Answers are grounded in this paper and cite their sources.
Sepsis-associated acute kidney injury (SA-AKI) is a common and severe complication in older intensive care unit (ICU) patients, yet static KDIGO staging fails to capture the dynamic evolution of renal function. We aimed to identify distinct serum creatinine (SCr) trajectory classes in older patients with SA-AKI and evaluate their independent association with 28-day mortality. In this retrospective cohort study, we analyzed data from the Medical Information Mart for Intensive Care IV (MIMIC-IV) database, version 3.0 (2008–2022). Older patients (≥ 65 years) who developed SA-AKI within 48 h of ICU admission and remained in the ICU for ≥ 96 h were included. The analysis is therefore restricted to 96-hour survivors. Latent class mixed models were applied to longitudinal SCr measurements during the first 96 h after sepsis onset. Multivariable logistic regression and Cox proportional hazards models were used to assess associations with 28-day all-cause mortality, with time origin at hospital admission, adjusting for age, sex, baseline SCr, SOFA score, and Charlson Comorbidity Index. Time-to-event analyses used a landmark at 96 h after sepsis onset, when the creatinine trajectory used for class assignment was complete, with follow-up to original hospital day 28. A total of 1,376 older patients with SA-AKI were classified into five SCr trajectory classes: Class 1 (transient mild AKI with rapid recovery, n = 920), Class 2 (early rise then decline, n = 254), Class 3 (severe AKI with delayed peak then decline, n = 92), Class 4 (U-shaped biphasic pattern, n = 39), and Class 5 (persistent worsening without recovery, n = 71). Class 5 exhibited the highest 28-day mortality (49.3%). Class 3 had the lowest crude 28-day mortality (15/92, 16.3%). In adjusted analyses, compared with Class 1, Class 5 was independently associated with increased 28-day mortality (adjusted OR 2.48, 95% CI 1.47–4.20; adjusted HR 1.88, 95% CI 1.30–2.73). Class 3 had an adjusted logistic OR of 0.52 (95% CI 0.27–0.98); the landmark Cox HR was 0.63 (95% CI 0.37–1.08). Adding trajectory class increased AUC from 0.661 to 0.685 (absolute difference 0.024; DeLong P < 0.001). Both models had only modest discrimination. The AUC increment is exploratory. A CKD-stratified analysis showed no significant interaction ( P > 0.05) but was imprecise, especially Class 4 (only two patients with CKD). Among older 96-hour ICU survivors with SA-AKI, early SCr trajectory classes are independently associated with 28-day all-cause mortality. A persistent worsening trajectory identifies a high-risk subgroup. A delayed-peak then declining pattern had lower crude mortality, but that contrast is hypothesis-generating. These findings indicate a prognostic association in a single-center cohort. The five-class solution has not been externally validated. Not applicable.
No takes yet. Share an insight, caveat, or question.
Chen et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: