PropertyValue
is nif:broaderContext of
nif:broaderContext
is schema:hasPart of
schema:isPartOf
nif:isString
  • The study and control groups were identified via the SRPC. During 2015, data from 66% of all deaths in Sweden were gathered by the SRPC [24]. The SRPC collects data with a web-based end-of-life questionnaire (ELQ), which is completed by healthcare staff after the death of a patient. Items focus on the last week of life and concern several aspects of the quality of end-of-life care. The ELQ was developed based on the principles of a good death proposed by the British Geriatrics Society [25]. Stored data are matched weekly with the central population register for validation purposes [19]. The validity of the ELQ has been examined in a specialised palliative care setting and was shown to vary between different items, but with overall improvements in an updated version of the ELQ [26,27]. The SRPC database also contains individual-level information about the cause of death according to the 10th revision of the International Classification of Diseases (ICD-10). These data are retrieved from the Cause of Death Register at the Swedish National Board of Health and Welfare. Inclusion criteria for the study group were as follows: data were reported to the SRPC, patient deceased during March 2012 to February 2015, 18 years or older, no forensic port-mortem examination was planned, death was reported as expected and dementia was the underlying cause of death according to the Cause of Death Register. Exclusion criterion was report of cancer as contributing cause of death in the Cause of Death Register data (ICD-10 C00–D48, including subcategories). Dementia was classified according to the ICD-10 (F00, F01, F02, F03 and G30, including subgroups). Subgroups of patients with Alzheimer’s disease (ICD-10 F00 and G30, including subgroups) and other and unspecified causes of dementia (ICD-10 F01, F02 and F03, including subgroups) were identified. The control group had the same inclusion criteria as the study group but with cancer as underlying cause of death according to the Cause of Death Register. Exclusion criterion for the control group was report of dementia as contributing cause of death in the Cause of Death Register data. Cancer patients were chosen as controls because several studies that have compared end-of-life care quality between patients with cancer and other diseases using SRPC data have shown that patients with cancer generally receive better end-of-life care compared to patients with other diseases [21–23], thus enabling a comparison with “best standard” end-of-life care. The control group of cancer patients was also chosen because Swedish cancer deaths have better coverage in the SRPC compared to cardiovascular diseases and lung diseases [24]. The following quality indicators based on items collected by the ELQ in the SRPC database were analysed: documented decision by the responsible physician to shift treatment/care to end-of-life care; pressure ulcers at death; assessment of oral health during the last week of life; someone present at the moment of death; information to next of kin about transition to end-of-life care; offer of a follow-up talk to next of kin after death of the patient; fluids via enteral tube or intravenously during the last 24 hours of life; assessment of pain and other symptoms during the last week of life; prescriptions of PRN drugs against pain, anxiety and nausea; and consultation of a specialised palliative care team during the last week of life. A robust validity for items about symptom prevalence and symptom alleviation has not been established in the SRPC [27], and these items were therefore not deemed suitable as quality indicators in this study. We did not include the item about place of death in line with the patient’s last stated wishes because of the uncertain nature of data collection for the dementia group for this item. We estimated that we would be able to identify at least 800 patients per subgroup and year. To be able to detect a 5 percentage point difference in the outcomes between subgroups, with significance level below 0.05 and power of 0.8, the study would require 1600 patients per subgroup. The study period was set from March 2012 to February 2015, a period when no changes were made to the ELQ. Chi-square test (gender and place of death) and t-test (age) were used to compare groups and subgroups. The outcomes were dichotomised after exclusion of the “Don’t know” answers. When analysing whether a specialised palliative care team was consulted, only patients cared for outside specialised palliative care were included. Patients with no next of kin were excluded from the analyses regarding next of kin. The outcomes were then analysed with logistic regression, with cause of death (dementia vs cancer) as independent variable. A model adjusted for age and gender was developed. A limit of 5 percentage point for clinically significant differences in outcomes between subgroups was chosen. The working procedure and study design were approved by the Regional Ethical Review Board in Umeå, Sweden (registration number 2017/11-31).
rdf:type