Factors associated with the preparedness for bereavement in families of patients with cancer: A secondary analysis of a nationwide bereaved family survey

Insufficient preparedness for bereavement can affect a family's psychological health status after bereavement. However, factors associated with preparedness remain unclear. This study aimed to identify factors associated with preparedness for bereavement in families of patients with cancer.


| BACKGROUND
Psychological care for bereaved families of cancer patients is an important issue.These bereaved families suffer from serious psychological distress in some cases.Bereavement can increase the risk of depression, pathological grief (i.e., complicated grief and prolonged grief disorder), and other mental disorders. 1,2Approximately 10%-55% [2][3][4][5] of bereaved families had depressive symptoms, and 2%-24% 2,3,[6][7][8] were suffering from pathological grief.Previous studies have shown that early support, especially pre-bereavement care, and communication with health care providers, can improve the mental health of bereaved family members. 9,10eparedness for bereavement could be a key factor that modifies family members' mental health status after bereavement, which could be modified through care provided by health care providers in the pre-bereavement phase.2][13] Furthermore, insufficient preparedness for bereavement is associated with sleep disturbances, 14 chronic pain, 15 and risk of suicidal ideation after bereavement. 16For instance, insufficient preparedness increased the risk of depression/ complicated grief by 2-3 fold; conversely, sufficient preparedness is found to buffer suicidal ideation among the bereaved with an odds ratio of 0.59. 1,16,17Thus, sufficient preparedness for bereavement may help to improve a family's psychological health status.However, the preparedness itself, that is, what preparedness for bereavement is and what those determinants are, remains unclear.
Preparing for bereavement involves a family's life experience, sociodemographic factors, coping with uncertainly (e.g., medical, practical, psychosocial, and spiritual), and good communication between those involved in the patients' end-of-life. 18,19Some studies have explored factors that contribute to families' preparedness for bereavement.1][22] However, previous studies have some limitations, such as being was conducted at a single site 21 or being limited to the participants' spouses 20 or patients receiving hospice care. 22Therefore, this study used data from a multicenter bereavement survey to study a larger number of bereaved families, was not limited to spouses, and identified factors associated with preparedness for bereavement in families of patients with cancer.

| METHOD
This study conducted a secondary analysis of data from the Japan Hospice and Palliative Care Evaluation Study 3 (J-HOPE 3) by the Japan Hospice Palliative Care Foundation.4][25] The J-HOPE 3 study was a cross-sectional, anonymous, self-report questionnaire survey conducted in 2014, and the details of this study are described elsewhere. 24We sent letters to 396 member institutions of Hospice Palliative Care Japan, comprising 49 acute hospitals, 296 inpatient palliative care units (PCUs), and 51 home hospice services.Of these, 175 institutions (20 acute hospitals, 133 PCUs, and 22 home hospice services) participated in this study.
Regarding PCUs, the number of institutions registered as PCUs nationwide was 333 in 2014.This represents about 40% of the total institutions in Japan.To identify potential subjects, we asked each institution to identify and list consecutively up to 80 bereaved family members of patients who had died between October 2011 and January 2014.The inclusion criteria were (1) the patient died from cancer; (2) the patient was aged ≥20 years (the age at which one is considered an adult in Japan); and (3) the bereaved family member was aged ≥20 years.The exclusion criteria were (1) the patient received palliative care for less than 3 days; (2) the bereaved family member was unavailable or could not be identified; (3) death occurred in an intensive care unit or was associated with treatment; (4) the potential participants were likely to suffer extra emotional distress after participation (i.e., history of mental diseases), as determined by a primary physician and a nurse; and (5) the potential participants were incapable of completing the self-report questionnaire because of health issues such as cognitive impairment or visual disability.Questionnaires were sent to the bereaved family members identified by each participating institution.
We identified 15,632 potential participants who were bereaved family members of patients with cancer, and 1921 met the exclusion criteria.The questionnaires were sent to 13,584 potential participants, excluding the number of bounces because of unknown addresses.Of the 10,153 responses, 1030 refused to participate.
We enclosed a letter explaining the survey concisely and explaining that returning a completed questionnaire would indicate that the individual had consented to participate in this study.Ethical approval for the study was granted by the institutional review boards of Tohoku University (No. 2013-1-334) and all participating institutions.

| Preparedness for bereavement
To assess the participants' preparedness for bereavement, we asked the question, "How prepared were you for the patient's death?"Their answers were based on a four-point Likert-type scale (1: sufficiently prepared, 2: somewhat prepared, 3: not much prepared, 4: not at all prepared).The participants were categorized into two groups: not prepared group (responses to the question on whether bereaved family members were prepared for the patient's death included 'not prepared at all' or 'not very prepared') and prepared group (responses included 'somewhat prepared' or 'prepared').

| Participant characteristics
We asked the participating institutions to search their medical databases and collect data concerning age, sex, number of hospitalization or home care days, and primary cancer site for each decedent.We asked bereaved family members about the patient's annual household income under medical treatment, period of anticancer treatment, and physical and psychological status during final anticancer treatment (using the same question, although separately, for each item: 'How was your physical/psychological health during caregiving, especially in the last month?',with possible responses ranging from 'good' to 'very bad').The bereaved family members were also asked for details concerning their age; sex; physical and psychological health status during the caregiving period; relationship to the patient; time spent with the patient in the week before death; educational background; perceived social support (feeling loved and cared for, and how they perceived that other people listened to their worries or problems); whether other caregivers were present; religion; belief in the survival of the soul after physical death; medical expenditures during the month preceding the patient's death; communication about the disease with the patient ('How much did you talk to the patient about patient's illness and future life?'), medical expenses for the month before the patient died, and pre-existing mental impairment ("Have you ever seen a psychiatrist or taken antidepressants or other medications?").

| Preference of bereaved family and end-of-life discussions
We asked for details concerning the family's preference of patient treatment and recuperation place, their need for information about what will happen to the patient in the future, and their end-of-life (EOL) discussions with the physician before the patient died, as well as when they started the EOL discussions before the patient died and their feeling about the timing and quality of discussion with the physician.These questions are listed in the Supporting Information S1.

| Analysis
We conducted descriptive analyses of the demographic characteristics and items concerning bereaved family members' preference and EOL discussion with doctors regarding to the preliminary theoretical framework of preparedness for bereavement identified in previous studies. 18,19To identify factors associated with preparedness, univariate analysis was conducted using the likelihood ratio chi-square test, with preparedness as the objective variable and the other survey items (patient and bereaved family background and bereaved family responses) as explanatory variables.
Subsequently, multivariate logistic regression analysis (backward selection) was performed to explore the effects of the independent

| RESULTS
Table 1 shows the characteristics of patients and bereaved family members.Of the patients, 9% died in acute hospitals, 80% in PCUs, and 11% in a hospice at home.Patients were most often over 80 years (35%).Lung cancer (23%) was the most frequent primary site.The bereaved family members were mostly in their 60s (30%), and 82% of the relationships with patients were spouses (46%) or children (36%).In total, 85% of the participants were prepared for bereavement, and 15% were insufficiently prepared for bereavement.
The results of the multivariate logistic regression analysis are shown in Figure 2. (Details of the multivariate analysis results are shown in Table S1).
In general, the factors associated with insufficient preparedness for bereavement were female sex (OR = 1.29, 95% CI: Several risk factors can be modified through the intervention of health care providers.Appropriate timing of EOL discussion between families and physicians was associated with a higher degree of preparedness for bereavement.Previous studies suggested that communication between patients/families and physicians can contribute to sufficient preparedness for bereavement. 18,26,27reover, our results indicate that while communication is important, timing can be more pertinent.Previous studies have reported that families who had EOL discussions more than a month preceding the patient's death were less likely to report depression and complicated grief than families who had EOL discussions less than a month before patient's death. 10Health care providers might need to assess families' and patients' physical and mental conditions to discuss EOL issues at least 1 month before the expected death according to the patient's setting, illness, and the family and patient's mental condition.
Patient and family background factors were also associated with preparedness for bereavement.Consistent with previous findings, sex, relation to the deceased, place of final care, mental health of caregiving family members, and family preferences for patient treatment were significantly associated with preparedness for bereavement.Families of patients whose place of care was acute hospitals were less prepared than families of those in hospice/PCUs, which was similar to the findings of previous studies. 20Previous studies showed that families who had hope for cure were less prepared for bereavement, and that receiving aggressive treatment at EOL contributes to insufficient preparedness for bereavement. 28In acute hospitals, patients are likely to receive aggressive treatment at EOL.In contrast, in PCUs, it is often confirmed, through the medical interview before admission, that the patient will not undergo aggressive treatment or resuscitation.Thus, families of patients at acute hospitals might be less prepared for bereavement than those of patients at hospice/PCUs.Although these background factors cannot be modified through care or intervention, they may be viewed as risk factors that can be used to carefully assess and utilize an individualized approach to promote preparedness.
Additionally, we found several factors associated with preparedness for bereavement that have not been previously reported in our study.First, we found that a lower patient's ADL was associated with sufficient preparedness for bereavement compared with a higher patient's ADL.This might be because perceiving and considering the death of patients is difficult for families of patients with high ADL at the last admission.However, a previous study showed that a higher caregiving burden was associated with a lower degree of preparedness for bereavement. 21Families perceiving a higher caregiving burden might have less time to think about what happens after the patient's death.Second, families who were unsure of their preference of the patient's treatment and who did not need detailed -5 of 8 information about what would happen to the patient in the future.
Families who were unsure of their preference of the patient's treatment may have lacked understanding of the patient's treatment or illness.Previous studies showed that a family's knowledge of medical information was associated with preparedness for bereavement. 18,20Additionally, families who did not need detailed information about the patient's future might have avoided thinking about the patient's death or those who lacked mental capacity to contemplate the patient's future.For these reasons, we believe these families were not prepared for bereavement.Therefore, health care providers need to monitor the level of fatigue caused by caregiving and their mental capacity, which allows them time to prepare themselves to think about the patient's death and after bereavement.Third, families of patients whose treatment duration for cancer was less than 3 months were insufficiently prepared for bereavement.This result suggested that families needed some time to prepare themselves for bereavement.We believe this is because families of patients with shorter duration of treatment have fewer opportunities to communicate with the patients about the disease, treatment and future, which previous studies have shown to be associated with sufficient preparedness for bereavement. 21However, previous studies have reported that while a family could perceive a patient's prognosis correctly when the patient's death approached, they could not prepare for bereavement. 29Hence, it is important to note that the passage of time might not be sufficient for families to prepare for bereavement.

| Study limitations
This study has several limitations.First, it is a cross-sectional study, which means that causal relationships between associated factors and preparedness for bereavement and changes in preparedness for bereavement over time cannot be clarified.Second, the study was F I G U R E 2 Factors associated with preparedness for bereavement.

| Clinical implication
The results showed that factors in which health care providers could intervene, such as communication, and patient and bereaved family background factors were associated with preparedness for bereavement.Our study's results suggest that by carefully assessing these risk factors and getting involved with families of patients with cancer, health care providers can identify family members who may have a low degree of preparedness for bereavement and intervene to promote preparedness for bereavement.

| CONCLUSION
This study examined the associated factors of preparedness for bereavement among bereaved family members of patients with cancer.We found that factors associated with preparedness for bereavement could be divided into factors in which health care providers could intervene and modify, as well as patient and bereaved family background factors.Future longitudinal studies of families of cancer patients with different institutional backgrounds should be conducted to make causal inferences about factors associated with preparedness for bereavement.

1
Recruitment of participants and response rate.Patient and bereaved family characteristics.Prior to formulating the regression model, the variance inflation factor (VIF) was calculated to examine multicollinearity, but there were no variables with VIF greater than 2.A p value of <0.05 was considered statistically significant, and all tests were two-tailed.All statistical analyses were performed with Those associated factors could be divided into (1) factors in which health care providers could intervene and modify (i.e., related to EOL discussions with physicians and patients), and (2) patient and family's background factors (i.e., place of death, sex, relation to deceased).
to cancer patients who received specialized palliative care; these results may not be generalizable to patients who did not receive specialized palliative care, had other illnesses, or had different institutional backgrounds.Further, the responses obtained from bereaved families may include recall bias because of the length of time that had passed since bereavement.Third, other factors could not be fully clarified, because this study is a secondary analysis of a national survey of bereaved families, and the questions asked were mostly about patient/family backgrounds.Future research should conduct an interview survey of families of patients with cancer in Japan to develop a theoretical framework and identify factors associated with preparedness for bereavement.Additionally, this study was answered entirely by bereaved families, including patients themselves.Despite the response bias of the family's assessment, the purpose of this study was to clarify families' preparedness for bereavement, and we believe that this is a reasonable research method for achieving the study's objectives.Finally, this survey was conducted in 2014; however, the age of the data is outweighed by the benefits of the large, multi-site sample and the detailed caregiving and cancer-related data, which are not available in most large existing datasets. limited