PROJECT SUMMARY/ABSTRACT
Cancer is a common diagnosis in the emergency department (ED) and by the time patients reach the ED, their
cancer has often progressed to later stages. EDs are not intended to diagnose cancer. Training in cancer
management is limited for ED physicians, and patient access to follow-up care after a visit to the ED is often
difficult. Research from other western countries tells us that, compared to non-emergency settings, cancer
diagnosis through the ED is an independent predictor for worse outcomes, including poorer overall survival.
Very little is known about emergency diagnosis of cancer in the United States (U.S.). From a handful of studies
conducted in a few cancer types, we know that up to 30% of cancer patients are diagnosed as emergencies,
and low income and racial/ethnic minority patients are substantially more likely to be affected. However, across
different cancer types, the burden and risk factors of cancer diagnosis as an emergency have not been
studied. Furthermore, no population-based studies in the U.S. have compared survival differences in patients
with cancers diagnosed in EDs vs other non-emergent settings. And among those who are diagnosed as
emergencies, it is unclear why they visited the ED, rather than going to a primary care provider for their
diagnosis. Our long term goal is to establish the epidemiology of emergency cancer diagnosis in the U.S. We
will describe the burden of cancer diagnosis in the emergency department including the patient characteristics
of those most affected and quantify disparities across vulnerable populations defined by socioeconomic status,
race/ethnicity, and geographic isolation (Aim 1). We will estimate the relative importance of ED (compared to
non-ED) diagnosis on patient survival, after controlling for known risk factors like cancer stage, treatment,
patient age and chronic conditions (Aim 2). Finally, we will investigate modifiable drivers of disparities among
patients diagnosed as emergencies by examining their pre- and post-diagnostic patterns of care to identify
opportunities for prevention and improved outcomes (Aim 3). Our highly efficient study design uses
epidemiologic methods to analyze high-quality, population-based SEER-Medicare data for 1.8 million,
Americans who were diagnosed cancers of the esophagus, stomach, colon/rectum, liver, pancreas, lung,
breast, uterus, ovary, prostate, bladder, kidney, non-Hodgkin’s lymphoma, myeloma, and leukemia. The
investigative team includes experts in cancer epidemiology, healthcare delivery, cancer health disparities,
medical sociology, and oncology, family, and emergency medicine. This will be the first large-scale study of
emergency cancer diagnosis in the U.S. Establishing a base of evidence about this important problem,
including who is affected, where, and why, will illuminate a largely unrecognized issue, and identify modifiable
drivers of disparities in patients with the poorest prognoses.