Emergency Department Visits made by Patients with Cancer; Analysis of Data from a Single Community Cancer Center

Abstract

Purpose: Cancer-related Emergency Department Visits (EDV) are costly and may indicate poor care. Most studies of cancer-related EDV identify patients using inclusive diagnostic codes but lack precision since they don’t distinguish active cancer. We compared estimates of oncology-related EDV made by diagnostic code methods to a more specific method followed by chart review. We also studied characteristics of validated EDV. Methods: EDV from cancer patients at a single acute care hospital were measured using any inclusive oncology codes and was compared to EDV made by patients who were active attendees at cancer clinics. We then reviewed the records of a 50% random sample of the ‘active’ patients to estimate how many were related to cancer or cancer treatment. Results: Over 5 months, 790 oncology-EDV were identified by coding, but only 554 (70%) were made by ‘active’ patients. After review, 29% of active patient EDV was determined not to be related to an oncology problem or treatment. 48% of EDV occurred during daytime clinic hours. 79% were preceded by one or more contacts with the oncology care team within a week. There was variability in the number of EDV by patients of different oncologists. Conclusion: The impact of cancer in overall EDV counts is over-estimated by coding because coding cannot distinguish between active and inactive cancer nor discriminate between symptoms likely due to unlikely due to cancer or cancer treatments. Cancer programs should study the experiences of their own patients to design effective programs to reduce potentially avoidable utilization.