To explore thoughts about death and perceived health status in elderly patients with heart failure during a 6month period after a deterioration needing hospitalisation.
To explore thoughts about death and perceived health status in elderly patients with heart failure during a 6month period after a deterioration needing hospitalisation.
A descriptive, mixed methods design was used. Health was measured with EuroQol-5D, thoughts about death with multiple choice and open questions. A total of 145 patients with New York Heart Association class II–IV heart failure, mean age 70 years, 70% males, were included.
During deterioration, 16% of the patients were afraid of dying and 4% had this fear very often. Fear of death did not change significantly during the 6 months after deterioration. Both during deterioration (r=0.26, P<0.01) and 6months later (r=0.40, P<0.001), fear of death and anxiety/depression were correlated. Content analysis of the open question produced 5 categories: (1) Death as a natural part of life; (2) Death as a relief from symptoms and disability (3) Death as fearful (4) Arrangements for time after death (5) A wish for an extended life.
Elderly patients with heart failure had a lot of thoughts about death. Higher levels of anxiety/depression were correlated to fear of death. Many expressed death as a natural relief from suffering, others were afraid of pain, loss of independence and dignity.
Heart failure is a public health problem affecting approximately 10–14 million individuals in Europe and 5–6 million in North America [1  b3]. It has a poor prognosis, 30–40% of patients with moderate to severe heart failure die within one year after the diagnosis and 60–70% die within 5 years. The cause of death is equal between worsening pump failure and sudden death caused by ventricular arrhythmias [4,5] leaving the patients in uncertainty not only about when, but also about how they will die. Heart failure is more malignant than most common forms of cancer , but surveys have shown that it is not clear to the public nor to all health care professionals how serious the condition is [7,8]. Further, little is known about how patients with heart failure perceive and reason about their prognosis and health status. A recent position paper on advanced heart failure highlights “thoughts about death and dying when having advanced heart failure” as a priority for clinical research . There are a few qualitative studies [10  b12] exploring thoughts about death, but to our knowledge there are no quantitative studies estimating how common thoughts and fear of death are in the heart failure population and how perceived health and other demographic and clinical factors are related to thoughts about death. The aim of the present study was therefore to explore thoughts about death and perceived health status in elderly patients with heart failure over a 6 month period after deterioration needing hospitalisation.
A descriptive, mixed method design was used, integrating both quantitative and qualitative approaches. The study was performed at four outpatient heart failure clinics in southern and central Sweden, situated at 2 university hospitals and 2 county hospitals. Heart failure clinics in Sweden are mostly nurse-led and include similar components (patient education, psychosocial support, assessment of symptoms/signs, uptitration of heart failure treatment, increased access to care through telephone consultations) as described previously [13,14]. The four clinics in the study were representative in this respect. The study had a 6 month follow up. Inclusion criteria were: previous diagnosis of heart failure according to the European Society of Cardiology guidelines , in New York Heart Association Classification (NYHA class) II-IV, hospitalised due to deterioration of heart failure symptoms with a follow up at an outpatient heart failure clinic after discharge. Exclusion criteria were: dementia or severe psychiatric illness, inability to speak, read or write Swedish or refusal to participate.
The study was performed in accordance with the principles outlined in the Declaration of Helsinki. Ethical approval was obtained from the Regional Ethics Committees for Human Research at the University of Linköping, Sweden. The patients received both written and verbal information about the study before they agreed to participate.
EuroQol-5D, a generic single index defining health in 5 dimensions (mobility, self-care, usual activities, pain/discomfort and anxiety/depression), was used to measure perceived health [15,16]. For each dimension the patients rated their health on one of three levels, no problem (coded as 1), some or moderate problems (coded as 2) and unable or extreme problems (coded as 3). Theoretically, 243 different health states can be generated by this classification. The ratings can be converted to health-state utility scores. These scores range from −0.594 to 1, with scores ≤0 being regarded as worse than death and 1 representing full health [15,16].
In addition, one global health question, asking whether the patient's general health status was improved, unchanged or worse over the past 12 months, was also asked.
There were two questions on thoughts about death. The first question was a closed question from the Cardiac Health Profile  addressing if the patient had a fear of dying. The answers could be rated as:  very seldom having a fear of dying,  having a fear of dying from time to time or  very often having a fear of dying. The second question was an open question, where the patients were invited to freely write down their thoughts about death.
All patients completed the EuroQol-5D, the global health question, the question about their fear of dying and the open-ended question regarding thoughts about death, at baseline. The baseline visit took place during a follow up visit to an outpatient heart failure clinic, 2–3 weeks after discharge from hospital following an admission due to heart failure deterioration. The same questionnaires, except for the open question, were sent to each patient by post 6 months after the baseline visit. Non-responders were sent two reminders over a period of 5 weeks. The rationale for repeating the questions after 6 months was to evaluate whether fear of dying changed when the majority of the patients were in a more stable phase.
The comparison between and within groups was performed using the chi-square test for discrete variables, student's t-test was used for normally distributed continuous variables, and if the distributions were not approximately normal or ordinal the Mann-Whitney U test and Wilcoxon signed rank test were used. Correlations were analysed with Spearman rank correlation coefficient. A multiple binary logistic regression (Enter method) was used to determine the relationship between socio-demographic factors, dimensions of perceived health and fear of death. The model was constructed by using the logistic regression procedure available in SPSS 15.0. Factors included in the regression analysis were those which had significantly correlated to fear of death in univariate analysis (sex and anxiety), age was forced into the model despite a non-significant correlation. The first category:  very seldom having a fear of dying, we dichotomised as no/little fear of dying. The other two categories:  having a fear of dying from time to time and  very often having a fear of dying, were merged and dichotomised as having a fear of dying. A P-value of less than 0.05 was considered significant.
The open question was analysed by manifest content analysis . All patients gave an answer to the open-ended question. The statements varied in length from 99 to 1653 characters. All statements were included in the analysis since they were focused in relation to the aim. Meaning units were extracted from the written answers, condensed, abstracted and labelled with a code. The codes were compared based on differences and similarities and sorted into categories that were internally homogenous and externally heterogeneous. The whole context was reread several times in order to preserve the original meaning of the codes during the categorisation. When the categorisation was finalised the distribution of meaning units in each category were was calculated and presented as percentages.
Credibility in the data collection was achieved by including patients with different experiences of heart failure in terms of disease severity, time since diagnosis, age, sex, civil status and education. This also influenced transferability of the findings. Credibility in the analysis was achieved by discussions between the two researchers during the process of sorting codes and naming categories.
A total of 145 consecutive patients with systolic heart failure, verified by echocardiography, gave informed consent and were enrolled in the study. Eight patients died during follow up, the remaining 137 patients completed the 6 month follow up after two reminders. Patient characteristics are shown in Table 1. Demographic data and baseline anxiety scores for the eight patients who died did not significantly differ from the baseline scores of the patients who completed the 6 month follow up. Thirty percent of the patients included in the study were women, the only gender difference in socio-demographic variables was that significantly fewer women were living with someone (P<0.001).
|Living with someone||88 (64%)|
|Compulsory school||96 (66%)|
|Aetiology of HF|
|Ischaemic heart disease||95 (66%)|
|Dilated cardiomyopathy||18 (12%)|
|Heart valve disease||9 (6%)|
|Duration of HF in months|
|Severity of heart failure by echocardiography|
|Mild heart failure||20 (14%)|
|Moderate to severe heart failure||125 (86%)|
The health-state utility score (median, inter quartile range) calculated from the five dimensions in the EuroQol-5D at baseline was 0.73 (0.62–0.85). Utility scores were significantly improved after 6 months 0.80 (0.66–0.94) (P=0.01). The ability to perform daily activities (e.g. work, household chores and social activities) increased after 6 months (P=0.01). Pain decreased after 6 months (P=0.001). Very few patients had problems with performing self-care, such as washing and dressing themselves, both at baseline and 6 months follow up. The psychological dimension measuring anxiety and depression did not change significantly over time. Table 2 summarises the results on perceived health. General health status improved significantly during the 6 month follow up (P=0.004), with fewer patients rating their health as worse or unchanged and more patients perceiving their health to be improved compared to the previous year.
|EuroQol-5D parameter||Baseline||6 months|
|No problems with walking around||80 (58%)||72 (53%)|
|Some problems with walking around||57 (42%)||65 (47%)|
|Confined to bed||0||0|
|No problems with self-care||135 (98%)||133 (97%)|
|Some problems with washing or dressing||1 (1%)||2 (1%)|
|Unable to wash or dress||1 (1%)||2 (1%)|
|No problems performing usual activities||72 (53%)||91 (66%)|
|Some problems performing usual activities||52 (38%)||34 (25%)|
|Unable to perform usual activities||13 (9%)||12 (9%)|
|Neither pain nor discomfort||52 (38%)||82 (60%)|
|Moderate pain or discomfort||78 (57%)||49 (36%)|
|Severe pain or discomfort||7 (5%)||6 (4%)|
|Neither anxious nor depressed||75 (55%)||88 (64%)|
|Moderate anxious nor depressed||57 (42%)||45 (33%)|
|Severe anxious nor depressed||5 (4%)||4 (3%)|
|No problems in any of the 5 dimensions||24 (18%)||34 (25%)|
|EQ-utility score (median, inter quartile range)||0.73 (0.62–0.85)||0.80 (0.66–0.94)|
At baseline 16% of the elderly heart failure patients were afraid of dying compared with 13% of patients at the 6 month follow up. Few patients were severely afraid of dying and fear of death did not change significantly over time, see Table 3. There was a strong correlation between fear of death at baseline and after 6 months (r=0.50, P=0.001). Further, fear of death and anxiety/depression were significantly correlated at baseline (r=0.26, P=0.003) and 6 months (r=0.40, P=0.001). Women were more afraid of death at 6 months follow up (P=0.04). No other correlation to fear of death, except sex and anxiety/depression, was found among the dimensions of the EuroQol-5D, utility scores and demographic variables. In the logistic regression, only anxiety and depression were significantly correlated to fear of death at both baseline (odds ratio 2.49, P<0.03) and 6 months (odds ratio 6.46, P<0.001).
|Very seldom||115 (84%)||119 (87%)|
|From time to time||20 (15%)||12 (9%)|
|Very often||2 (1%)||6 (4%)|
The analysis of the answers from the open-ended question on thoughts about death showed that almost two thirds of the patients (64%) expressed in various ways that they often thought about death. The content analysis produced 5 categories:  Death as a natural part of life (38% of the meaning units);  Death as a relief from symptoms and disability (13% of the meaning units);  Death as fearful (22% of the meaning units) ; Arrangements for time after death (7% of the meaning units);  A wish for an extended life (20% of the meaning units).
This category described death as a natural part of life. Death was described as a certainty, the only sure thing that would happen; the uncertainty was not knowing how and when death would occur. Death was something that everyone had to accept and learn how to deal with. The patients realised that with increasing age death became more of a reality. Many of their friends and family members had already died and they had dealt with life and death issues during periods of grief several times. Some spoke openly about death and their prognosis. Nowadays, when I have become so old I often think about death. I am not afraid of death, but I think about it and I joke about it. You are always aware of death, it is natural.
This category described death as relief from pain, shortness of breath and lost functions. Some expressed a wish to die quickly and painlessly. I hope death will be peaceful and quick, that I will die during my sleep. Just go to bed at night in your own home and never wake up again. That is much better than ending your days in a nursing home….
Individuals that believed in God or had had a near death experience perceived death as something positive. I have seen the light in the tunnel and that made me realise that death is something good.
This category described the fear of death. Some patients had always been afraid of dying. From a very young age they had feared the death of their own and other loved ones and had had to deal a lot with this fear. To some, death was so emotionally difficult to deal with that they always tried to repress these thoughts. I just cannot handle to think about death. To others the fear of death had come in connection to the heart failure diagnosis and in periods of deterioration. After I got this heart disease I am often afraid at night, I am afraid to go to sleep, I fear that I will not wake up again.
Many patients were afraid of deterioration and a painful and difficult time before death. That they would lose their independence and that they would not receive good care. I am afraid of becoming helpless, and lose my dignity. Limited health care resources for the elderly and in palliative care were described as a threatening cloud by some patients. I do not think that I am afraid of dying, but I am afraid of what might happen before death like severe pain and shortness of breath or that I will not get good care at the end of my life.
This category described that for many patients there were practical aspects connected to the thoughts about death. Many patients had made a testimony (Will) and planned their funeral. Some were occupied with trying to throw away old things and organise their possessions and financial issues. They did not want to leave all these practical things for their partners and children to deal with after their death. One woman described how she always made the house ready in case she died during the night. Death can come quickly when you are in my age; I do not leave anything to the next day.
Some patients had no expectations for the future, since they were so old and sick and they just waited to die. Others were very motivated to try to do everything to live longer. They expressed how, despite physical limitations and symptoms, they enjoyed life and hoped that they would have an extended life. The patients described how they tried to improve their health through lifestyle changes, adherence to treatment and other self-care activities. The reasons given for the wish for a longer life was that they wanted to support their family, wanted to see their grandchildren grow up or finalise tasks they were dedicated to. I want to live longer. I like life and I have just started to write my memoirs, so right now is not a good time to die.
This study underlines that it is important to start addressing existential and end of life care issues before heart failure becomes too advanced. At advanced stages of disease, the patient may be unable to participate in decision making or to receive counselling and support, to help them accept deterioration and death. The patients in our study, who had a poor prognosis due to moderate to severe heart failure with deterioration needing hospitalisation, described that they had a lot of thoughts about death and dying making this an area that clinicians should not neglect. The 5 categories from the content analysis, gave a more in-depth view of the different aspects of what heart failure patients think about their prognosis and how this relates to their thoughts about death. Only a few patients in our study stated that they were severely afraid of dying. A study by Willems et al.  also reported, that despite a poor prognosis, very few patients with heart failure feared death. This might be explained by a lack of knowledge and awareness about the poor prognosis of heart failure among both health care professionals and patients and their families [7,8]. Another explanation found in our qualitative analysis was that the patients with heart failure had accepted that they were in the process of dying and had adapted to this condition. Adaptation to illness provides a probable explanation for the occurrence of response shift, meaning that death is not fearful anymore. It is also known that patients adapt their rating of quality of life over time as the illness progresses [19,20]. Levenson et al.  found that when death came closer, disability and symptoms increased. However, there was no significant decrease in quality of life as death approached. During the last month of life, many patients described a good to excellent quality of life.
Only anxiety and depression correlated to fear of death in the present study, no correlation to general health or specific dimensions of health were found. Willems et al.  stated that very few patients thought about death, except during deterioration. Our data revealed that the patients often thought about death, but they were no more afraid of dying during the time of deterioration at baseline. This suggests that fear of death might be more influenced by personality and mental health than severity of symptoms and perceived health.
Our patients described that they were more afraid of the process of dying, with suffering and pain, than the fact that their life would end. Some wished for a pain free and quick death during sleep, while others were reluctant to go to sleep since they might not wake up again. Limited health care resources were mentioned as a threat to receiving palliation and dignity at the end of life. Fitzsimons et al.  also found that patients dying from chronic disease had many concerns and unmet needs. Aiken et al.  showed that provision of intensive home-based palliative care to patients with severe chronic heart failure can effectively decrease symptom distress and improve self management of illness, vitality, physical functioning and perceived health in the last years of life . Brännström et al.  also highlighted the important role of advanced home care facilities. Rabow et al.  showed that consultations by a palliative team improved patient outcomes for dyspnoea, anxiety, and spiritual well-being, but failed to improve pain or depression, highlighting an area in need of improvement.
Most of the patients in the present study suffered from moderate to severe heart failure and the majority had problems in at least one of the five dimensions of health measured by EuroQol. Almost half of the patients had problems with mobility. Previous studies have shown that many patients with heart failure have limitations within the physical dimension [24,25]. In the present study, the patients described that they were very limited in their daily life and that death could be a relief from always being tired, having pain and being short of breath.
The mixed method used in this study integrated quantitative and qualitative approaches to collect and analyse data. By combining the methods their complementary strengths can be utilized and the analytic power improved. The mixed method made it possible for us to use both numbers and narratives in order to provide a more complete view of the phenomenon under study [26,27].
Although clinically relevant, this study was restricted by its small sample size and the descriptive design. Since all patients were enrolled in a disease management programme at a heart failure clinic the results may reflect a rather stable and well controlled sample with access to specialised care; therefore it may not be appropriate to generalise these findings to heart failure patients with mild to moderate heart failure in primary care or those in need of palliative or advanced heart failure care. The study population had a lower mean age and consisted of more men than in the general heart failure population. This was not unexpected, since the patient population receiving follow up at heart failure clinics tend to be younger and more often male. Some of the most elderly patients are not referred to heart failure clinics due to problems with communication, cognition or mobility. The aetiology was representative for a hospitalised heart failure population.
Further, when measuring fear of death we found the combination of one open and one closed question relevant. However, only these two questions were used and for future research a more extensive instrument, covering the different aspects of death and dying, would be relevant.
We are aware that asking questions about thoughts on death could be upsetting for severely ill heart failure patients. However, the study received ethical approval, and patients had the opportunity to withdraw from the study and to call a heart failure nurse if they wanted counselling or support.
Care and support for heart failure patients at the end of life is an area that needs to be addressed in research and clinical practice. Future research should further explore how patients with heart failure and their partners reason about death and dying, through interviews and larger quantitative studies comparing heart failure patients with age and sex matched controls.
Elderly patients with heart failure have a lot of thoughts about death and this needs to be recognised in clinical practice. Higher levels of depression and anxiety were correlated to a fear of death, but no other relation was found among the dimensions of the EuroQol-5D, utility scores and demographic variables. Many patients regarded death as natural and as a relief from suffering, others were afraid of pain and loss of independence and dignity.
Anna Strömberg has a research position grant from the Swedish Research Council and additional funding from the Vardal Foundation. She is a member of the Vardal Institute. Tiny Jaarsma has been a guest researcher funded by Östergötland County council and has also a research position funded by the Netherlands Heart Foundation and the department of Cardiology of the University Medical Centre, Groningen.