Table 4  Characteristics of studies

Yamaguchi, Maeda, Hatano, Mori, Shima, Tsuneto, Kizawa, Morita, Yamaguchi, Aoyama&Miyashita, 2017 [46]

Effects of end-of-life discussions on the mental health of bereaved family members and quality of patient death and care

Transversal

Retrospective

Mean time between death and completion of the protocol = 283.4 days

Palliative care services, Japan

N = 9123

Age:

 < 50: 18.8%

51–60: 25.7%

61–70:30.2%

 > 71:25.2%

Woman = 66.7%

Spouse = 46.4%

Primary cancer site

- Lung 23.1%

- Gastrointestinal tract: 45.8%

- Breast: 4.9%

- Others:29.3%

Sociodemographic data:

-patient: physical condition, treatment preferences, need for information, awareness of terminal illness

-caregiver: education, communication with the patient about the disease, treatment preferences, information needs

End-of-life discussions

Quality of Death (Good Death Inventory – Short Version)

Quality of end-of-life care (Care Evaluation Scale—Short Version)

Depression (Patient Health Questionnaire-9)

Complicated grief (Brief Grief Questionnaire)

Discussions about the end of life are associated with lower prevalence of depression and complicated grief in bereaved

The earlier discussions about the end of life took place, the lower the prevalence inherent in depression and complicated grief

Discussions about the end of life are associated with a better perception (of bereaved) of the quality of death and end of life care

12/32

Dumont, Dumont &Mongeau, 2008[47]

End-of-life care and the grievingprocess: Family caregivers who have experienced the loss of a terminal-phase cancer patient

Transversal

Retrospective

Average time between death and completion of the protocol: 137 days

Hospital specializing in palliative care, Quebec, Canada

N = 18

Age [33—75 years old]

WomaN = 61%

Spouse = 66.6%

Cancer

Semi-structured interview guide: building blocks of caregiving that could have facilitated or hindered the grieving process

- Nature of attachment

- Circumstances surrounding death

- History (caregiver)

- Personal data (caregiver)

- Social data

- Stress factors

Risk factors for adjustment to bereavement:

- Caregiver: difficulty expressing feelings, presence of a psychological and emotional burden, unsatisfactory informal and formal support;

- Patient: patient denial of his illness, aggressiveness, presence of symptoms (confusion, major changes in behavior, cachexia, uncontrollable pain), departure to a hospital or palliative care;

- Caregiver-patient: communication problems, ambivalent or dependent relationship, presence of family tensions

Protective elements for adjustment to bereavement:

- Caregiver: optimism, ability to assert oneself, religious and spiritual beliefs, past experiences, satisfactory formal and informal support, preparation for loss;

- Patient: acceptance of the disease, control of pain and suffering, respect for the sanctity of death and the dignity of the patient;

- Caregiver-patient: presence of significant relationships, presence of the family at the time of death

9/32

Kelly, Edwards, Synott, Neil, Baillie & Battistutta, 1999 [43]

Predictors of bereavement outcome for family carers of cancer patients

Longitudinal

Prospective

Time 1: before the patient's death

Time 2: 122 days after the patient's death

Hospice Homecare Service at Mt Olivet Community Services, Brisbane, Australia

N = 178

Age = 57 years old

WomaN = 76%

Spouse = 60%

Cancer diagnosis:

- Lung: 28%

- GI: 32%

- Breast: 7%

-- Melanoma: 5%

- Genito-urinary: 6%

- Prostate: 5%

- Gynaecologic: 3%

- Haematologic: 2%

- Brain:CNS: 2%

- Other: 4%

- Unknown primary: 6%

Caregiver

Social support (Social Health and Social Contact Instrument)

Number of adverse life experiences (Life Events Inventory)

Ways of Coping Measure

Perception by caregivers of the frequency of specific patient symptoms + distress they cause to caregivers

Quality of the relationship between the caregiver and the patient (Intimate Bond Measure)

Anxiety and Depression Symptoms (DSSI Anxiety and Depression Scale)

Psychic distress (The General Health Questionnaire)

Guilt Scale

Post-traumatic stress disorder (Impact of Events Scale)

Symptoms of bereavement (Bereavement Phenomenology Measure)

Patient

Functional state (Karnofsky Palliative Scale)

Cognitive state (Mini Mental State)

Quality of life (QL Index)

During step 1, as a caregiver, being a woman, having anxious depressive symptoms and a lower level of practical assistance, having a perception of a higher level of control in the caregiver-patient relationship, to identify a greater number of undesirable life events and that the patient be more seriously are so many elements which predict, at step 2, the anxiety and depressive symptoms in the bereaved

In step 1, the psychic distress experienced by the caregiver predicts, in step 2, the bereavement symptoms in the bereaved

In step 1, the caregiver's use of avoidance as a coping mechanism, a lower quality in the caregiver-patient relationship, and separation from parents in childhood predict, at step 2, post traumatic stress disorder

During step 1, as a caregiver, the fact of experiencing a feeling of guilt, of having the perception of less attention to the patient-caregiver relationship, of experiencing a lower level of practical assistance, separation from the parents during childhood and a quality of life and a general state of health of the patient which appear to be altered, are so many elements which predict, at step 2, the feeling of guilt

10/32

Hirooka, Fukahori, Taku, Togari& Ogawa, 2017 [45]

Quality of death, rumination, and posttraumatic growth among bereaved family

members of cancer patients in home palliative care

Transversal

Retrospective

Mean time between death and protocol completioN = 912.5 days

Palliative home care services, Japan

N = 805

Age = 63 years old

WomaN = 71.8%

Spouse = 52%

Primary cancer sites:

- Lung: 23.3%

- Stomach: 16.7%

- Pancreas: 9.5%

- Colon: 9.4%

- Others: 32%

Sociodemographic data:

- patient: age, gender, 1st location of cancer

- caregiver: age, gender, relationship with the patient, religious status

Quality of Death (Good Death Inventory-Short Version)

Post Traumatic Growth Inventory

Being woman as a caregiver predicts post-traumatic growth in the bereaved

The caregiver's religious beliefs predict post-traumatic growth in the bereaved

Better quality of death is associated with higher post-traumatic growth

9/32

Thomas, Hudson, Trauer, Remedios & Clarke, 2014 [48]

Risk factors for developing prolonged grief during bereavement in family carers of cancer patients in palliative care: A longitudinal study

Longitudinal

Prospective

Time 1: before the patient's death

Time 2: 183 days after the patient's death

Time 3: 396 days after the patient's death

Palliative care services, Melbourne, Australia

N = 

- 301 at Time 1

- 167 at Time 2

- 143 at Time 3

WomaN = 76.7%

Spouse = 50.9%

Cancer

Self-esteem, family support, financial impact, impact on the organization, impact on health, (Caregiver Reactions Assessment)

Partner dependency (Bereavement Dependency Scale)

Social support (Multidimensional Scale of Perceived Support)

Family functioning (Family Environment Scale)

Optimism (Life Orientation Test)

Preparedness for Caregiving Scale

Perceived competence for the caregiver role (Caregiver Competence Scale)

Complicated mourning before loss (PG-12)

Anxiety and Depression (Hospital Anxiety and Depression Scale)

Complicated mourning (PG-13)

Demoralization Scale

Post-traumatic stress disorder (PTSD checklist)

In step 1, as a caregiver, living with the patient, being the patient's spouse, not having completed high school, and the patient's young age are so many predictors, at times 2 and 3, symptoms of complicated grief

During step 1, as a caregiver, the fact of experiencing anxiety-depressive symptoms, symptoms of complicated grief, and having a low level of optimism, are so many predictors, at step 2 and 3, symptoms of complicated grief. These predictions are improved with the contribution of socio-demographic and psychosocial factors of caregivers

In step 1, the caregiver's gender and age, length of care, previous caregiving experiences, and the patient's level of functioning did not predict symptoms of complicated grief at step 2 and 3

Symptoms of complicated grief and a low optima level at time 1 predicted complicated grief at time 3

11/32

Koop & Strang, 2003 [49]

The bereavement experience following home-based family caregiving for persons with advanced cancer

Transversal

Retrospective

Mean time between death and protocol completioN = 161 days

Capital Health Authority Home Palliative Care Services, Edmonton, Alberta, Canada

N = 15

Age = 58.5 years

WomaN = 73.33%

Spouse = 60%

Cancer

Semi-structured interview guide

Difficult decisions leading to disagreements and harsh words between the caregiver and the patient, the image of the patient's physical decline, and the sense of horror associated with the patient's treatment room contributed to grief adjustment difficulties for the bereaved

The fact of having succeeded in accompanying the patient creates a feeling of capability and pride in the bereaved. Likewise, the latter may be made to feel gratitude and relief

7/32

Yamashita, Arao, Takao, Masutani, Morita, Shima, Kizawa, Tsuneto, Aoyama & Miyashita, 2017 [50]

Unfinished business in families of terminally ill with cancer patients

Transversal

Retrospective

Palliative care services, Japan

N = 642

Age = 61.48 years

WomaN = 64.6%

Spouse = 42.3%

Cancer

Unfinished business

Depression (Patient Health Questionnaire 9)

Complicated grief (Brief Grief Questionnaire)

Families with unfinished business are more likely to develop depression and complicated bereavement (than families with no unfinished business)

10/32

Chiu, Huang, Yin, Huang, Chien &Chuang, 2009 [44]

Determinants of complicated grief in caregivers who cared for terminal cancer patients

Transversal

Retrospective

Mean time between death and protocol completioN = 271 days

Palliative care services, Taiwan

N = 668

Age = 42.9 years

WomaN = 60.6%

Spouse = 29.0%

Diagnosis of patients:

- Lung: 22.8%:

- Buccal cancer: 17.1%

- HCC: 9.3%

- Colon cancer: 8.0%

- Esophageal cancer: 7.3%

- Pancreatic cancer: 6.2%

- Breast cancer: 5.2%

- Hypopharyngeal cancer: 4.9%

- Gastric cancer: 4.1%

- Rectal cancer: 1.8%

- Others: 13.2%

Sociodemographic data:

- patient: age, sex, cancer, terminal symptoms, duration of treatment, places of care

- caregiver: sex, relationship with the patient, religious practices, education, income, medical and psychological history

Satisfaction with social support

Inventory of Complicated Grief

As a caregiver, being woman, the marital relationship, the parent–child relationship, the absence of religious belief, unavailable family support, a history of mood disorders predispose the bereaved to complicated grief

The following factors, namely a length of long-term care, a patient's stay in palliative care and caregivers with a medical history, would be protective factors for complicated bereavement

12/32

Allen, Haley, Small, Schonwetter, & McMillan, 2013 [51]

Bereavement among hospice caregivers of cancer patients one year following loss: predictors of grief, complicated grief, and symptoms of depression

Longitudinal

Prospective

Time 1: before the patient's death

Time 2: 365 days after the patient's death

Palliative Home Care Services, Florida

N = 188

Age = 66.41 years

WomaN = 74%

Spouse = 66%

Cancer

Caregiver

Sociodemographic data: age, sex, ethnicity, education, employment status, relationship with the patient

Depression (Center for Epidemiological Studies-Depression)

Current feelings of grief (Texas Revised Inventory of Grief)

Inventory of Complicated Grief

Satisfaction with social support

Patient

Sociodemographic data: age, sex, ethnicity, education

Functional capacities (Palliative Performance Scale)

Number of cancer symptoms (Memorial Symptom Assessment Scale)

Time 1 factors, including caregiver's depressive symptoms, fewer patient deficiencies and fewer years of caregiver training predict, at time 2, depression in the bereaved

The factors of time 1, namely the depressive symptoms of the caregiver, the young age of the patient, fewer years of training for the caregiver predict at time 2, "worse" bereavement in the bereaved. The deficiencies of the patient and the resources of the caregiver are not significant. Lower social satisfaction correlates with worse grief, but not in the regression analysis

Time 1 factors, including the caregiver's depressive symptoms, the patient's young age and fewer years of caregiver training predict at time 2, complicated bereavement in the bereaved. The deficiencies of the patient and the resources of the caregiver are not significant. Lower social satisfaction correlates with complicated grief, but not in the regression analysis

There is no effect of gender, ethnicity, employment status, type of relationship, and number of cancer symptoms on depression, bereavement, and complicated bereavement

10/32

Ferrario, Cardillo, Vicario, Balzarini&Zotti, 2004 [52]

Advanced cancer at home: caregiving and bereavement

Longitudinal

Prospective

Time 1: before the patient's death

Time 2: 91 days after the patient's death

Time 3: 183 days after the patient's death

Time 4: 365 days after the patient's death

Palliative care team, Northern Italy

N = 

- 111 at Time 1

- 96 at Time 2

- 93 at Time 3

- 93 at Time 4

Age = 56.2 years

WomaN = 66%

Spouse = 42%

Tumours:

- Solid tumours: 97%

- Lymphoma: 3%

Performance Status (Eastern Cooperative Oncology Group Performance Status Scale)

Family functioning (Family Strain Questionnaire)

Personality dimension (Eysenck Personality Questionnaire)

Anxiety (STAI XI and X2)

Depression (Depression Questionnaire)

Subjective satisfaction with life (Satisfaction with Life Scale)

Perceptions of physical, emotional and social problems encountered in the last 3 (Caregiver Mourning Questionnaire)

Time factors 1, including being an elderly caregiver (over 61) and having a significant emotional burden, predict a poor adjustment to bereavement in the bereaved

Patient's age, social participation, knowledge about the disease, quality of family relationships, and thoughts about death at time 1 have no effect on grief adjustment in the patient. bereaved

11/32

Ringdal, Jordhøy,

Ringdal &Kaasa, 2001 [42]

Factors affecting grief reactions in close family members to individuals who have died of cancer

Longitudinal

Retrospective

Time 1: 30 days after the patient's death

Time 2: 91 days after the patient's death

Time 3: 181 days after the patient's death

Time 4: 396 days after the patient's death

Palliative Medicine Unit, Trondheim, Norway

N = 

- 183 at Time 1

- 92 at Time 4

Age = 56.8 years

WomaN = 68%

Spouse = 63%

Cancer origin:

- Gastrointestinal: 46.4%

- Lung: 14.2%

- Breast/female genitals:10.9%

- Prostate: 8.7%

- Others: 19.7%

Sociodemographic data: age, relationship with the patient, child living at home, education, employment status, number of days between diagnosis and death, places of death

Current feelings of grief (Texas Revised Inventory of Grief)

Being a woman caregiver, being an older caregiver (> 60 years old) and having lost a young family member are all things that are likely to trigger grieving reactions more important

There is no effect of the relationship with the patient, nor of children living at home, nor of employment status, nor of the number of days between diagnosis and death, nor of place of death on adjustment to bereavement

10/32

Nanni, Biancosino& Grassi, 2014 [53]

Pre-loss symptoms related to risk of complicated grief in caregivers of terminally ill cancer patients

Longitudinal

Prospective

Time 1: before the patient's death

Time 2: 183 days after the patient's death

Hospice, Ferrara, North-East Italy

N = 60

Age = 60 years old

WomaN = 75%

Spouse = 65%

Primary cancer site:

- lung: 32.9%;

- gastrointestinal: 27.6%,

- other: 39.5%

Sociodemographic data: relationship with the patient, caregiver living alone at home

Symptoms of complicated pre-loss grief (Inventory of Complicated Grief—Pre Loss)

Inventory of Complicated Grief –Structured Clinical Interview

Pre-loss complicated grief criteria including traumatic distress, separation distress, and emotional symptoms appear to be linked to a post-loss complicated grief diagnosis

There is no effect of the duration of pre-loss distress on complicated bereavement in the bereaved

9/32

Götze, Brähler, Gansera, Schnabel, Gottschalk-Fleischer & Köhler, 2016 [54]

Anxiety, depressionand quality of life in family caregivers of palliative cancer patients during home care and after the

patient’sdeath

Longitudinal

Prospective

Time 1: before the patient's death

Time 2: 61 days after the patient's death

Ambulatory palliative care teams and palliative care service at the University Medical Center in Leipzig, Germany

N = 

- 106 at Time 1

- 72 at Time 2

Age = 65 years old

WomaN = 69.4%

Spouse = 79%

Most common diagnosis of sample: cancer of the:

- prostate: 17.0%,

- lung:14.2%,

- pancreas:13.2%,

- colon:11.3%

Social support (Oslo Social Support Scale)

Quality of Life (Short Form-8)

Anxiety and Depression (Hospital Anxiety and Depression Scale)

In time 1, when caregivers experienced high levels of mental distress and insufficient social support, then they were more likely in time 2 to have high levels of anxiety and depression

At step 1, when the caregivers were the spouses of patients and had their own weak physical functions, they were particularly depressed at step 2

There is no effect of gender, age, marital status, education, religion, period of care and location of death on anxiety and depression

10/32

Mori, Yoshida, Shiozaki, Morita, Baba, Aoyama, Kizawa, Tsuneto, Shima& Miyashita, 2018 [55]

‘‘What I Did for My Loved One Is More Important

than Whether We Talked About Death’’: A Nationwide Survey of Bereaved Family Members

Transversal

Retrospective

Mean time between death and protocol completioN = 236 days

Palliative care units, Japan

N = 678

Age = 61.4 years

WomaN = 60.9%

Spouse = 38.5%

Primary cancer site:

- Lung: 22.4%

- Esophagus, stomach, colon, rectum: 27.3%

- Liver, gall bladder, pancreas:18.0%

- Breast:4.6%

- Head and neck:4.4%

- Kidney, prostate, bladder:8.3%

- Uterus, ovary:3.8%

- Blood, lymph node:1.5%

- Other:1.6%

Sociodemographic data of the patient and caregiver

Actions in preparation for death

Discussion around death

Depression (Patient Health Questionnaire-9)

Complicated grief (Brief Grief Questionnaire)

Caregivers who had acted and talked about the death before the patient died were significantly less likely to experience depression after the patient died (than those who had neither acted nor spoken)

Caregivers who acted before the patient's death were significantly less likely to experience complicated grief after the patient's death, whether or not they spoke (compared to those who neither acted nor spoke)

12/32

Bradley, Prigerson, Carlson, Cherlin, Johnson-Hurzeler, Kasl, 2004 [56]

Depression Among Surviving Caregivers:

Does Length of Hospice Enrollment Matter?

Longitudinal

Prospective

Time 1: before the patient's death

Time 2 = 183 days after patient death

Hospice, Connecticut, United States

N = 174

Age: < 65 years = 71.8%

WomaN = 72.4%

Spouse = 30.5%

Cancer

Sociodemographic data: age, sex, education, annual income, marital status, religious status, relationship with the patient

Patient prognosis

Number of days in the hospice before death

Duration of care provision before enrollment in the hospice

Use of services

Caregiver Burden (Zarit Burden Interview)

Social support

Caregiver health

Depression (Structural Clinical Interview for DSM)

Caregivers of patients enrolled in a hospice for 3 days or less were significantly more likely to have major depressive disorder at time 2 (than caregivers of those with a longer hospice stay)

12/32

Gilbar, 1998 [57]

Length of Cancer Patients' Stay at a Hospice: Does it Affect

Psychological Adjustment to the Loss of the Spouse ?

Transversal

Retrospective

Average time between death and completion of the protocol: 336 days for 50.4% of the sample

Hospice, Northern Israel

N = 134

Age = 64 years old

WomaN = 61.2%

Spouse = 100%

Cancer

Sociodemographic data: sex, time in bed, length of stay in hospice, year of diagnosis

Psychological distress (Brief Symptom Inventory)

Psychosocial Adjustment to Physical Illness Scale

Texas Revised Inventory of Grief

A stay in short-term palliative care (1–7 days) had a beneficial effect on the bereavement of the surviving spouse (compared to a long stay, i.e. 8 days or more)

Grieving male spouses, a shorter time in bed and a short stay in a hospice would predict less psychological distress

9/32

Kristjanson, Sloan, Dudgeon &

Adaskin, 1996 [58]

Family Members' Perceptions of

Palliative Cancer Care: Predictors

of Family Functioning and Family

Members' Health

Longitudinal

Retrospective

Time 1: before the patient's death

Time 2: 91 days after the patient's death

Palliative care services, Manitoba, Canada

N = 

- 80 at Time 1

- 64 at Time 2

Age: > 65 years old = 40%

WomaN = 78.7%

Spouse = 55%

Cancer

Caregiver

Family Inventory of Needs

Care Expectations of Family Members (F-Care Expectations Scale)

Family Perception of Care (F-Care Perceptions Scale)

Family satisfaction with care (FAMCARE Scale)

Perception of family functioning (Self-Report Family Inventory)

Family Health (Symptoms of Stress Inventory)

Patient

Symptom Distress Scale (Quality of Life Scale)

The experience of caring for the family during the palliative phase affects the health of family members and the family's ability to function early in the grieving period

7/32

Liu & Lai, 2006 [59]

Find a way out: bereavement supportin Taiwan hospice

Longitudinal

Prospective

Time 1: before the patient's death

Time 2: 14 days after the patient's death

Mackay Hospice Palliative Care Center at Mackay Memorial Hospital, Taipei, Taiwan

N = 108

Age = 43 years old

WomaN = 65%

Spouse < 50.8%

Cancer

Chinese Anticipatory Grief Scale

Mourning (Chinese Perinatal Grief Scale)

Anticipated grief was significantly but moderately correlated with post-death grief

Neither age, relationship with the patient, nor gender are factors associated with bereavement

10/32