deep sedation (RASS = - 4). Secondary endpoints included mean scores of STAS and RASS, deep sedation as a result, and adverse events.
Among 398 patients who died during the period, 32 received proportional and 18 received deep sedation. The treatment goal achievement rate was 68.8% (22/32, 95% confidence interval 52.7–84.9) in the proportional sedation group vs. 83.3% (15/18, 66.1–100) in the deep sedation group. STAS decreased from 3.8 to 0.8 with proportional sedation at 4 h vs. 3.7 to 0.3 with deep sedation; RASS decreased from + 1.2 to - 1.7 vs. + 1.4 to - 3.7, respectively. Deep sedation was needed as a result in 31.3% (10/32) of the proportional sedation group. No fatal events that were considered as probably or definitely related to the intervention occurred.
CONCLUSION: The two types of intervention protocol well reflected the treatment intention and expected outcomes. Further, large-scale cohort studies are promising.
BACKGROUND: Actions in preparation for death and talks about death between advanced cancer patients and their families are considered essential to achieve a good death. However, little is known about the prevalence of such actions compared with talks and their association with bereaved families' psychological morbidity.
OBJECTIVE: To clarify the prevalence of bereaved families having acted in preparation for death and talked about death with their loved one, and to explore their associations with bereaved families' depression and complicated grief (CG).
DESIGN: A nationwide survey. Setting/Subject: A total of 999 bereaved families of cancer patients admitted to 133 inpatient hospices in Japan.
MEASUREMENTS: The prevalence of families' actions in preparation for and talks about death, Patient Health Questionnaire (PHQ)-9, and Brief Grief Questionnaire (BGQ).
RESULTS: Among 678 bereaved families (response rate = 68%), 513 (76%) acted in preparation for death, and 315 (46%) talked about death with their loved one. Those who acted and talked were significantly less likely to suffer depression (PHQ-9 = 10) than those who neither acted nor talked (odds ratio [OR], 0.405; 95% confidence interval [CI], 0.195–0.845; adjusted p = 0.016). Families who acted were significantly less likely to suffer complicated grief (CG; BGQ =8), whether they talked (OR, 0.394; 95% CI, 0.185–0.84; adjusted p = 0.016) or not (OR, 0.421; 95% CI, 0.191–0.925; adjusted p = 0.031).
CONCLUSIONS: Most families acted in preparation for death, and those who acted were less likely to suffer depression and CG. Clinicians may minimize families' later psychological morbidity by helping patients and families act in preparation for death.
CONTEXT: Acculturation is the phenomenon of the attitudinal changes of individuals who come into continuous contact with another culture. Despite the long history of Japanese immigration to America, little is known about the impact of acculturation on perceptions of a good death.
OBJECTIVES: To examine differences in perceptions of a good cancer death among Japanese Americans (JA/A), Japanese living in America (J/A), and the Japanese living in Japan (J/J).
Methods: We administered surveys among JA/A and J/A, and used historical J/J data for reference. Primary endpoint was the proportion of respondents who expressed the necessity of core and optional items of the Good Death Inventory. Group differences =20% were deemed clinically important.
Results: 441 survey responses in America and 2,548 in Japan were obtained. More than 80% of respondents consistently considered 9 of 10 core items necessary without significant group differences. No core item reached a =20% group difference. Three of the 8 optional items reached =20% group difference: ‘fighting against disease until one’s last moment’ (49%, p<0.0001; 52%, p<0.0001; and 73% in JA/A; J/A; and J/J; respectively), ‘knowing what to expect about one’s condition in the future’ (83%, p<0.0001; 80%, p<0.0001; and 58%, respectively), and ‘having faith’ (64%, p=0.0548; 43%, p=0.0127; and 38%, respectively).
Conclusions: While most core items of a good death were preserved throughout the levels of acculturation, perceptions of some optional items shifted away from Japanese attitudes as individuals became more acculturated. Understanding of different levels of acculturation may help clinicians provide culturally-sensitive end-of-life care.
CONTEXT: Talking about death is an important issue for terminally-ill cancer patients and their families. Little is known about how often and which bereaved families regret not having talked about death with their deceased loved one.
OBJECTIVES: To explore the prevalence of a regret of not having talked about death with a deceased loved one among bereaved family members of adult cancer patients, and to systematically explore factors contributing to their regret.
METHODS: We conducted a nationwide survey of 999 bereaved families of cancer patients admitted to 133 inpatient hospices in Japan and surveyed families' regret on talking about death. Exploratory analyses identified the underlying structures of process, option, and outcome subscales of factors contributing to regret.
RESULTS: Among 678 bereaved families (response rate, 68%), 224 (33%) regretted not having talked about death sufficiently, while 40 (5.9%) conversely regretted having talked about death. Three process factors ( "prognostic disclosure to patient" (beta=0.082, p=0.039), "upsetting of patient and family" (beta=0.127, p=0.001), and "family's sense of uncertainty about when to act based on terminal awareness" (beta=0.141, p=0.000)) and an outcome factor ("having achieved a good death" (beta=-0.152, p=0.000)) contributed to the regret of talking insufficiently.
CONCLUSION: A third of bereaved families of adult cancer patients regretted not having talked about death sufficiently. Clinicians may minimize this regret by facilitating a shared understanding of the disease and prognosis, advising families explicitly when to talk based on terminal awareness, providing continuous emotional support, and validating their decision on talking about death.
Purpose : Palliative care referral is primarily based on clinician judgment, contributing to highly variable access. Standardized criteria to trigger automatic referral have been proposed, but it remains unclear how best to apply them in practice. The authors conducted a Delphi study of international experts to identify a consensus for the use of standardized criteria to trigger automatic referral.
Methods : Sixty international experts stated their level of agreement for 14 statements regarding the use of clinician-based referral and automatic referral over two Delphi rounds. A consensus was defined as an agreement of =70% a priori.
Results : The response rate was 59/60 (98%) for the first round and 56/60 (93%) for the second round. Twenty-six (43%), 19 (32%), and 11 (18%) respondents were from North America, Asia/Australia, and Europe, respectively. The panel reached consensus that outpatient palliative care referral should be based on both automatic referral and clinician-based referral (agreement = 86%). Only 18% felt that referral should be clinician-based alone, and only 7% agreed that referral should be based on automatic referral only. There was consensus that automatic referral criteria may increase the number of referrals (agreement = 98%), facilitate earlier palliative care access, and help administrators to set benchmarks for quality improvement (agreement = 86%).
Conclusions : Our panelists favored the combination of automatic referral to augment clinician-based referral. This integrated referral framework may inform policy and program development.
Context: End-of-life discussions are crucial for providing appropriate care to patients with advanced cancer at the end of their lives.
Objectives: The objective of this study was to explore associations between end-of-life discussions and bereaved families’ depression and complicated grief and the quality of patient death and end-of-life care.
Methods: A nationwide questionnaire survey of bereaved family members was conducted between May and July 2014. A total of 13,711 bereaved family members of cancer patients who were cared for by specialist palliative care services at 75 institutions throughout Japan and died before January 2014 participated. We evaluated the prevalence of depression (defined as the Patient Health Questionnaire-9 = 10) and complicated grief (defined as the Brief Grief Questionnaire = 8) in bereaved family members. Moreover, we evaluated the quality of death and end-of-life care with the Good Death Inventory and the Care Evaluation Scale, respectively.
Results: A total of 9123 questionnaires were returned (response rate 67%), and 80.6% of the respondents reported that they had end-of-life discussions. After propensity score–weighted adjustment, the results showed that bereaved family members who had end-of-life discussions had a lower frequently of depression (17.3% vs. 21.6%; P < 0.001) and complicated grief (13.7% vs. 15.9%; P = 0.03). End-of-life discussions were associated with better quality of death (the Good Death Inventory score, 47.2 ± 8.5 vs. 46.1 ± 9.4; P < 0.001) and end-of-life care (the Care Evaluation Scale score, 84.1 ± 11.4 vs. 78.9 ± 14.3; P < 0.001).
Conclusion: End-of-life discussions may contribute to reducing depression and complicated grief in bereaved family members and enable patients to experience quality end-of-life care and a good death.