Keerat Grewal: Emergency Department Outcomes for patients with Cancer


Almost half of all Canadians will develop cancer during their lifetime. Studies show that patients with cancer require the emergency department (ED) frequently. However, little research has been done where oncology and emergency medicine converge. The NIH recognized this deficiency and developed CONCERN, a collaboration between ED and oncology researchers, to bridge this gap in the U.S.

Our proposed program of research is a Canadian collaboration of experts in emergency medicine, oncology, and health services research. The goal is to improve ED care for oncology patients. We have identified several priority research questions, including analysis of why oncology patients require the ED, impact of ED boarding, and ED triage for cancer patients.


We will use provincial administrative data from ICES and Cancer Care Ontario to examine the management and outcomes of oncology patients undergoing treatment who require emergency care in Ontario. We will include adult patients who received chemotherapy in the 30 days prior to an ED visit between Jan 2013 and Dec 2017.

  • ED Use: Exploratory analysis to examine the most common ED presenting complaints and diagnoses. We will also assess hospital admissions, 30-day mortality, and ED LOS.
  • ED LOS for Boarded Patients: We will include oncology patients who required inpatient admission to a general ward. The primary outcome of interest is 30-day mortality. Secondary outcomes are: in-hospital mortality, hospital LOS, ICU transfer. The main exposure variable is ‘prolonged ED boarding’, defined as ≥ 8 hours from time the decision was made to admit the patient to the time of transfer to an in-patient bed. To examine the impact of prolonged ED boarding on 30-day mortality, in-hospital mortality, and ICU transfer, we will conduct multivariable logistic regression using a generalized estimating equations approach to account for clustering by emergency department. For hospital LOS, we will use quantile regression given the non-normal nature of the data.
  • ED Triage: We will include patients triaged using eCTAS. We will examine patients with presenting complaints of fever, chest pain, and shortness of breath. The primary outcome of interest is hospital admission. Secondary outcomes are: 30-day mortality and repeat ED visits. The main exposure variable is to compare patients with and without cancer. For analysis, we will hard match patients on presenting complaint. We will then use propensity score matching to match patients with cancer to those without cancer during the same time period. Conditional logistic regression accounting for the paired nature of the data will be conducted and outcomes will be stratified based on the CTAS score to determine whether there are differences in outcomes between patients with and without cancer in each triage category.


Cancer Care Ontario has identified unplanned emergency visits by patients with actively treated cancer a strategic priority for the cancer system. Therefore, it is important to examine why oncology patients require the ED, their management and outcomes in the ED. By examining these factors, high impact interventions can be proposed to improve ED care and ultimately reduce emergency visits by oncology patients.

Leave a Reply

Your email address will not be published. Required fields are marked