Background: Sepsis, a syndrome characterized by diverse clinical and biological features, remains a leading cause of mortality in intensive care units (ICUs), with limited availability of effective targeted treatments. Assessing treatment effects across stratified patient subgroups may provide insights into differential responses and guide precision interventions.
Objectives: This study aims to evaluate the differential effects of intravenous (IV) amino acid treatment on mortality across propensity score–based clusters in critically ill patients with sepsis.
Methods: A retrospective cohort study was conducted using the Korean National Health Insurance claims database from 2017 to 2021, including adult patients admitted to intensive care units (ICUs) for sepsis. Patients with ICU stays ≤2 days, prior ICU admissions for sepsis within a year, pregnancy-related conditions, palliative care, cardiac arrest, hepatic encephalopathy, or moderate-to-severe liver disease with lactulose administration were excluded. Patients were categorized into treatment and control groups based on exposure to IV amino acid treatment, and 1:1 propensity score matching was applied. The primary outcomes were all-cause mortality at 30, 60, 90, and 180 days. Between-group comparisons were conducted using the Chi-square test for categorical variables and the Mann–Whitney U test for continuous variables. Logistic regression was used to estimate adjusted odds ratios with 95% confidence intervals. Subgroup analysis was performed by stratifying the matched cohort into five groups based on propensity score ranges (0–0.2, 0.2–0.4, 0.4–0.6, 0.6–0.8, 0.8–1.0) to explore potential heterogeneity in treatment effect. All analyses were performed using SAS software (version 9.4; SAS Institute, Cary, NC, USA).
Results: Among 510,864 ICU admissions for sepsis, 99,732 patients received IV amino acid treatment and 411,132 did not. After 1:1 propensity score matching, 92,662 pairs were analyzed. In the overall cohort, IV amino acid treatment was not associated with a reduction in mortality. However, in the 0.6–0.8 propensity score stratum, it significantly reduced 30-day mortality (adjusted OR: 0.80, 95% CI: 0.71–0.89) and 60-day mortality (adjusted OR: 0.87, 95% CI: 0.78–0.96), with a marginal effect observed at 90 days (adjusted OR: 0.90, 95% CI: 0.82–1.00). No significant benefit was found in other strata. In the 0.6–0.8 stratum, the treatment group received more intensive care, including longer ICU stays, vasopressor use, and mechanical ventilation, suggesting that the observed mortality reduction occurred despite greater baseline severity.
Conclusions: IV amino acid administration showed varying effects across propensity score–based subgroups, indicating heterogeneity in treatment response among sepsis patients. These findings highlight the need for further research to validate the effectiveness within each subgroup and to identify the mechanisms underlying the observed differences in response.
