Frailty and Hospitalization After Emergency Department Evaluation Within Triage Categories in Older Patients: A Retrospective Cohort Study
Natsumi Hata, Masaaki Nagae, Hiroyuki UmegakiABSTRACT
Objectives
Triage reflects acute urgency, whereas frailty may also influence disposition in older emergency department (ED) patients. This study aimed to estimate the association between Clinical Frailty Scale (CFS)‐defined frailty and hospitalization after ED evaluation within Japan Triage and Acuity Scale (JTAS) categories and assess whether the strength varied across triage categories.
Methods
We conducted a single‐center retrospective cohort study in an urban tertiary hospital in Japan. We identified consecutive triaged ED visits by patients aged ≥ 65 years from November 2025 to March 2026. Visits with CFS scores and covariates constituted the complete‐case cohort. CFS ≥ 5 was defined as frailty. The primary outcome was admission to the study hospital or acute‐care interhospital transfer. We used a mixed‐effects logistic regression model with a CFS‐by‐JTAS interaction term, adjusting for age, sex, mode of arrival, injury‐ or poisoning‐related visit, living arrangement, and long‐term care needs certification, with patient‐level random intercept. The global interaction was assessed using a likelihood‐ratio test.
Results
Among 3127 ED visits, 1526 resulted in hospitalization. In JTAS 1, the adjusted odds ratio (OR) for CFS ≥ 5 was not estimated because of quasi‐complete separation. In the model excluding JTAS 1, the adjusted OR was 1.00 (95% confidence interval [CI], 0.65–1.55) in JTAS 2, 1.54 (95% CI: 1.19–1.99) in JTAS 3, and 2.01 (95% CI: 1.13–3.58) in JTAS 4/5. The global CFS‐by‐JTAS interaction was not statistically significant (P for interaction = 0.091).
Conclusions
Category‐specific estimates suggested little association in JTAS 2 but were numerically larger in JTAS 3 and JTAS 4/5. However, the global interaction was nonsignificant, providing insufficient evidence that the association differed across categories. Prospective studies should examine the clinical, geriatric, social, decision‐related, and healthcare‐system factors that may underlie this association among older patients with comparable triage‐assessed urgency.