TOPLINE:
In a matched cohort study of 410,120 patients with cancer, emergency department (ED) visits within 90 days before diagnosis were associated with a significantly increased risk for mortality, and the elevated risk persisted through 7 years of follow-up.
METHODOLOGY:Researchers conducted a retrospective study of 410,120 adults diagnosed with cancer between 2014 and 2021 in Ontario, Canada, with follow-up until death, 7 years, or March 31, 2024.A total of 205,060 patients who had ED visits in the 90 days before cancer diagnosis were matched 1:1 with patients without such visits, based on sex, year of diagnosis, and propensity scores.The primary outcome was all-cause mortality.TAKEAWAY:Overall mortality was 61.7% in patients with ED use vs 37.8% in those without ED use; the risk remained consistently higher among patients with ED use across all timepoints.The elevated risk for mortality among patients with ED use was highest at 30 days (hazard ratio [HR], 4.49; 95% CI, 4.40-4.58) and remained significant over 7 years (HR at 7 years, 1.05; 95% CI, 1.01-1.09).Patients hospitalized at the ED visit had a higher risk for mortality at all timepoints (HR at 30 days, 5.83; 95% CI, 5.69-5.99; HR at 7 years, 1.30; 95% CI, 1.23-1.37).Additionally, discharged patients had a higher risk for mortality up to 3 years (HR at 30 days, 2.68; 95% CI, 2.59-2.77; HR at 3 years, 1.38; 95% CI, 1.34-1.41; HR at 7 years, 1.03; 95% CI, 0.98-1.10), but the association was not significant for follow-up beyond 3 years.IN PRACTICE:
Higher risk for mortality among patients presenting to the ED prior to cancer diagnosis emphasizes the need for “established systems of care to ensure timely cancer workup for patients with suspected cancer in the ED” and “health care system improvements to enhance early cancer detection and management, thereby reducing the reliance on emergency care for initial cancer presentations,” the study authors wrote.
SOURCE:
The study was led by Keerat Grewal, MD, MSc, Schwartz/Reisman Emergency Medicine Institute, Sinai Health, Toronto, Ontario, Canada, and was published online on July 22, 2025, in JAMA Network Open.
LIMITATIONS:
Unrelated ED visits may have been included. The observational design limited causal inference. The cancer stage at diagnosis and race and ethnicity were not accounted for.
DISCLOSURES:
The study was supported by Canadian Institutes of Health Research and ICES, which is funded by the Ontario Ministry of Health and the Ministry of Long-Term Care. One author reported receiving a salary from ICES outside the submitted work. Several authors reported receiving research grants or personal fees or serving as experts for various organizations.