Predictors of 30-Day Recurrent ED Visits for Hyperglycemia
Justin W. Yan et al.
Academic Emergency Medicine, August 27, 2025
🩺 Clinical Implications for EM
-
Substance use history is a strong predictor—screen and consider social support/referral.
-
Higher initial glucose levels = higher revisit risk—even in absence of DKA/HHS.
-
Patients with a new DM diagnosis or recent hospital discharge may be lower risk for near-term return.
-
Consider these factors when deciding:
-
Discharge vs. observation
-
Need for early outpatient follow-up
-
Involvement of social work or addiction services
🧪 Study Design
-
Type: Prospective multicenter cohort study
-
Setting: 4 Canadian tertiary academic EDs
-
Population: Adults (≥18 years) presenting with:
-
Hyperglycemia
-
DKA (diabetic ketoacidosis)
-
HHS (hyperosmolar hyperglycemic state)
📈 Key Findings
-
Total patients: 594
-
30-day recurrent ED visits: 13.5% (n = 80)
🔍 Independent Predictors of 30-Day Revisit
-
Complete case analysis:
-
🔺 Increased risk:
-
Substance abuse history → OR 2.32
-
Higher initial blood glucose → OR 1.04 per unit increase
-
🔻 Decreased risk:
-
New diabetes diagnosis → OR 0.29
-
Sensitivity analysis (with imputation):
-
🔺 Increased risk:
-
Substance abuse history → OR 2.55
-
ED visit in prior 14 days → OR 2.14
-
Higher initial blood glucose → OR 1.04
-
🔻 Decreased risk:
-
Recent hospitalization (last 30 days) → OR 0.40
-
New diabetes diagnosis → OR 0.37
Yan, J.W., Van Aarsen, K., Thorne, J., Karp, I., Spaic, T., Liu, S.L., Woods, N. and Stiell, I.G., 2025. A multicenter prospective cohort study evaluating 30-day outcomes after an emergency department visit for hyperglycemia in Canadian adults with type 1 or 2 diabetes. Canadian Journal of Emergency Medicine, pp.1-8.