Effect of early enteral nutrition on short-term outcomes and survival in patients in the cardiothoracic intensive care unit: An inverse probability weighted retrospective study

Nutrition. 2024 Mar:119:112328. doi: 10.1016/j.nut.2023.112328. Epub 2023 Dec 15.

Abstract

Objectives: The merits of early enteral nutrition (EEN) in patients in the cardiothoracic intensive care unit (CTICU) remain unclear. This retrospective study aimed to address this issue.

Methods: We analyzed data from the MIMIC IV v2.0 database, including patients with a CTICU stay of ≥4 d. Patients were divided into early and delayed enteral nutrition (EN) groups. Differences in baseline data were corrected using an inverse probability weighting (IPW) approach. Generalized linear models (GLMs) were used to compare trends over time between groups, and survival effects were evaluated with weighted logistic and Cox regression, supplemented by weighted Kaplan-Meier curves. Subgroup analysis facilitated the exploration of potential interactions.

Results: The study included 720 CTICU patients. Following IPW, all baseline variables were balanced. EEN led to shorter hospital and CTICU stays, lower incidence of respiratory and blood infections, and reduced total insulin usage in the first week of CTICU admission, albeit with an increased total gastric residual volume. Mortality risk between the groups did not significantly differ at 28 d or at 1 y. Excessive early energy and protein intake elevated the risk of 28-d mortality, but the relationship may not be linear. Overweight patients or those with fewer comorbidities had a higher mortality risk with EEN.

Conclusions: EEN may improve short-term outcomes in CTICU patients without a clear survival benefit. Early high caloric and protein intake could lead to adverse outcomes, suggesting a careful evaluation for initiating EN in specific patients.

Keywords: Cardiothoracic surgery; Early nutrition; Enteral nutrition; Intensive care unit; Mortality.

MeSH terms

  • Enteral Nutrition*
  • Humans
  • Intensive Care Units*
  • Length of Stay
  • Probability
  • Retrospective Studies