Palliative Medicine 2006; 20: 171 /175
The sound of death rattle I: are relatives distressed by hearing this sound? Bee L Wee Sir Michael Sobell House, Oxford Radcliffe Hospitals NHS Trust and University of Oxford, Oxford, PG Coleman Faculty of Medicine, Health and Life Sciences, University of Southampton, Southampton, R Hillier St Joseph’s Hospice, London and SH Holgate Faculty of Medicine, Health and Life Sciences, University of Southampton, Southampton Background: Death rattle is the noisy, rattling breathing that occurs in many dying patients. Health professionals intervene because the sound is said to distress attendant relatives. We found no formal study to confirm or refute relatives’ distress, so we decided to ask the relatives. Method: Face-to-face semi-structured interviews with 27 bereaved relatives to investigate their experience of terminal care and what their response had been to the sound of death rattle if this had occurred. Interview transcripts were subjected to thematic content analysis. Results: We found that almost half of the 12 relatives who had heard the sound of death rattle had been distressed by it. The others were either neutral about the sound or found it a helpful signal of impending death. Conclusion: We confirmed that some relatives do find it distressing to hear the sound of death rattle. However, our expectation that relatives are universally disturbed by this sound was unfounded. There is no justification for a ‘blanket’ approach to therapeutic intervention when death rattle occurs. A better understanding is required of how relatives make sense of the sound of death rattle. Palliative Medicine 2006; 20: 171 /175 Key words: bereaved relatives; death rattle; ethical dilemma; terminal care
Introduction Noisy, rattling breathing which occurs towards the end of life is commonly described as ‘death rattle’. It occurs in 23/50% of dying patients in hospices,1 3 and is a phenomenon well-known to palliative care professionals. Airway secretions are the presumed cause of death rattle, so interventions, such as repositioning, airway suction and anti-muscarinic drugs, are used to reduce the secretions. Even when the patient does not appear to be disturbed, this treatment is carried out because the patient’s attendant relatives are usually distressed.4 There is wide informal acceptance, amongst palliative care practitioners, that relatives are usually distressed by this noise. This view has been strengthened by studies which document the perception of palliative care nurses that relatives are distressed by the sound.5,6 However, there is no published report asking relatives of dying patients about how they were affected by death rattle. The aim of this study was to answer the question: ‘are relatives distressed by hearing the sound of death rattle?’ This is crucial to know because the current rationale for intervention is the alleviation of the relatives’ perceived distress at a time when the patient is no longer able to Address for correspondence: Bee Wee, Sir Michael Sobell House, Churchill Hospital, Oxford, OX3 7LJ, UK. E-mail:
[email protected] # 2006 Edward Arnold (Publishers) Ltd
give or withhold consent or to report adverse effects, such as excessively dry mouth.
Methods We adopted a qualitative interpretive approach, using face-to-face semi-structured interviews with bereaved relatives. This offered a suitable methodology to explore a sensitive subject that has not been extensively investigated. Ethical approval was granted by the local research ethics committee. Over a 2-month period, we wrote to all registered next-of-kin (bereaved relatives) at a time when they were being invited to a bereavement support evening, between 2 and 4 months after the death of their relative. The patients had all been under the care of specialist palliative care services and death had occurred in hospitals, hospices or at home. The study letter invited bereaved relatives to participate in a single face-to-face, audio-taped interview with the investigator (BW) and to discuss their recent experience of terminal care. They were offered the option of meeting in their own home, in a hospice or elsewhere. Each meeting lasted up to 30 minutes and was carried out using an interview schedule. First, relatives were invited to describe their experience of the terminal phase of the patient’s life, in particular the last day or two. 10.1191/0269216306pm1137oa
172
BL Wee et al.
Second, they were explicitly asked if they had observed any change in the patient’s breathing. If they had, the investigator then asked about relatives’ feelings, perceptions and responses to this change. Thematic content analysis was used to analyse the data. Each transcript was read at least twice by BW. Coding categories were initially developed from the study aim. These categories were applied to the first five transcripts and additional coding categories were developed until saturation had been achieved and no new categories were emerging from the data. The coding categories were then applied to the entire dataset to identify broad themes from the data. A non-exclusive approach was used as some of the data could be coded into more than one category. Two random transcripts were also independently coded by each of two health professionals, otherwise unconnected to the study, to check for consistency in coding. The following measures were put in place to support the bereaved relatives in case their participation caused undue distress. At the end of each interview, BW provided explicit ‘wind-down’ time for the relatives to ask her questions once her ‘research hat’ had been taken off. The first four relatives to be interviewed (10% of the proposed sample population) were also invited to submit confidential written feedback to SH about their experiences of participating in this study. Finally, all respondents were offered access to additional support, in confidence, from two palliative care professionals (a consultant and a social worker).
was their gender and relationship to the deceased patient. This information is set out in Table 1. Not all relatives who were interviewed had experienced the sound of death rattle or noisy breathing (Table 2). Nine had observed no change in their relatives’ breathing. Six noted a change only in the pattern of breathing, ie, breath sounds becoming irregular, shallow or deeper. In this paper, we report the data from interviews with the remaining 12 relatives. Ten of these relatives clearly recalled a change in the breathing sound in their dying relative and the other two remembered hearing this in other patients. The descriptors of breathing sound in Table 2 were the only ones offered by these relatives when asked to comment, describe or mimic any observed change in breathing. Ten of the respondents were female and two were male. Ten were spouses of the dying person; the remaining two were daughters. Although a sample population of 40 had been initially proposed, further recruitment was not required as theoretical saturation had been reached. No new data emerged about the way the sound of death rattle affected relatives in the data from the final three interviews. All four relatives who had been invited to submit written feedback on their experience of participating in this study made positive comments. They stated that they had found the interviews helpful. This concurred with verbal feedback from most other interviewees, even when they had clearly found it painful to relive a difficult experience through telling their story.
The effect of the sound on patients’ relatives Broadly, the effect of the sound of death rattle on patients’ relatives can be divided into:
Results The participants Introductory letters about this study were sent to 82 bereaved relatives and 26 agreed to be interviewed. One relative brought his daughter with him, so 27 relatives were interviewed in total. All interviews took place between 4 and 8 months after the patient’s death and, except for three interviews, all were held in the relative’s home. The only information available about those who did not reply to the introductory letter (non-responders)
. Negative feelings . Neutral or positive feelings . Negative feelings associated with the sound of death rattle Five of the 12 relatives interviewed expressed explicit negative feelings about hearing the sound of death rattle. Some found the experience itself distressing, using
Table 1 Known characteristics of responders and non-responders
Gender
Relationship to deceased
Female Male Total Female:male ratio Husband Wife Adult children Other
No. invited
No. of responders (% of nos. invited)
No. of non-responders (% of nos. invited)
52 30 82 1.7:1 17 36 19 10
21 (40) 6 (20) 27 (33) 3.5:1 5 13 4 5
31 (60) 24 (80) 55 (67) 1.3:1 12 23 15 5
The sound of death rattle I 173 Table 2 Observations reported by bereaved relatives interviewed in this study Observations reported by bereaved relatives
No. of relatives
No change noted in relatives’ breathing 9 Change in breathing pattern only, eg, irregular, 6 shallower, deeper breathing Change in breathing sound reported by 10 relatives, eg, ‘noisy’, ‘rattling’, ‘harsh’, ‘gurgling’ No change noted in recent experience of 2 relative’s death, but reported previous experience of hearing death rattle in other dying relatives or patients nearby
descriptors such as ‘awful’, ‘horrible’ and ‘terrible’. The tone of intense distress was particularly striking with two respondents, where the noisy rattling breathing had continued for a long time, in one case for almost 24 hours, and in the other for 2 days: it was soul destroying to see her there. . . she was lying there sort of in a daze but all she was doing was like this horrible noise from her mouth (Respondent 5a) it went on for two days and I knew she must be in pain (Respondent 10) Some of the distress around the terminal event was clearly focused around other things that seemed to be happening to the patient at the time. One was the continuous, sometimes copious, leakage of fluid from the patient’s mouth that occurs with some patients during death rattle: I don’t think he was aware of anything that was happening. . . but I would have been more worried if he was aware ‘cos the fluid scared me more than anything that night (Respondent 7)
Neutral or positive feelings associated with the sound of death rattle Five of the respondents whose dying relative had developed death rattle stated that they had not been ‘bothered’ or distressed by the sound. it wasn’t bad (the sound). . . In fact, I probably would have said it was nothing at all (Respondent 1) he’d been breathing like that for a little while. . . were the things I’d thought would happen. . . so I didn’t speculate (Respondent 3) Some had drawn their own conclusions from the sound of death rattle and appeared to be comfortable with this: I wasn’t distressed by it. I had heard that people who were near the end did breathe a bit oddly (Respondent 21)
we sort of knew, I suppose, in ourselves, that it must be this fluid in the throat and he probably couldn’t swallow anymore (Respondent 12) One interviewee was particularly explicit about the relief that the onset of death rattle brought: ’cos I knew about like the death rattle and things, you know, and when that started, I just thought oh thank goodness for that, it’s near the end you know. . . I felt relieved ‘cos I knew it wasn’t long then you know. . . it’s horrible to hear, it’s a horrible noise to hear and it’s a noise that you never forget but it was also to say that he’s going to be out of this pain within an hour or so, you know (Respondent 7) Two interviewees reported that their most recently deceased relative had not experienced death rattle, but they had previous experience of the sound. One stated that she had heard it in a neighbouring patient. She had not been disturbed by the sound, but had hoped that her husband would not experience this in case it disturbed other patients, even though she was sure he would not have been aware of it. The other relative regretted that her dying husband had not developed death rattle in the end because this would have prompted her to remain by his side through his final night. As things turned out, she was asleep when he died.
Discussion The findings To our knowledge, this qualitative study provides the first direct evidence that some relatives are distressed when they hear the sound of death rattle. In this study, distress was evident in five out of 12 bereaved relatives. This was sometimes, but not always, associated with their concerns about the patient’s suffering. A second, less expected, finding was that the assumption that relatives are usually distressed by this sound was unfounded. At least half of the relatives interviewed were not distressed by the sound; some even found it helpful, as a warning sign of impending death. We used a qualitative, interpretive methodology because there was no existing literature about the impact of death rattle, as reported by relatives themselves, and we anticipated that sensitive questioning of relatives offered the opportunity for in-depth exploration of their perceptions about it. However, a large proportion of those interviewed (15 out of 27) could not offer data that was useful for this investigation because their dying relative had not experienced death rattle. Limitations of this study Invitations to participate in this study were sent out with invitations to attend the bereavement support evening.
174
BL Wee et al.
The mailing list consisted mainly of patients’ registered as next-of-kin, who were not always the people who had spent most time by the patient’s bedside or were the people most affected by the sound. As with any such study, self-selection is an issue. Little is known about the relatives who declined to participate. Previous studies looking at characteristics of bereaved relatives who declined to participate in research have come up with limited information.7,8 In this study, the only information available was the gender and relationship to the deceased patient. Another potential difficulty was that the investigator (BW), although not previously known to any of the relatives, was identified as a palliative care doctor. Despite clear explanations about the ‘research’ role being distinct from the ‘doctor’ role, this was bound to have some impact on the interview and type of information offered by the interviewees. Another issue was the timing of the interview. Relatives were interviewed several months after the patient’s death. The time interval since the event and the bereavement process itself may influence relatives’ recollection of events and their feelings at the time.9,10 The optimum time for interviews needs to be balanced between seeing the relative whilst the memory remains fresh with not intruding at a difficult time and allowing acute grief reactions to settle down. The vividness of descriptions by the relatives interviewed in this study suggested that the experience remained clear in their minds. A further limitation was that coding of the interview transcripts were mainly carried out by only one person (BW), although two random transcripts were independently coded by two other professionals. Finally, these findings apply to relatives of patients from a single centre, suffering only from cancer. Links to previous studies So how does this link in with previous studies? There have been two previous published reports about relatives’ distress at hearing the sound of death rattle. In 1999, Watts et al . reported that of 23 nurses, 52% agreed with the statement that ‘bereaved relatives mentioned distress caused by death rattle’.5 In the same study, 18 respondents (79%) felt that death rattle distressed nurses. This was the first serious attempt to document the impact of the sound of death rattle, but was based on nurses’ views, not the views of relatives themselves. It is possible that the nurses’ views about relatives might have been influenced by their own experiences and by the caring relationships they had developed with the dying patients and their families. The second report was an audit of the use of anti-muscarinic drugs to treat death rattle, in which attending nurses had judged and documented the distress level of relatives present when the sound occurred and at 30-minute intervals thereafter.6 Interestingly, in
this study, distress appeared to diminish in many relatives, even when death rattle persisted until the patient died. It is difficult to expect the nurses, or any other health professional, to have been completely detached in their evaluation of the sound of death rattle, particularly when they are closely involved in the care of dying patients and their relatives. Implications for practice In our study, bereaved carers were directly questioned about their personal experience of hearing the sound of death rattle. We confirmed that some relatives were distressed by this sound, but found that others were not bothered by it, or even regarded it as a useful warning of impending death. The belief that relatives are universally distressed by this sound is not substantiated. This has implications for the management of patients with death rattle: they are unable to give explicit consent to, or report adverse effects from, pharmacological treatment. Any decision to instigate treatment must be taken thoughtfully. Careful exploration of how the sound affects the relatives concerned should occur first. This may be difficult, as members of a single set of relatives may all hold diverse views. In another study, reported in a separate paper in this journal, we explored how relatives tried to make sense of the sound of death rattle when it occurred. A better understanding of relatives’ perceptions about death rattle should help health professionals to frame their explanations and interventions more effectively. Further studies are required to confirm our findings and investigate relatives’ response in different settings and in different diseases.
Acknowledgements BW carried out this study as part of her PhD thesis and wrote the first draft of this paper. PC provided methodological advice for the study, RH and SH supervized the thesis and all three commented on drafts of this paper. We would like to thank the relatives who participated in this study, Celia Cooke for being on standby to provide additional support to them if needed and David Field and Peter Speck for advice. No funding was received and there were no conflicts of interest.
References 1 Bennett MI. Death rattle: an audit of hyoscine (scopolamine) use and review of management. J Pain Symptom Manage 1996; 12: 229 /33. 2 Morita T, Tsunoda J, Inoue S, et al . Risk factors for death rattle in terminally ill cancer patients: a prospective exploratory study. Palliat Med 2000; 14: 19 /23.
The sound of death rattle I 175 3 Wildiers H, Menten J. Death rattle: prevalence, prevention and treatment. J Pain Symptom Manage 2002; 23: 310 /17. 4 Furst CJ, Doyle D. The terminal phase. In Doyle D, Hanks G, Cherny N, Calman K eds. Oxford textbook of palliative medicine, third edition. Oxford University Press, 2004. 5 Watts T, Jenkins K. Palliative care nurses’ feelings about death rattle. J Clin Nurs 1999; 8: 615 /18. 6 Hughes A, Wilcock A, Corcoran R, et al . Audit of three antimuscarinic drugs for managing retained secretions. Palliat Med 2000; 14: 221/22.
7 Stroebe MS. New directions in bereavement research: exploration of gender differences. Palliat Med 1998; 12: 5 /12. 8 Parkes CM, Brown R. Health after bereavement: a controlled study of young Boston widows and widowers. Psychosom Med 1972; 34: 449 /61. 9 Hinton J. How reliable are relatives’ retrospective reports of terminal illness? Patients’ and relatives’ accounts compared. Soc Sci Med 1996; 43: 1229 /36. 10 Higginson I, Priest P, McCarthy M. Are bereaved family members a valid proxy for a patient’s assessment of dying? Soc Sci Med 1994; 38: 553/57.
