In-hospital development of new-onset frailty and discharge outcomes among patients with sepsis: a Korean multicenter registry study.

backgroundFrailty is an important marker of vulnerability and poor outcomes in patients with sepsis. Although prior studies have primarily focused on pre-existing frailty or post-discharge outcomes, little is known about the in-hospital development of frailty among patients who are non-frail at baseline.

methodsWe conducted a multicenter cohort study using the Korean Sepsis Alliance registry, a prospectively maintained database of adult patients with sepsis across 15 hospitals in South Korea. Analyses were restricted to patients without baseline frailty, defined as a premorbid Clinical Frailty Scale score < 5. The primary analysis categorized patients as remaining non-frail at discharge, having new-onset frailty at discharge, or dying during hospitalization. Multinomial logistic regression with prespecified covariates was used to evaluate clinical factors associated with these mutually exclusive outcomes, with in-hospital death treated as a competing outcome. Among hospital survivors, discharge destination was described according to discharge frailty status. In an exploratory analysis restricted to patients with new-onset frailty, we evaluated clinical factors associated with home discharge.

resultsAmong 6336 patients without baseline frailty, 1453 (22.9%) had new-onset frailty and 1527 (24.1%) died during hospitalization. Older age (per 1-SD increase, adjusted odds ratio [aOR], 1.38), higher Charlson Comorbidity Index (per 1-SD increase, aOR, 1.27), higher Sequential Organ Failure Assessment score (per 1-SD increase, aOR, 1.41), intensive care unit admission (aOR, 1.47), and pulmonary infection (aOR, 1.77) were associated after adjustment with new-onset frailty (all p < 0.001). Several of these factors were also associated with in-hospital death, with a larger adjusted estimate for SOFA score. Among hospital survivors with new-onset frailty, 42.8% were discharged home. In exploratory analyses restricted to patients with new-onset frailty, higher Sequential Organ Failure Assessment score, solid malignancy, and nursing home- or nursing hospital-acquired infection were associated with a lower likelihood of home discharge. In sensitivity analyses using inverse probability of censoring weighting, the overall pattern of associations for new-onset frailty remained generally consistent.

conclusionsNew-onset frailty and in-hospital death were common among patients with sepsis who were non-frail at baseline. Assessment of frailty at admission and discharge may help characterize functional decline and inform discharge planning after sepsis.