The aim of this study was to investigate family members’ experiences of support received during the last three months of life, at the time of death and after the death of a person with advanced illness.
This study employed a retrospective cross-sectional survey design using The VOICES (SF) (Views of Informal Carers – Evaluation of Services) (Short Form) questionnaire, and multiple methods for data analyses.
Setting and Sample
The study sample consisted of adult bereaved family members of persons with advanced illness, who died in four hospitals, located in two Swedish healthcare regions, between August 2016 and April 2017. In Sweden, care at the end of life can be provided at home, in hospitals, in nursing homes and in specialized palliative care units, e.g., hospices. The deceased persons had all died in hospital but had received care in several care places and settings, in one of the healthcare regions. The hospitals were used as recruitment settings since 42% of the Swedish population die in hospital .
The inclusion criteria were that the deceased persons had identifiable bereaved family members; both deceased persons and bereaved family members needed to be 18 years old or older; and the deceased person had to have died from underlying causes of death (ICD-10 codes) in accordance with the Murtagh et al.  model: HIV/Aids; Malignant Neoplasms (Cancer); Alzheimer’s disease, dementia and senility; Neurodegenerative diseases; Heart diseases including cerebrovascular diseases; Respiratory diseases; Liver diseases and Renal diseases.
The questionnaires were sent to the bereaved family members four to twelve months after the death of the deceased person, which is in accordance with experiences from previous studies using VOICES (SF), in which it was concluded —after testing different time frames— that 4–12 months after death was a good time period, striking a balance between intrusion in the grieving process and avoiding the likely gradually reduction in the ability to recall over time [13-17].
The VOICES (SF) questionnaire
The VOICES (SF) is a questionnaire that retrospectively evaluates the quality of care received in several different care places, during an ill person’s last three months of life, based on bereaved family members’ reports. The full version of VOICES (SF) has been translated and validated into other languages by other research groups [18-20]. The Swedish version  of the questionnaire is divided into domains such as Care at home; Care homes; Hospital care; and Specialized palliative care units/hospice care. It contains 75 items, initially about characteristics (e.g., age, sex, educational attainment, relationship to the deceased person), followed by, e.g., items about symptom relief, communication, involvement in decision-making, being treated with respect and dignity by care staff and satisfaction with care. The questionnaire includes structured items about help and support received during the illness, at the death of the ill person and after the death. Additionally there are open-ended questions — some in connection with the items about support and help, and three open-ended questions at the end of the questionnaire. There have been previous studies using VOICES in different patient groups and in various healthcare settings, both in cross-sectional studies and at a population level, mainly in the United Kingdom, where the questionnaire was developed [13, 17].
Variables used to describe the characteristics of the bereaved family members were age, sex and educational level. The same variables were used for the deceased persons, with additional ones also included — namely time of illness before death, diagnosis, care place/care provider, number of care places and the relationship between the deceased person and their family member.
The variables chosen for this study were questionnaire items about the following: support during the illness; having been contacted in time to be present at the death; support from the staff at the time of death; being treated with respect by staff at the time of death; and support after the death. Five open-ended questions were also included. Of these, two were comments linked to the items about support during the illness and being treated with respect by staff at the time of death. The other three were open-ended questions in the final part of the questionnaire; one asking if the respondent would like to add anything else about the care and support received, and two questions asking whether anything was good or bad about the care (Table 1).
Recruitment and data collection
Of all the patients who died in the recruitment hospitals during the study period, about 80% (n=1277) were eligible for inclusion. Based on the inclusion criteria, hospital administrators identified the deceased persons, and their bereaved family members were identified via the hospital’s patient records by one healthcare professional at each hospital (assigned to assist Author 1). Postal addresses of the bereaved family members were retrieved from publicly available databases. Written information about the study was sent, including contact information for one of the researchers (Author 1), information stating that the study was performed in cooperation with the hospital in which their family member had died and the VOICES (SF) questionnaire along with a pre-paid return envelope. The written information assured confidentiality and the right to withdraw from the study at any time without explanation. Consent was considered to have been obtained upon return of a questionnaire; no other written informed consent for participation in the study was obtained. For ethical reasons — i.e., being sensitive towards the family members who may be considered vulnerable due to bereavement and may not wish to participate— no reminders were sent.
Characteristics of the deceased persons and their bereaved family members
The deceased persons were between 40 and 90 years or older (64% were 80 or older) and 50,3% were men. Of the deceased persons, 72,5% had lower secondary education, 11,1% higher secondary education and 15,5% higher education. The most common underlying cause of death was heart diseases, including cerebrovascular diseases (56,3%), followed by cancers (15,8%) and respiratory diseases (15,1%) (Table 2). The participating family members were between 18 and 90 years old or older and 70,7% were women. Of the family members, 29,5% had lower secondary education, 30,5% higher secondary education and 39,4% higher education. About half (51,4%) were children of the deceased person and 34,3 % were spouses or partners. Of the deceased persons, 79,2% had been cared for at home at some point during the last three months of life, with 52% receiving care from general practitioners (GPs); 17,9% from specialised palliative home care and 36,7% received care from district- and county nurses. Furthermore, 90,7% had received hospital care, 27,4% nursing home care and 15,7% care in a specialised palliative care unit. The number of places of care during the last three months of life ranged from 1 to 4, with two places being most common (63,4%), followed by three places (20,4%), one place (12,9%) and, least commonly, four places of care (3,3%).
The characteristics of the bereaved family members who chose to answer one or more of the open-ended questions differed slightly from the total sample regarding sex (77,7% women), educational attainment (52,2% higher education and 19,3% lower education) and the deceased person’s educational attainment (20% higher education and 67,4 % lower education).
The response rate was 37,9%, resulting in a total of 485 bereaved family members participating in the study. The non-responding family members’ individual characteristics (e.g., age, sex, educational attainment) were not available. The deceased persons’ profiles linked to the non-responders did not differ from the sample. The responses to the open-ended questions are presented in Figure 1.
Both statistical and qualitative methods were used for data analyses (Table 1).
Descriptive statistical analyses were used to explore the variables derived from the questionnaire items and characteristics of the deceased persons and their family members. For statistical computations, Statistical Package for the Social Sciences (SPSS) version 21.0 (IBM Corp., Armonk, NY, USA) was used.
Initially, the quantitative data was analysed, after which the open-ended responses were analysed qualitatively to deepen the understanding of the family members’ responses to the items about support received. On average, the responses were 1 or 2 sentences long, ranging from a couple of words to full pages. Moreover, there was a variety in scope of content; some were quite exhaustive stories, while others were shorter substantive responses. After excluding responses with non-relevant content, 529 responses (20 full pages of single-spaced text) were analysed. The analysis was guided by ideas inherent in interpretive description, which is a viable approach for generating knowledge applicable to clinical practice. . The analysis involved both descriptive and interpretive processing of data. In line with interpretive description, the authors’ theoretical and clinical knowledge —together with knowledge gained from previous studies— was used for the present study e.g.in the study design and analysis of data. Initially, with the study aim in mind, all the open-ended responses were read to obtain an overall picture. The responses were read again in the context of the family members’ varied responses to the items about support, to discover patterns and deepen the understanding of the quantitative responses. Thereafter, the text was broadly coded, identifying meaningful content and variations of contextual descriptions of support, which were organised into patterns of experiences. The patterns were then read again and interpreted through a process of asking questions such as “what is the underlying meaning of this?” The descriptions and interpretations of the text were continuously discussed and revised within the research group for clarification and further development of the analysis.