Journal of the American Medical Directors Association · Nov 16, 2025

Diabetes Medications in Post-Acute Care: Association With Emergency Department Visits and Hospitalization

By Bogin MH, Takahashi PY, Johnson MG et al.

OBJECTIVES: Diabetes is common, but optimal diabetes management for patients undergoing post-acute rehabilitation is unclear. Our primary aim was to determine whether various diabetes medication classes were associated with emergency department (ED) visits and/or hospitalization (termed the combined outcome) for this patient population. DESIGN: Retrospective cohort study. SETTINGS AND PARTICIPANTS: Eligible patients (aged ≥60 years) were discharged from hospitals to skilled nursing facilities (SNFs) for postacute care in southeast Minnesota from October 1, 2019, through June 30, 2021. METHODS: The primary outcome was the total number of ED visits and/or hospitalizations (the combined outcome) during the 30 days after SNF post-acute care admission. The predictors were insulin, metformin, sulfonylureas, dipeptidyl peptidase-4 inhibitors, thiazolidinediones, sodium-glucose cotransporter-2 inhibitors, and glucagon-like peptide-1/glucose-dependent insulinotropic polypeptide receptor agonists. To identify associations, we used logistic regression, with adjustment for age, gender, Charlson Comorbidity Index score, and marital status. RESULTS: The cohort comprised 1784 patients (95.3% were White, 46.1% were female). Within 30 days after their SNF admission, 526 patients (29.5%) were evaluated in the ED and/or hospitalized. Insulin use was positively associated with the combined outcome, and this association persisted after adjusting for age, gender, and comorbid conditions (odds ratio, 1.275; 95% CI, 1.020-1.594). In contrast, metformin had a reduced association with the combined outcome (odds ratio, 0.714; 95% CI, 0.576-0.884). Other diabetes medication classes (eg, dipeptidyl peptidase-4 inhibitors, thiazolidinediones) were not associated with the combined outcome. CONCLUSIONS AND IMPLICATIONS: Post-acute rehabilitation patients who used insulin had significantly higher rates of ED visits and/or hospitalizations within 30 days after SNF admission than those who did not use insulin. Further studies are needed to determine the reasons underlying this finding and to identify an optimal pharmacotherapy regimen for residents with diabetes receiving post-acute care in SNFs, who constitute a unique patient population.

doi.org/10.1016/j.jamda.2025.105997PubMed 41241384Interventions