INTRODUCTION
The place of death for people dying from cancer has been widely studied and documented in recent decades, especially in North American and European countries (Brazil, Reference Brazil, Bedard and Willison2002; Beccaro et al., Reference Beccaro, Costantini and Rossi2006). Study sizes range from small samples of <100 patients (Montel et al., Reference Montel, Laurence and Copel2009), to large surveys of whole regions (Costantini et al., Reference Costantini, Beccaro and Merlo2005; Davies et al., Reference Davies, Linklater and Jack2006; Lin et al., Reference Lin, Lin and Liu2007). The purpose behind such studies may be to establish a means of addressing existing or future healthcare policies (Gagnon et al., Reference Gagnon, Mayo and Hanley2004; Gomes & Higginson, Reference Gomes and Higginson2006), or to reflect on cultural differences and norms (Decker & Higginson, Reference Decker and Higginson2006). Observations have been made about the trend away from home deaths in some countries, with the impact this has on healthcare budgets and future social and healthcare planning. (Bruera et al., Reference Bruera, Russell and Sweeney2002; Fukui et al., Reference Fukui, Fukui and Kawagoe2004; Cohen et al., Reference Cohen, Bilsen and Addington-Hall2008) This trend has been observed over two decades in more developed societies (Bowling, Reference Bowling1983; Davies et al., 2006; Gomes & Higginson, Reference Gomes and Higginson2008).
ONCOLOGY CARE IN SOUTH AFRICA – A LOCAL CONTEXT
Research into the place of death for those with cancer on the African continent are few and far between, partially because of a lack of accurate statistics (Parkin et al., Reference Parkin, Sitas and Chirenje2008). It is possible that in many parts of Africa most cancer deaths occur at home due to a shortage of in-patient facilities. The growing influence of the palliative care discipline and the development of the hospice movement in Africa should see an improving awareness of where and how those with terminal disease in Africa die (Kikule, Reference Kikule2003; Clemens et al., Reference Clemens, Kumar and Bruera2007), while affording patients and communities better services.
Healthcare in South Africa is broadly divided between state health services and private health providers. State services are available to the entire population, with particular attention to the needs at primary and secondary healthcare levels, and a pay-for-service situation for patients with higher incomes. Oncology treatments therefore occur at tertiary healthcare centers such as academic hospitals, which operate in most of the major cities in South Africa.
Private healthcare is available to paying customers, who may self-fund, or may be reliant on medical aid plans that wholly or partially cover medical expenses dependent on the type of membership selected.
The Constantiaberg Oncology Unit in this study is essentially a treatment clinic, forming part of a group of oncology practices within the country. This unit, which was founded in 1993, is situated in the southern suburbs of Cape Town, in a predominantly middle- to upper-class English-speaking area. The unit uses an independent acute care hospital (Constantiaberg Medi-Clinic) for inpatient admissions, where patients are accommodated, under the care of their referring oncologist. Patients who are cared for in the medical ward may be admitted for long-stay chemotherapy sessions, for protracted investigations, and for symptom control, which may include end-of-life care. The unit treats adult patients most of the time, as there is a specialized pediatric oncology unit at a local children's hospital, where chemotherapy is available. Occasionally children are referred to the adult unit for radiation therapy, but this is a rare occurrence.
Care for patients in the community generally falls to that patient's family, and is seen as being their responsibility by the medical profession. The southern suburbs of Cape Town are well served by private nursing agencies that can assist families with caring for patients at home with basic nursing care. The costs for such services are high, compared with an average household income, and may be covered partially by certain of the medical aid plans, when the need for them is verified by the oncologist involved. Prolonged use of such home nursing care can seriously erode savings or family income, putting additional strain on families trying to care for their dying loved one.
Patients and families who are without financial means, and have no medical aid plan to rely upon for assistance, may be able to make use of home-based carers, provided as a community service by the South African government.
The St. Luke's Hospice supports the community in this area. This hospice has a core of registered nurses who are supported by volunteer visitors, and has a small inpatient unit, comprising ten beds. Unfortunately, referrals to the hospice are not always made sufficiently early, which may be the result of reluctance from the patient and family to allow for hospice intervention, and can ultimately increase the stress on informal carers when symptom control at home becomes difficult. The challenge of appropriately timing of referrals to hospice is noted by several authors, in different countries, across the years (Friedman et al., Reference Friedman, Harwood and Shields2002; McCarthy et al., Reference McCarthy, Burns and Davis2003; Waldrop & Rinfrette, Reference Waldrop and Rinfrette2009).
By looking at where patients of this particular oncology unit died, and how long they spent in acute-care beds prior to their deaths, it was hoped that some indications for future care needs for both patients and family carers could be found.
Although the Constantiaberg Oncology Unit performs many treatments daily, there is an important component of palliative care, with much of the liaison regarding care arrangements falling to the nursing sisters and the oncology social worker employed by the unit. The appropriateness of terminally ill cancer patients being cared for in an acute-care medical ward, is understandably a debatable issue (Willard & Luker, Reference Willard and Luker2006), but may at times be unavoidable in the absence of other resources.
It is important to be cognizant of the emotional burden placed on ward staff members who are confronted with newly admitted patients who are in extremis, together with their families who are understandably anxious and needing of support. Likewise, the stress and pressure of moving terminally ill patients through a treatment clinic to the medical ward carries a toll for the Oncology Unit staff who may be required to manage emergency situations unexpectedly in addition to their regular duties. These factors and the impact they have on staff could form the basis for additional studies.
The information presented here is representative of the social and demographic details of the patient population and will therefore be difficult to compare with other centers in Africa. However, certain parallels with similar studies completed elsewhere are apparent and are therefore referred to.
STUDY SAMPLE AND METHODS
This retrospective study was undertaken over a period of three consecutive calendar years, January 2006–December 2008, and presents an overview of where patients at the Constantiaberg Oncology Unit died in that time, with a total of 426 deaths recorded.
The data analyzed excludes two deaths of people who were registered with the practice for intervention, but were not people with a cancer diagnosis — one had Franconi's Anemia, and the second was thought to have brain metastases requiring radiation, but on investigation had had a cerebrovascular accident as a result of end-stage AIDS. This reduced the total of deaths to 424.
Medical records are manually recorded at this unit, with notes and reports being stored in a folder for each individual patient. The author recorded details of where and when each patient died on MS Excel sheets, as the information became available. Patient details were noted in terms of: age at death, gender, and diagnosis on presentation. In addition, it was noted whether or not the patient was a member of a medical aid plan.
Additionally, each folder was reviewed by the author to ascertain the number of days spent under the oncologist's care prior to death if the patient had died in the medical ward or been referred to the local hospice, and whether the death was known to be as a result of cancer or not.
In terms of the physical place of death, this was categorized according to information available from the patient's folder and notes, and the report of death from family members, the community hospice staff and records, and, in some cases, from the acute hospital's records. Place of death was classified as being either home, under the unit oncologist's care in the hospital's medical ward, in acute hospital care elsewhere (for example, the bone marrow transplant unit, the intensive care ward, or in other hospitals), in frail care (including nursing homes or old-age home ward facilities), hospice care, or in a category for place of death unrecorded.
Missing data was sought in the unit's central computerized registration program (Med-e-Mass™), as well as from the local hospice and the acute care hospital's admissions staff where possible. Deaths that were noted in the classified section of the daily newspapers were recorded and followed up as best possible. As this is an initial, exploratory study into place of death, it was decided not to contact family members directly for information required for this study.
It is possible that some deaths of patients may never have been brought to the attention of the unit's staff, especially when people may have moved away from the area, may not have attended regular follow-up appointments at the unit, or may have died of other causes than their cancer.
Data was stored initially in batches according to year. This would allow for a year-on-year analysis or comparison in the future. The data was then combined, so that statistics could be drawn from the grand total. Statistical analysis of the data collected was conducted using simple descriptive statistics, to obtain averages, means and ranges for the categories of information required. The author was the sole researcher and stored data away from staff in the unit.
RESULTS
Basic Demographics
A total of 424 patients, registered with a cancer diagnosis at the GVI Oncology Unit at Constantiaberg, died in the time frame from January 1, 2006 to December 31, 2008, compared with a total of 2,338 new patients being seen by those oncologists as new patients in that same period.
Of the 424, 44.81% of the deceased were female (190) and 55.19% were male (234). The average age of the deceased patients was 66.09 years, with the youngest patient being a 3-year-old and the eldest, a person of 102 years, giving a range of 99 years (Figure 1). One male patient's age is unknown, giving a total of 423 subjects for this category.
Fig. 1. Age at death, n = 423.
The majority of patients (93%) were beneficiaries of medical aid plans, with 7% being self-funders, or receiving financial backing from other sources.
Most of the patients in this study recorded as their home address, a location within the greater Cape Town municipal area. The city of Cape Town encompasses a wide range of socioeconomic neighborhoods, with the “catchment area” for this unit being predominantly English-speaking, middle- to upper-class residents. Some of the patients were referred to the unit despite living in rural areas outside of the city limits, and three were non-resident in South Africa – one each visiting from Hong Kong and from the United Kingdom respectively, and one being a resident of Uganda who traveled to South Africa for treatment but died upon return to Uganda.
Diagnosis on Registration
On presentation at this oncology unit, each patient's diagnosis is recorded, based upon histological studies, and in accordance with ICD-10 coding. This code is then used for all tests, procedures, and prescriptions and is required by law to be on documentation that will be presented to medical aid plans for reimbursement.
The largest number of fatalities in this study was from lung and bronchus cancer (23.82% all deaths, n = 101). This leading cause of death correlates with the GLOBOCAN 2002 study, which found that lung cancer has the highest cancer incidence and death rates worldwide (Parkin et al., Reference Parkin, Bray and Ferlay2005). The second largest group had breast cancer (15.56%, n = 66), with colorectal cancers as the next highest diagnostic group at 9.90% (n = 42). Thereafter, prostate cancer at 8.96% (n = 38) was the fourth largest diagnostic group, and pancreatic cancer was the fifth largest group (4.48%, n = 19). The remaining diagnoses combined account for 108 of the total deaths.
Melanoma is also a contributor to death in the study population (4% total deaths). This is particularly the case in the white community in South Africa, which has one of the highest melanoma incidences in the world (Mqoqi et al., Reference Mqoqi, Kellett and Sitas2004). One of the oncologists at the practice has a particular interest in melanoma, so it is possible that this specialist interest may attract a higher number of melanoma referrals to this oncology unit than to others.
Stomach cancers and cancers of unknown primary origin also accounted for 4% of deaths each, followed by head and neck cancers, and brain tumors at 3% of deaths recorded each. Figure 2 presents the ten most commonly occurring diagnoses at death, which represented a total of 316 patients.
Fig. 2. Ten leading diagnoses of deceased patients, n = 316.
Mesothelioma was identified apart from the other respiratory cancers, by way of highlighting that this form of cancer is still responsible for a number of deaths every year (1.65%).
Ovarian cancer (2.83%) was grouped separately from other gynecological cancers (1.89%), as this particular cancer reflects upon the socioeconomics of the patient population served by the unit. Likewise, of interest, the number of women dying of carcinoma of the cervix in this study was only 1 – in contrast to the number of deaths this cancer causes elsewhere in Africa (Parkin et al., Reference Parkin, Sitas and Chirenje2008). However, it must be remembered the study population largely drawn from upper- and middle-class urbanites.
As discussed previously, the unit from which this study drew data, treats solid tumors primarily. For this reason, the number of hematological cancers treated at the unit is small (2.83% of patients in this study).
Cause of Death
In order to ascertain whether patients died as a result of cancer of not, death certificates were consulted where possible, and enquiries were made of the relevant hospice staff, where applicable. Perusal of medical notes assisted in establishing the cause of death so that a death resulting from cancer-related symptoms or treatment-related complications would be regarded as being a “cancer death.” Deaths known to be as a result of other medical conditions were then categorized as “unrelated.” There were situations in which a patient died and was believed, on balance of evidence in their folder, to have died of their cancer – this was categorized as being “probable cancer death.” There was a group of patients who died without a precise cause of death being established for our records – this may have occurred when patients were lost to follow-up, had transferred to the care of other oncologists, or had moved out of the geographical area. The cause of such deaths was simply noted as “not known.” There were two known suicides that are recorded separately from other categories. Family members in both cases notified the unit staff of this sudden and unexpected death. Table 1 provides a breakdown of figures for the various causes of death.
Table 1. Cause of death
The majority of deaths recorded, 87.03%, were directly attributable to cancer. By contrast, 6.37% of the deaths recorded in this study are known to be unrelated to the cancer for which the patient had consulted the Oncology Unit.
Place of Death
Where a patient died was recorded as far as could be reliably established, and is reflected in Figure 3. People who died in their own home or that of a close relative or friend were deemed to have died at home (182 or 43%).
Fig. 3. Place of death, n = 424.
Certain of the patients of the Constantiaberg Oncology Unit are admitted to the medical ward at the hospital by their oncologist for symptom control and may die in that ward. Although this is an acute care situation, patients and family are supported as best as is possible by ward staff, and by the oncology social worker. The total number of these patients was the second highest category of deaths, at 145.
There are occasions when Oncology Unit patients die in acute care hospitals under the care of other medical specialists such as surgeons or physicians – this total over the 3-year period was 33.
Patients who died in a frail care or nursing home facility (n = 21) were classified as having died in frail care, as both terms tend to be used interchangeably in South Africa, and imply care in a facility (generally designated for care of the aged), with nursing staff available on a 24-hour basis, and the bulk of the care being delivered to the patient in a bed.
The smallest category in terms of place of death was the local hospice inpatient facility, where 20 of the unit patients died in the period under review. Twenty-three of the patients whose deaths were recorded died in unspecified locations, including several patients who returned to family or friends abroad prior to death or were effectively lost to follow-up.
Hospice Referral
It is the policy of the Constantiaberg Oncology Unit to encourage patients who move into the realm of palliative care, to make use of hospice support and guidance, especially in terms of emotional support and symptom control. Patients are still encouraged to attend follow-up appointments and to make use of the services at the oncology unit, but then benefit from the addition of expert advice and interventions from the community hospice staff.
This approach explains the high number of total referrals to the local hospice in the 3 years under review (52.36%). It stands to reason that where there were sudden, unexpected or non-cancer related deaths, referral to hospice may not have taken place. There are also some patients and families who choose not to request hospice support. There was a group (6.13% of the sample) for which it was not known whether there was a hospice referral or not, possibly as the patients in question may not have been under the care of the unit's oncologists at time of death.
Length of Stay in Hospital Immediately Prior to Death
Initially, the motivation for recording this particular information was to prove that work with even terminally ill cancer patients from this unit is short term, with the focus being on discharging a patient once symptoms have been stabilized, with an adequate care plan in place.
On examining the data, it was immediately apparent that many of the patients were admitted to the medical ward either via the hospital's casualty department or via the oncology unit as emergencies. Admissions were unplanned, requiring immediate medical assessment and management of the patient's symptoms. More than a third of the patients dying under an oncologist's care in the medical ward died within 3 days of admission (Figure 4). Sixty-seven percent of the patients who died in the medical ward were there for 7 days or less, again suggesting that the admissions are in response to a sudden or dramatic change in the patient's condition, coupled with the uncertainty of carers about coping with the situation. In this particular patient population with relatively easy access to hospital care and ambulance services, admission for symptom management is easier than in communities where there is little or no access to hospital care.
Fig. 4. Hospital days under oncologist's care immediately prior to death, n = 145.
LIMITATIONS
This study is, of necessity, time limited, and accuracy of recording information improved as all the staff involved became accustomed to noting details when registering the death of a patient. Three years is a sufficiently long period to give an overall indication of where patients die, together with a little about the circumstances around their death. It is not a long enough period to begin identifying trends or changes in trends to do with place of death. This could be the realm of extended future studies.
Given that the information presented in this article is reliant upon accurate reporting by healthcare workers, administrative staff, and family members of the deceased, it is to be expected that there may be oversights or omissions in the statistics presented. The families of patients who had not been seen at the unit for some time, or who had died of causes other than cancer may have forgotten or chosen not to notify the unit staff of a death.
On occasion, patients may be seen briefly on an emergency basis in the hospital ward, and may die soon after the consultation. In such instances, the unit folder may have been opened only briefly and hurriedly, and may be missing certain data.
Many of the observations reflect the socio-demographics of the patient population, which is not typical of the broader South African society. Culturally and economically, this unit's patient population is more privileged – as witnessed by the high percentage of patients belonging to medical aid plans. The reader is therefore cautioned not to extrapolate this information presented to the broader South African community.
This theme of greater affluence than in many other parts of South Africa carries over to the support families may have caring for their loved ones at home, and the conditions under which such care takes place. It is the broad impression that the majority of patients who are cared for at home up until, or immediately prior to death, are cared for primarily by an informal carer in the shape of a family member or loved one. Some families are able to employ an assistant to help with the caring (for example, a domestic worker), but many rely on the support of extended family members. There are a small number of patients whose medical aid plans allow for the provision of home nursing for a limited period of time, provided the home nurse comes from a registered nursing practice.
DISCUSSION
Diagnoses
The problem of data collation and accurate recording of cancer deaths in Africa has been documented (Setel et al., Reference Setel, Macfarlane and Szreter2007; Parkin et al., Reference Parkin, Sitas and Chirenje2008), especially when basic population statistics are lacking (Setel et al., Reference Setel, Macfarlane and Szreter2007), and when chronic diseases such as cancer do not enjoy priority attention from governments and healthcare sectors (Yach et al., Reference Yach, Hawkes and Linn Gould2004). Total figures for cancer deaths in South Africa are elusive, and mortality figures by diagnosis are even more difficult to locate.
According to data presented by the Medical Research Council of South Africa, based on figures gathered in 2000, the cancers most responsible for adult deaths are tracheal, bronchial, and lung cancer, which nationally account for 1.3% deaths, ranking as the number 14 cause of death across both genders. Esophageal cancer ranks as the number 17 cause of death, as 1.1% all deaths. (South African Medical Research Council, 2009)
Cause of Death
As people with cancer tend to be older, it is to be expected that patients may have co-morbid conditions that may be ultimately responsible for their death.
Where possible, and where issued by oncologists from the practice, the Department of Health's Death Certificate was consulted to establish cause of death. Logistically, it is not viable to source every death certificate of every patient, hence the reference to patient notes in the folder, and that fact that information was gathered from staff taking details of the death, and from hospice workers and from family members where appropriate.
Deaths that may have followed from complications or side effects of cancer treatments are included in the total for cancer-related deaths.
Place of Death
As long ago as the 1980's, academic literature has spoken out in favor of home deaths, and has encouraged the support of the patient in choosing and discussing their preference for place of death (Bowling, Reference Bowling1983; Costantini, Reference Costantini2008).
The majority of deaths (42.92%) among the patients at the Constantiaberg Oncology Unit occurred in a home environment, in contrast to the study of the urbanized, upper-middle-class population of South East England (Davies et al., Reference Davies, Linklater and Jack2006), which found that by 2002 only 23% of the sample group had died at home. A more recent study that aims at predicting trends in place of death in England and Wales, suggests that there is a general tendency in these countries away from home deaths, with a projection that < 1/10 people will die at home by 2030 (Gomes & Higginson, Reference Gomes and Higginson2008)
Although approximately a third (34.2%) of the deaths recorded were under the direct care of an oncologist in the acute medical ward, it is important to recognize that most of the patients would have come to the ward from a home environment.
Willard and Luker (Reference Willard and Luker2006) identify several contributory factors to the admission of patients into acute care during their terminal phase of illness–symptom control; quick and perhaps unexpected deterioration in the patient's condition; a lack of, or insufficient care from community resources; the carer's emotional and/or physical exhaustion.
In South Africa, where community facilities are scarce, even in the better-resourced areas, it stands to reason that many of these factors may be present at the time of admission to hospital.
Long-term care facilities for the frail and aged in South Africa are privately run, with a minority of beds being available to state pensioners, at a subsidized rate. Many nongovernmental organizations, such as churches, offer institutional care, but are reliant on fundraising and community support and their bed availability at short notice is scarce, with such old age homes in the area around the Constantiaberg Oncology Unit having waiting lists literally years long.
Privately run care facilities vary in cost and availability, and many provide an excellent service. However, these facilities usually require a monthly payment for the accommodation, and this amount is frequently far larger than an elderly person's pension, therefore requiring monetary support from family in order to secure the bed.
Medical aid plans in this country seldom provide cover for frail care or nursing care in an old-age home, so such care is self funded by patients and families.
Hospice Referral
This category of data was captured as a potentially interesting aside. As the local hospice battles with resources, there is frequently a delay between when a patient and family submit application forms to the hospice, and when services can commence. For this reason, not every patient in this category will have actively received services from the hospice before their death. As occurs elsewhere (McCarthy et al., Reference McCarthy, Burns and Davis2003), there remain some individuals who prefer not to have contact with hospice, even in the face of terminal illness. Unit staff respects this preference, once the possible referral has been thoroughly discussed.
As the St. Luke's Hospice in Cape Town functions independently of any hospital, serving the whole community and working with various diagnoses, it is necessary for people requesting their services to be formally referred for support.
Length of stay
Data on length of stay suggest that terminally ill patients frequently present to the unit for admission, or are admitted for end-of-life care via the hospital's emergency unit.
Of the patients dying under the oncology unit's care in the medical ward, 38% were in hospital for 3 days or less prior to death, with 15% of people admitted to the ward dying within a day of admission. This suggests that care of the patient prior to admission had been manageable, but that at the very end of life, symptoms may have developed or become alarming and extreme, causing the carers to bring the patient to hospital for attention.
There is also the possibility that when some of the patients and/or their carers realize that death may be imminent, there is a sudden urgency to have that person admitted to hospital care, rather than have them die at home. This aspect of “end stage admissions” has been examined by other authors (Bowling, Reference Bowling1983; Hallenbeck, Reference Hallenbeck2008) and could be further explored in future studies, or by extracting more detailed information from the existing cohort of subjects.
The Role of the Professional Team
In South Africa, palliative care is emerging as a specialty that is providing increasing training and support to healthcare professionals. However, staff members in most oncology treatment units are not palliative care experts. It is expected that staff at the unit familiarize themselves with palliative care practices and principles, and that they are comfortable with terminal care, death, grief, and bereavement. Staff such as the oncologists and the oncology social worker should be able to initiate and continue discussions around end-of-life issues, choices, and decisions, as well as being familiar with some of the cultural practices that the patient population may observe (Kane et al., Reference Kane, Hamlin and Hawkins2004; Nieumeyer & Hosking, Reference Nieumeyer and Hosking2006)
It is clear that excellent communication and respect among multi-disciplinary team members as they care for patients at the end of their lives, will enhance the quality of care delivered, whether to the patient at home, or in a care facility (Greeff & Gouws, Reference Greeff and Gouws2003). Coordination and cooperation between home nursing agencies, hospice staff and oncology unit personnel is essential to allow the patient feel secure and contained. Informal carers of the dying patient also anecdotally report feeling less anxious and better prepared when their professional team acts in a coordinated way and demonstrates continuity of care, even beyond the death of the patient.
IMPLICATIONS
Implications for Policy
Accepting that this article has focused on a group of patients in private medical care in urban South Africa, there are several concerns that could be taken up at a national level, both with policy makers and the funders of private healthcare
1. Of primary concern, although not addressed directly in this article, is the need for country-wide, accessible, and consistent palliative care for all people with advanced cancer. This requires ongoing advocacy and awareness raising, in a country where resources are extremely limited and are frequently prioritized to other life- threatening conditions.
2. The absence of a single, operational, and accurate population-based national cancer registry is a handicap to healthcare planners and providers. Needs, trends, and improvements are not being accurately monitored in the broader population, again reducing satisfactory service provision.
3. There is a clear need for improved support for caregivers in a home environment – recognizing that carers may be unskilled, but are still willing to care, and therefore in need of encouragement (Phipps et al., Reference Phipps, Braitman and True2003), education, and support.
4. Possible discussions with medical aids could be initiated, regarding the needs of informal carers of patients, as is evident in the growing body of literature arising from the United Kingdom, Europe, and North America. Carers may find that they provide support while simultaneously needing support themselves (Koffman et al., Reference Koffman and Mitchell2008). This is known to occur at a time when the caregiving role isolates people from other roles in society, leaving carers feeling burdened and overwhelmed (hence the potential need for mental health support, alongside the obvious need for better practical home care support)
5. There is an opportunity here for advocating to medical aid funders in South Africa for improved benefits for palliative care in terms of home nursing availability, and ambulance transfers from hospital to the patient's home if death is imminent and the family or patient have a strong need for death at home rather than in hospital.
Implications for Practice
A study of this nature will naturally raise issues, some of which could be taken forward into the clinical setting, and brought to the awareness of practitioners
1. Where it is not already in practice, it would be helpful to include a record of the patient's preference for place of death clearly in that patient's medical folder. This preference would become a guideline for the oncology social worker and the oncologist, acknowledging patient's needs while attempting to respect the family's opinions and limitations regarding care (Stajduhar & Davies, Reference Stajduhar and Davies2005; Agar et al., Reference Agar, Currow and Shelby-James2008).
2. The compilation of a unit-specific information sheet for families about practical aspects of home care – medication routines, nursing agencies, role of hospice, equipment, etc. should be considered (Mangan et al., Reference Mangan, Taylor and Yabroff2003). This would afford both families and patients relevant and local details that are recorded in one document and are easy to consult in a home environment.
3. Information about and preparation for end-of-life symptoms that may otherwise be distressing could also be provided in a sensitive manner where appropriate to discussions with family members. This could also assist with providing guidance in terms of timing and appropriateness of hospital or hospice admissions.
4. Family and staff preparation for the death of the patient could be improved by using the Liverpool Care Pathway, or a similar, culturally appropriate protocol, to assist with “diagnosing dying” (Ellershaw & Ward, Reference Ellershaw and Ward2003).
CONCLUSION
The findings in this study underline the fact that palliative care is being practiced at this oncology unit, alongside more aggressive oncology treatments for other patients.
In keeping with many developed communities, the patients of the Constantiaberg Oncology Unit predominantly die at home, with a further 34.2% of deaths occurring in acute care beds under the care of the treating oncologists.
It is probable that assistance from the local hospice, together with the possibility of using home nurses, may enable families to care for people in their homes. It is also speculated that the scarcity of affordable institutional care may be forcing families to manage the terminally ill within their homes.
The statistics presented here should provide a discussion point for staff in similar units, and could serve as a motivator to the hospital management to provide better emotional support and debriefing for medical ward staff who are faced with an average of 47 oncology deaths per year.
As always, with initial and pioneering studies, there is scope for further investigation into the role of the oncology unit in palliative care, possible improved liaison with the local hospice staff, advocacy with the medical aids around improved end-of-life care, and a reevaluation of practices in terms of staff training and preparedness for terminal care support and advice.
ACKNOWLEDGMENTS
Thanks to the staff at GVI Oncology, Constantiaberg for their support as this research project has grown. The authors thanks Dr. Garth Davids for his “editorial eye” and encouragement.
