Frailty, organ dysfunction and mortality in older ICU patients

Older patients now constitute a large proportion of ICU admissions, yet conventional severity scores may not adequately reflect their specific vulnerabilities. This study examined whether pre-ICU frailty, functional status, comorbidity burden and acute organ dysfunction at ICU admission were associated with 30-day all-cause mortality in patients aged ≥ 65 years, and whether the associations of geriatric vulnerability measures persisted in adjusted models that also included conventional physiology-based severity scores. In this multicentre prospective observational cohort study, 55 ICUs across 27 provinces in Türkiye enrolled consecutive eligible patients aged ≥ 65 years who had an ICU length of stay exceeding 48 h during a 1-month period. Pre-admission frailty and functional status were assessed using the Clinical Frailty Scale (CFS), FRAIL questionnaire and Katz Index of Activities of Daily Living (Katz ADL); comorbidity burden was assessed using the Charlson Comorbidity Index (CCI); and acute illness severity was assessed using the Sequential Organ Failure Assessment (SOFA), Acute Physiology and Chronic Health Evaluation II (APACHE II) and modified Nutrition Risk in the Critically Ill (mNUTRIC) scores. The primary outcome was 30-day all-cause mortality after ICU admission. Kaplan–Meier and Cox regression analyses were framed as conditional 48-h landmark analyses. Follow-up began at the 48-h landmark; observed death times were retained, and all patients alive at day 30 were administratively censored at landmark day 28. A total of 1,529 patients were included (median age, 77 years), and 30-day all-cause mortality was 33.7%. Frailty was common, with 71.2% of patients having a CFS ≥ 5. In the conventional primary Cox model, CFS (aHR 1.190 per one-point increase, 95% CI 1.129–1.255) and SOFA (aHR 1.182 per one-point increase, 95% CI 1.152–1.212) were associated with mortality. In the centre-clustered robust model, CFS (aHR 1.190, 95% CI 1.104–1.283) and SOFA (aHR 1.182, 95% CI 1.143–1.222) remained statistically significant, whereas CCI was estimated with less precision and was not statistically significant (aHR 1.034, 95% CI 0.989–1.081). The extended conventional and centre-clustered robust models including APACHE II yielded the same substantive pattern. Comparisons of organ-support therapies and ICU-acquired complications across vulnerability strata were descriptive and unadjusted. In this selected cohort of ICU patients aged 65 years or older with an ICU length of stay exceeding 48 h, pre-admission frailty and early organ dysfunction showed the most robust associations with 30-day all-cause mortality, whereas chronological age was not independently associated in the adjusted models. These findings should not be interpreted as evidence that chronological age is irrelevant, but rather that frailty and acute organ dysfunction remained associated with mortality after adjustment for chronological age in this cohort.