Background: Despite cancer patients preferring to spend their last days out-of-hospital, many make difficult visits to the emergency department (ED). Family physician continuity of care has been shown in some clinical situations to reduce ED utilization.
Objective: To determine if greater family physician continuity of care for cancer patients during the end-of-life is associated with less ED utilization.
Method: This retrospective, population-based study involved secondary analysis of linked administrative data files for 1992 to 1997. Sources included the Nova Scotia Cancer Registry, Vital Statistics, the Queen Elizabeth II Health Sciences Center Oncology Patient Information System and Palliative Care Program (PCP), Hospital Admissions/Separation data, and Physician Services information. Subjects included adults with a recorded date of cancer diagnosis who died of cancer and who had made at least three visits to a family physician during their last 6 months of life. The relationship between total ED visits and family physician continuity of care, developed using the Modified Modified Continuity Index (MMCI), was examined using negative binomial regression with adjustments for survival, year of death, sex, age, cancer type, region, PCP admission, specialty visits, hospital days, death location, income quintile, and total ambulatory visits.
Results: In total, 8702 subjects made 11,551 ED visits (median = 1.0); median MMCI was 0.83. Adjusted results indicate those experiencing low continuity (MMCI < 0.5) made 3.9 times more ED visits (rate ratio [RR] = 3.93; 95% CI [CI] = 3.57-4.34) than those experiencing high continuity (MMCI > or = 0.8) and patients experiencing moderate continuity (MMCI = 0.5-0.8) made twice as many ED visits (RR = 2.28; CI = 2.15-2.42).
Conclusion: Given this significant association between family physician continuity of care and ED visits during the end-of-life, and given international trends to reform primary care, active planning of strategies to facilitate such continuity should be encouraged.