Utility of Restricted Mean Survival Time Analysis for Renal Replacement Therapy in Patients with Sepsis Associated...

Sepsis-associated acute kidney injury (SA-AKI) is a common and high-mortality complication in the intensive care unit (ICU), and renal replacement therapy (RRT) is frequently used, yet its survival benefits remain unclear. This study aims to assess survival differences via restricted mean survival time (RMST) in patients with SA-AKI.

This retrospective cohort study included 577 patients diagnosed with SA-AKI from the ICU of Shenzhen People's Hospital between January 2017 and July 2019. All patients were categorized into RRT group (n=343) and non-RRT group (n=234) according to whether they need to received RRT. Survival analyses were conducted using Kaplan-Meier curves. Univariate and multivariable logistic regression analyses were employed for the selection of covariates. RMST analysis was performed to quantify the survival benefits associated with RRT.

Patients who need to received RRT had greater baseline illness severity and more frequent organ support (eg, higher SOFA scores and higher rates of shock and mechanical ventilation), suggesting potential confounding by indication. In unadjusted analyses, 28d mortality was higher in the RRT group than in the non-RRT group (44% vs 20%). After adjustment for multidrug-resistant (MDR) infection and SOFA score (both P < 0.05), the RMST difference between the RRT and non-RRT groups was 3.25 days (95% CI: 0.67 to 5.84; p = 0.01), which was statistically significant.

In this retrospective cohort study, after adjusting for SOFA score and MDR infection, the potential survival benefits of RRT in patients with SA-AKI were demonstrated. These findings suggest that poor outcomes in RRT recipients may be attributable to underlying risk factors rather than RRT itself. However, given the observational design, these results are hypothesis-generating and require prospective validation.