Primary Care Continuity and Location of Death for Those with Cancer

Background: Continuity of primary care is known to be associated with both improved processes and outcomes of care. Despite continuity being a desired attribute of end-of-life care and despite the desire by most patients with cancer to die at home, there has been no health services research examining this relationship. Aim: To examine the association between family physician continuity of care and the location of death for patients with cancer. Design of study: A retrospective population-based study involving secondary data analysis of four linked administrative health databases spanning 6 years of information (1992-1997). Setting: Nova Scotia, Canada Participants: All those who died of cancer from 1992 to 1997 and had made at least three ambulatory visits to a family physician. Methods: The relationship of provider continuity of care and an out-of-hospital death was examined using logistic regression. Results: Out-of-hospital deaths accounted for 31.6% of the 9714 deaths in the study population. The mean provider continuity of care was 0.78 (standard deviation [SD] 0.22). Those who died out-of-hospital had a greater odds of having received high provider continuity (adjusted odds ratio [OR] = 1.54, 95% confidence interval [CI] = 1.22, 1.93) when compared to those who died in-hospital. There appears to be a modification of this effect by gender with a significant association found for males and not for females. The trends in the point estimates are, however, similar for both sexes. Conclusions: This study demonstrates an association between family physician continuity of care and the location of death for those with advanced cancer. Such continuity should be fostered in the development of models of integrated service delivery for end-of-life care.