Preferred and Actual Places of Death Among Patients with Advanced Cancer: A Single-centre Cohort Study in Japan

Study design

This study was a follow-up of a previous cross-sectional study conducted at a single centre in Japan. The parent study investigated the association between physician communication and prognostic awareness in patients with unresectable/recurrent solid cancers.^[13,14] During the recruitment phase of the parent study, the researcher in charge relocated to another hospital, which limited our ability to reach the target sample size of 250. Patients with advanced cancer who underwent chemotherapy at Tohoku University Hospital between January 2015 and January 2016 were enrolled and followed up for 5 years.

Patients included in the study were aged ≥20 years, had unresectable/recurrent solid cancer or had failed first-line chemotherapy. The following exclusion criteria were applied: (1) unwillingness to provide written informed consent, (2) cognitive impairment, (3) unconfirmed lesions by imaging, (4) lack of awareness of the diagnosis, (5) inability to understand or complete a questionnaire in Japanese and (6) unsuitability to participate in the opinion of the primary attending physician. Eligibility was initially evaluated using the medical records. The primary attending physician, defined as the medical oncologist who examined the patient most frequently, was consulted to confirm eligibility and allow the researcher to contact the patient.

Both inpatients and outpatients were eligible for enrolment in the study. The limited number of available researchers necessitated the enrolment of patients by convenience to some extent; potentially eligible patients were sometimes excluded if several were identified on the same day or for other practical reasons. The primary attending physician obtained informed consent from each patient.

Collection of data

The enrolled patients were surveyed using a questionnaire developed specifically for the study and were required to complete and return the questionnaire to our office within 2 weeks of enrolment. We developed the questionnaire based on three previous reports^[15-17] and included questions that had been used in prospective longitudinal studies in the United States^[15,18,19] after applying the back-translation method. First, the principal investigators examined the validity of the professionally translated Japanese statements. Second, the revised translations were sent back to the translator for retranslation into English. Finally, we confirmed the uniformity of the questions with two authors of the three previous papers.^[15,18,19]

Information on the following items was sought in the patient survey: Preference for treatment (prolongation of life or palliation), hopes for participating in a clinical trial, preference for place of death, whether they had an end-of-life discussion, awareness of their terminal status (understanding of how unlikely it was that their cancer would be cured),^[19] desire for prognostic information, the potential for cure (how likely patients thought chemotherapy would cure their cancer),^[17] understanding of the purpose of chemotherapy, understanding of physicians’ goals when administering chemotherapy,^[19] characterisation of communication with physicians, score on the Patient Health Questionnaire (PHQ)-9, which has been validated in Japan^[20] as a screening tool for depression, the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire Core 30 (EORTC-QLQ-C30), which has been validated in Japan^[21] to assess the quality of life.

End-of-life discussion was defined as a discussion on palliative care or a do-not-attempt-resuscitate order. We used the commonly accepted definition of palliative care,^[1,22] specifically (1) an approach that improves the quality of life for patients and their families facing the problems associated with life-threatening illness; (2) treatment of pain and other problems, which may be physical, psychosocial or spiritual and (3) care provided in the patient’s own home or a palliative care unit (PCU).

We calculated a summary score for communication with the primary attending physician^[23] by asking patients five questions; ‘How often did your doctors listen carefully to you?’ ‘How often did your doctors explain things in a way you could understand?’ ‘How often did your doctors give you as much information as you wanted about your cancer treatments, including potential benefits and side effects?’ ‘How often did your doctors encourage you to ask all the cancer-related questions you had?’ and ‘How often did your doctors treat you with courtesy and respect?’ The response options for each question were ‘Always’ (score 3), ‘Usually’ (score 2), ‘Sometimes’ (score 1) and ‘Never’ (score 0). We then calculated the total score, with 0 being the worst possible total score and 15 being the best. The PHQ-9 scores were calculated based on a previous study and ranged from 0 to 27 points. The EORTC-QLQ-C30 scores were transformed into a 0–100-point scale according to the scoring manual.^[24]

We also obtained baseline patient data from the medical records, including age, sex, primary cancer site, Eastern Cooperative Oncology Group performance status, employment status, marital status, highest education level and number of family members living at home. Actual survival time was calculated by reviewing medical records after a follow-up period of 5 years from enrolment, where the date of enrolment was subtracted from the date of death. Patients who were alive at the final follow-up were censored. We also obtained information from physicians regarding whether they discussed the patient’s preference for place of death with the patients or their families. The physicians were asked two questions: ‘Did you discuss the preferred place of death with your patient?’ and ‘Did you discuss the preferred place of death with your patient’s family?’ The response options included ‘explained that we should discuss your preferred place of death’, ‘explained that there are death places such as general wards, hospices and your own home’, ‘explained specifically about your preference for death place’ and ‘did not explain’. For analysis, we categorised patients whose physicians answered ‘explained that we should discuss your preference for death place’ or ‘explained that there are death places such as general wards, hospices and your own home’ or ‘explained specifically about your preference for death place’ as an ‘explained group’ and those whose physicians answered ‘did not explain’ were classified as a ‘not explained group’.

Preference for place of death

We asked the enrolled patients the following question about their preference for place of death: ‘Where would you like to spend your end of life?’ The response options were ‘own home,’ ‘general ward’ and ‘palliative care unit’

In addition, we compared each patient’s preferred and actual place of death through a follow-up review of their medical records.

Statistical analysis

First, we performed descriptive analyses to summarise baseline characteristics. Second, we compared patients preferred and actual places of death. If a patient’s preference and actual place of death were in the same category, we termed their result concordant. If a patient’s preference and actual place of death were in different categories, the result was deemed to be non-concordant. Third, we performed Wilcoxon rank-sum tests for continuous variables, Cochran– Armitage trend tests for ordinal variables and Fisher’s exact tests for categorical variables to identify the factors associated with concordance. Statistical significance was set at P < 0.05 as statistically significant. All the statistical analyses were performed using JMP version 16 for Windows (SAS Institute Inc., Cary, NC, USA).

Ethics statement

The study met the ethical guidelines for medical and health research involving human subjects outlined by the Ministry of Health, Labour and Welfare in Japan and was conducted in accordance with the ethical standards of the Declaration of Helsinki. The primary attending physicians obtained written informed consent from all patients. The study was approved by the independent ethics committee of Tohoku University School of Medicine (approval no. 2016-1-689).

Patient characteristics

In total, 157 patients with advanced cancer were enrolled between January 2015 and January 2016 [Table 1]. After a follow-up period of 5 years, 10 patients were still alive, 135 had died, and 12 were lost to follow-up. There was missing data regarding the preference for place of death for 33 patients.

Finally, data for 102 patients who died during the follow-up period were available for analysis [Figure 1]. The study group consisted of 64 men (62.7%) and 38 women (37.3%) with a mean age of 63.9 years (standard deviation, 12.3). The primary cancer sites were the gastrointestinal (47.1%), hepatobiliary pancreas (11.8%) and sarcoma (11.8%). The median survival time was 273 days (range: 18–1,731).

Close

Figure 1:

Flow chart outlining participant selection.

Export to PPT

Preference of death place and actual death place

Table 2 shows the patients’ preferred and actual place of death in the study population. Forty-eight patients (47.1%) answered ‘Own home,’ 25 (24.5%) answered ‘General ward’ and 29 (28.4%) answered ‘Palliative care unit’. Regarding the place of death, 17 (16.7%) patients died in their homes, 55 (53.9%) in general wards and 30 (29.4%) in PCUs. As for the comparison between patients’ preference for death place and actual death place, 37.3% of patients died at the place they wanted.

Table 3 shows the factors associated with the concordance between patients’ preferences and the actual place of death. No significant factors were identified in the univariate analysis. Meanwhile, understanding physicians’ goals of chemotherapy (P = 0.10) and discussions regarding patients’ preference for the place of death with patients (P = 0.17) and their families (P = 0.11) had a weak association with the concordance.