Journal of Palliative Care

We compare the effectiveness of terminal cancer pain management delivered by primary care physicians with that of specialists in home-based hospice settings. Each visit record for 223 outpatients in three hospice programs was abstracted for physician type, patient age, medication usage, level of pain reported, cancer type, and metastatic status. Thirteen percent of patients reported no pain at any visit and 19% reported pain at all visits; half of the patients reported pain at two thirds of their visits. No difference was found in the presence or absence of pain between primary care and oncologic patients. When available, the level of pain reported (0–10 scale) was statistically (p<0.01) but not clinically different between physician groups; average pain rating for primary care patients was 3.7 while the mean pain rating for oncologic patients was 3.1. The reported pain level varied significantly among facilities, as did physician mix. Multivariate analysis revealed that program and an interaction term between program and physician type, but not physician type independently, explained a significant amount of variation in pain level. Overall, reported pain remained higher than optimal. Research elsewhere has shown that application of the World Health Organization (WHO) cancer pain management guidelines can control 70%–90% of cancer pain. Strategies for implementing pain guidelines that emphasize a systems approach may be effective.

The timing of death has received much attention, particularly in the area of sudden cardiac death. Many studies have demonstrated that sudden cardiac death and other sudden deaths follow a circadian pattern. Deaths have also been reported to vary around dates that are especially meaningful to patients and families. To test these reported observations in a cancer palliative care population, we reviewed the date and time of death of 626 consecutive patients admitted to the palliative care unit of a western Canadian hospital. All patients were adults with advanced metastatic or locally recurrent cancer. A cir-cadian distribution in the time of deaths was observed; 225 deaths occurred between 20:00 and 06:00 (261 deaths expected) versus 401 deaths between 06:00 and 20:00 (365 deaths expected) (p=0.0037). The distribution of deaths did not change significantly according to day of the week or month of the year. Based on patient birthday, 41 deaths occurred during the three weeks before a birthday (33.5 deaths expected) versus 26 deaths during the three weeks after a birthday (33.5 deaths expected) (p=0.067). There appear to be fewer deaths during the evening and night; there does not appear to be a decline in deaths before the patient's birthday; and any temporal distribution of deaths in this population appears to be minimal.

To ensure quality of care, palliative care programs need to document the effectiveness of their relief of physical and psychological distress. The Edmonton Symptom Assessment Scale (ESAS) is a validated, reliable instrument developed to measure 9 different symptoms in palliative care patients. To see if symptom management could be compared across institutions, we first reviewed the charts of 188 successive admissions to the palliative care unit at St. Boniface General Hospital, Winnipeg, Manitoba. Our study showed that the ESAS is a useful audit tool for assessing patterns of palliative symptom control that allows for institutional comparisons. Procedures that ensure completeness of data collection remain to be developed.