br Total patients enrolled br Patients
Total patients enrolled
Patients still alive
Patients who died
Preferred place of death (POD)
Fig. 1. Flowchart indicating analysis of preferred and actual place of death.
only significant factor associated with death in the preferred setting.12
Our study has a number of strengths: its prospective longitudinal design, relatively large sample size, and complete follow-up with remarkably little missing data. However, our data come from a single center serving urban public sector patients; our findings therefore may not be generalizable to other popula-tions in South Africa. We did not collect data on care-givers’ circumstances that might have affected actual POD or on caregivers’ preferences for place of care or POD for the patient. We also did not consider the number of palliative care consultations each patient received.
Sociodemographic and cultural factors associated with POD may not be amenable to intervention by health care providers. However, there are some factors that may be modifiable with timely palliative care pro-vision. Patients who were taking morphine for pain were more likely to die in their preferred place, espe-cially if at home. Although consumption of morphine-equivalent strong opioid analgesics per cancer death is considered by the WHO to be an indicator of palliative care service provision,25 an association between morphine usage and palliative care service provision cannot be assumed in this RVX-208 study owing to insufficient data.
Where caregivers indicated that patients would pre-fer to have family or friends present, patients were more likely to die at home, if preferred, but where caregivers were unsure of the patient’s preference, pa-tients were more likely to die in hospital even if they preferred to die at home. While the patients were not asked directly by the interviewer regarding who they would like to be present, and the caregivers were not asked if they had asked the patients them-selves, which is a limitation in this study, Cordycepin is of concern that 47 (24.6%) of the 191 caregivers were
Table 4 Determinants of Death in the Preferred Place
Preferred Place of Death (Hospital)
Preferred Place of Death (Home)
Died in Hospital
Died at Home
Died in Hospital Died at Home
(Not Preferred) (Preferred)
Highest level of
Site of primary tumor
Did the patient want
family and/or friends
death, days, median
CHBAH ¼ Chris Hani Baragwanath Academic Hospital; IQR ¼ interquartile range; N/A ¼ not applicable (cell empty); NCD ¼ noncommunicable diseasedhypertension, diabetes, cardiac diseases/stroke, asthma/ chronic obstructive pulmonary disease. Variables with P < 0.05 are shown in bold face. aFisher’s exact test. bStudent’s t-test. cWilcoxon rank-sum test. d23 of the 28 were not sure. e4 of the 5 were not sure. f17 of the 22 were not sure.
Blanchard et al.
Table 5 Multivariate Logistic Regression Model of the Factors Associated With Congruence Between Preferred and Actual Place of Death
Characteristics Univariate Analysis OR (95% CI) Multivariate Analysis OR (95% CI) P-Value
Preferred place of death