“He has come home to die,” said Cara.
 “What is it?”
 “His heart; some long word at the heart. He is dying of a long word.”
 Evelyn Waugh, Brideshead Revisited

More than one third of people dying within a hospital do so within the first few hours of being admitted to hospital.¹ Many of these deaths occur within the emergency department (ED) and are both sudden and unexpected. The very nature of these unanticipated deaths increases the possibility of an abnormal grief reaction in the bereaved, which may be prolonged.^2–6 The landmark Harvard Bereavement Study^7,8 involved undertaking a series of structured interviews with a group of widows and widowers, aged <45 years, over a 2–4 year period. A control group of matched married people was also interviewed. The interviews assessed different aspects of social, physical and psychological adjustment, and also looked at a variety of factors that might be used to identify those at risk of an abnormal grief reaction (box 1). The study suggested that the bereaved must develop ‘an acceptable account of how the loss occurred’ if they are to return to full social and emotional function after bereavement.⁵

The provision of care by EDs to these bereaved relatives and close friends has developed enormously over the past 25 years,^1,12–14 and the evidence strongly suggests that it beneficially influences the subsequent grief reaction.¹⁵ While much has been written about the process of dealing with bereaved relatives in the ED, the evidence base for following up these relatives is limited to a number of descriptions of service models.^1,14,16–21 The effectiveness of these services was not formally evaluated by psychometric testing; however, less rigorous evidence derived from telephone surveys and questionnaires indicated that the bereaved appreciated such services, which allowed them to discuss outstanding issues and concerns. The bereaved also frequently express the idea that follow-up shows that somebody was caring for their loved one at the moment of death and subsequently care about them.^14,16,20

In spite of this paucity of evidence, it has become widely accepted that a follow-up bereavement service is valuable, although the scope and quality of such services is still unfortunately variable.^16,22 A document, When a patient dies: advice on developing bereavement services in the NHS,²³ was published in October 2005 in direct response to surveys of NHS Trusts in 2001²² and 2005²⁴ and a subsequent public consultation document in the first half of 2005. It also importantly built upon recommendations from the Frenchay and Alder Hey inquiries²⁵ and subsequent advice from the Chief Medical Officer.²⁶ Similarly, the British Association of Accident and Emergency Medicine and the Royal College of Nursing in 1995²⁷ and Victim Support in 1994²⁸ recommended that EDs provide such a service. Unfortunately, all of these documents failed to provide specific, detailed guidance about the form that such a service should take.

The ED of South Manchester University Hospital offers a bereavement service to relatives of those patients who have died within the ED. Four to six weeks after the patient dies, the named next of kin is contacted by letter; they are offered the chance of either meeting with or telephoning the consultant responsible for the bereavement service to ask questions and discuss issues regarding the death of their relative. The relatives meet with one of the consultants who participate in the service. The consultant will have gathered the required information from the patient’s ED and hospital records and the postmortem examination reports, before the meeting. The departmental bereavement pack contains an information leaflet about the service, which is given to the next of kin while they are in the department. In addition, a form with the name of the deceased and the name and contact details of the next of kin is sent to the secretary who co-ordinates the service.

It is important to distinguish this follow-up service from a counselling service, a specific intervention dealing with grief reactions, provided separate from the ED by trained counsellors. In the service provided at South Manchester, it is stressed to the relatives that it is not designed to be a counselling session; however, follow-up may allow the identification of bereaved relatives who have developed or at risk of developing an abnormal grief reaction. In addition, the contact details of the hospital bereavement counsellor are included.

As part of the continuing review and development of the service, a retrospective audit was carried out and is presented as one particular model of care.

METHODS

A secure computer database was developed to record the following data: (a) dates that the invitation letters were sent; (b) whether replies were received; (c) whether meetings were accepted; (d) which members of the bereaved families attended the meeting; and (e) a brief outline of the matters discussed.

These details were recorded immediately after each meeting. A retrospective audit of a 4-year period (2001–2005) of the database was carried out. Information was derived from the database itself, from the case notes (which included documentation of the ED attendance when the patient died) and the free text notes made by the consultant at the follow-up appointment. Cause of death was determined either from the hospital notes or from post-mortem examination reports. One of the departmental secretaries was responsible for co-ordinating the service and sending out the letters, booking the appointments and updating the database.

Tables 1–3 show the results of the review of the bereavement survey in the South Manchester University Hospital Trust ED.

RESULTS

In total, 1286 deaths (which included both adults and children) were recorded on the bereavement service database between 1 July 1999 and 12 February 2005. Letters were sent to 1179 named next of kin; this 8% discrepancy was due to the next of kin contact details being unavailable to the department.

Nearly half (41%) of relatives (table 1) who were contacted replied either by post (returning the tear-off slip at the bottom of the letter), phone, or e-mail. Of these, 14% accepted the offer of a meeting; 27% declined but all of these expressed thanks for the offer of the meeting. One person out of the 478 who replied to the letter found it to be obtrusive and inappropriate. General feedback about the letter was positive and included comments that thanked the department for the letter and complimented the good quality of care provided at the time of death by the staff involved.

In 81% of meetings, either information was given about voluntary organisations such as Cruse Bereavement Care or contact was made on behalf of the relatives for follow-up with their general practitioner (GP) or occasionally a hospital consultant. If counselling was thought to be appropriate, this was arranged by the bereaved person’s GP. In total, 21 (13%) of relatives were advised to have medical follow-up, mostly for screening for familial cardiac risk factors.

DISCUSSION

Although many articles that discuss the issues surrounding the treatment of relatives at the time of the death of their relative exist, few articles have been published about the role of the ED in supporting relatives in the weeks after the death. Of those articles that directly examine or discuss the follow-up of bereaved relatives, all had slightly different models of care and examined different aspects of their model either formally or informally. Definite conclusions are therefore difficult to make, and issues such as the timing of contacting the relatives, the service to be provided, which members of staff should provide such a service and the training needs of such staff are all unresolved. Whether such a service is beneficial is also unclear, but the studies do suggest that the provision of information and the chance to ask questions is an important component of such a service and is needed and appreciated.^9,29

In setting up this particular service, it was initially felt that it would be best to delay contacting the relatives, to allow them time to overcome some of the initial shock of the bereavement and to settle some of the important practical problems, such as arranging the funeral. It was decided to wait for 4–6 weeks before sending the letter; although there was no firm evidence in choosing this figure, it simply seemed to be reasonable at the time and subsequent experience has not yet refuted this decision.

The group of relatives who accessed this service wanted answers to specific questions: the audit of our service provides useful information about the nature of questions asked (table 3), allowing us to anticipate requests from the relatives at the bereavement meeting and therefore collate information before the meeting. The information may be useful for those planning to develop such a service.

It is interesting to note that questions about the cause of death, such as the results of post-mortem examinations, and the meaning of the medical terminology on the death certificate, were asked in 75% of meetings (a similar figure to that found by Yates et al¹). Relatives can find out the results of the necroscopy from the Coroner’s Office, either directly or via the funeral directors, and it is estimated that 40–50% of relatives will do so (personal communication, Manchester Coroner’s Office).

The second most commonly asked set of questions concerned the degree of distress experienced by the patients (19%), in particular questions about the rapidity of onset of the final illness, whether the patients were conscious at the end, and what treatments were administered before and during their stay in the department.

In 7% of meetings, family members were concerned that they or their children might need screening for certain conditions, most commonly after their relatives had died of cardiovascular disease at a relatively young age. No note has been made of what arrangements were made in every case, although in our experience, if the deceased had previously been under the care of a consultant, an appointment was arranged for the family to meet the consultant. Otherwise, the relatives were advised to see their GPs for appropriate referral (especially if the family was geographically dispersed); however, as a result of the audit, it has been decided to change this informal practice and to contact the GP by letter, and to add such information to the database.

It was interesting to note that this service was used by relatives of patients who had died at a great age (table 4), and it would be informative to compare the proportion of patients in each age group whose relatives attend meetings and if any particular issues are pertinent to different age groups.

Our audit, whilst based only on the comments from bereaved families, seems to reinforce the principle of providing follow-up to allow relatives to ask questions and seek information that they may not have been given or had forgotten. The effectiveness of our approach is difficult to measure, but the number of relatives contacting us does provide some measure of the usefulness of the service. Importantly, there has been only one case of negative feedback from any relative, who thought that the letter was intrusive. In fact, a large number of relatives responding to the initial letter stated that they were grateful for the letter but were coping well and had no concerns or questions; it would be interesting to ascertain the reasons why the other 58% of relatives who failed to reply did not make contact.

The service provided is not labour intensive, and does not require specialised staff; although not formally recorded or audited, from the personal observations of the consultants involved, meetings last between 40–60 minutes, with the shortest lasting approximately 15 minutes and the longest approximately 2 hours. As a result of the audit, the way information is entered onto the database has been changed; initially data was added as free text but it is now added in tick boxes according to the categories shown in the tables above (with an area for free text if necessary). This speeds up the process of updating the database and also simplifies the process of accessing the information; in particular, it will make it easier for future audits to identify any changes in patterns of questions asked by relatives, which will ensure that the service remains flexible to meet their needs.

The results are necessarily limited by this being a retrospective audit with there being no formal evaluation of the bereavement service and any feedback that is available is based only on informal comments by relatives. While we believe this audit to include all deaths that have occurred within the ED, it is possible that a small number of deaths has escaped being recorded on the database.

It would be interesting to ascertain why relatives do not respond to the service; presumably there are a number of reasons, including those who are progressing uneventfully through their bereavements and do not need external help and those who are undergoing abnormal grief reactions. We feel that it is outside the remit of this service to follow up these relatives because it may become intrusive and actually make the situation worse for them.

In conclusion, contacting relatives after a few weeks allows them to ask the questions that had been forgotten previously, and to clarify the facts that happened on the day of their bereavement. This may be a useful way of reducing unnecessary guilt, providing a degree of comfort, and allowing an understanding of the situation that gives power back to the relatives to help them in their healing process. Bereaved relatives must not be left with unanswered questions, otherwise there may be a risk that the grieving process may be prolonged.

Finally, a small but significant number of patients die shortly after having passed through the ED (e.g. those who do not survive surgery for an abdominal aortic aneurysm). The South Manchester University Hospital Trust ED does not encompass the bereaved of such patients, but there is no reason why this model should not applicable to other departments.