Predicting ICU Admission and Length of Stay in Geriatric ED Patients: Comparison of Clinical Scores, Frailty and Comorbidity

Introduction: The growing aging population has led to an increasing number of geriatric patients presenting to emergency departments (EDs), many of whom have complex clinical profiles characterized by frailty, comorbidity, and polypharmacy. Early identification of patients requiring intensive care is challenging in this population, as clinical presentations are often atypical and physiological responses may be blunted. Although various clinical scoring systems are available, their comparative performance in predicting intensive care unit (ICU) admission and ICU length of stay (LOS) in the general geriatric ED population remains unclear.Methods: This retrospective, single-center observational cohort study included patients aged ≥65 years presenting to the emergency department of a tertiary-care hospital. Clinical data, including vital signs and scores such as qSOFA, NEWS2, Clinical Frailty Scale (CFS) and Charlson Comorbidity Index (CCI) were collected. The primary outcome was ICU admission and the secondary outcome was ICU length of stay. Receiver operating characteristic (ROC) analysis and multivariable regression models were used to evaluate predictive performance and identify independent predictors.Results: A total of 294 patients were included, of whom 34 (11.6%) required ICU admission. Lower diastolic blood pressure, lower oxygen saturation and higher respiratory rate were significantly associated with ICU admission. Among the evaluated scores, qSOFA demonstrated the highest discriminatory performance (AUC = 0.702), followed by CCI (AUC = 0.653), NEWS2 (AUC = 0.633) and CFS (AUC = 0.625). In multivariable analysis, qSOFA (OR 3.33, 95% CI 1.78–6.24), CFS (OR 2.42, 95% CI 1.51–3.88), CCI (OR 1.24, 95% CI 1.03–1.49) and age (OR 1.07, 95% CI 1.01–1.13) were independent predictors of ICU admission, whereas NEWS2 was not. For ICU LOS, only CCI was independently associated with prolonged stay (β = 0.152, p = 0.030).Conclusion: In geriatric patients presenting to the emergency department, qSOFA appears to be a useful tool for predicting ICU admission, while frailty and comorbidity burden also play important roles. However, the overall discriminative performance of these scores is modest, suggesting that they should be used in conjunction with clinical judgment. ICU length of stay is primarily influenced by comorbidity burden rather than acute physiological parameters. An integrated approach combining acute and baseline patient characteristics may improve early risk stratification in this population.Keywords: Aged, Emergency Service, Hospital, Intensive Care Units, Frailty, Comorbidity, Risk Assessment