Do patient-reported symptoms predict emergency department visits in cancer patients? A population-based analysis.

STUDY OBJECTIVE Since 2007 in Ontario, Canada, the Edmonton Symptom Assessment System has been routinely used for cancer outpatients. The purpose of this study is to determine the relationship between individual patient symptoms and symptom severity, with the likelihood of an emergency department (ED) visit. METHODS The cohort included all cancer patients in Ontario who completed an Edmonton Symptom Assessment System between January 2007 and March 2009. Using multiple linked provincial health databases, we examined the adjusted association between symptom scores and the likelihood of an ED visit within 7 days of assessment. RESULTS The cohort included 45,118 patients whose first assessment contributed to the study, of whom 3.8% made a subsequent ED visit. A severe well-being score was associated with the highest odds of a subsequent ED visit (adjusted odds ratio [OR] 1.9; 95% confidence interval 1.5 to 2.4). Nausea, drowsiness, and shortness of breath with moderate or severe scores were associated with ED visits (adjusted OR 1.2 to 1.5), whereas pain, tiredness, poor appetite, and well-being had a significant association for mild scores (adjusted OR 1.2, 1.3, 1.2, and 1.3, respectively), moderate scores (adjusted OR 1.3, 1.5, 1.5, and 1.7, respectively), and severe scores (adjusted OR 1.4, 1.7, 1.7, and 1.9, respectively). Anxiety and depression were not associated with ED visits. CONCLUSION Worsening symptoms contribute to emergency visits in cancer patients. Specific symptoms such as pain are obvious management targets, but constitutional symptoms were associated with even higher odds of ED usage and therefore warrant detailed assessment to optimize both patient outcomes and resource use.