Predictors of 30-Day Recurrent Emergency Department Visits for Hyperglycemia in Patients With Diabetes: A Multicentre Prospective Cohort Study.

Abstract

Identifying predictors of increased healthcare utilization for hyperglycemia may have important implications for designing interventions to improve patient outcomes and reduce costs. Studies examining predictors of 30-day recurrent ED hyperglycemia visits have been limited due to their retrospective nature. This study's objective was to prospectively identify predictors of 30-day recurrent ED visits for hyperglycemia in patients with diabetes. We conducted a multicentre, prospective cohort study of adults ≥ 18 years at one of four Canadian tertiary care, academic EDs with a diagnosis of hyperglycemia, diabetic ketoacidosis, or hyperosmolar hyperglycemic state. Multivariable logistic regression analysis was used to identify variables independently associated with recurrent 30-day ED visits for hyperglycemia. We enrolled 594 patients; 80 (13.5%) had a recurrent ED visit for hyperglycemia within 30 days. Independently associated predictors of 30-day recurrent visits on complete case analysis include substance abuse history (odds ratio [OR] 2.32, 95% confidence interval [CI]: 1.23-4.38) and initial laboratory blood glucose (OR 1.04, 95% CI: 1.01-1.07), while a new diabetes diagnosis was negatively associated (OR 0.29, 95% CI: 0.09-0.94). Sensitivity analysis using multiple imputation for missing data found the following independently associated variables: substance abuse history (OR 2.55, 95% CI: 1.34-4.85), previous ED visit within the past 14 days (OR 2.14, 95% CI: 1.02-4.48), and initial laboratory blood glucose (OR 1.04, 95% CI: 1.01-1.07). Two variables were negatively associated: recent hospitalization within the past 30 days (OR 0.40, 95% CI: 0.19-0.98) and new diabetes diagnosis (OR 0.37, 95% CI: 0.14-0.97). This multicentre prospective study reports predictors independently associated with 30-day recurrent ED visits for hyperglycemia. These predictors should be considered by ED clinicians when making disposition and follow-up plans for this important patient population, and future interventions should explore the interaction between hyperglycemia and substance use to prevent recurrent ED visits and reduce healthcare system costs and utilization.