Geriatric nutritional risk index independently predicts delirium in older patients in intensive care units: A multicenter cohort study

Fangxin Wei,Hongtao Cheng,Ruiyao He,Xin Yang,Ziqi Hu,Jun Lyu,Yu Wang
DOI: https://doi.org/10.1016/j.archger.2023.105288
Abstract:Background: Malnutrition is associated with clinical outcomes in several diseases. However, the role of malnutrition in the risk of delirium in elderly patients in intensive care units (ICU) remains unclear. This study aimed to elucidate the association between the Geriatric Nutritional Risk Index (GNRI) and the risk of delirium in elderly patients in the ICU. Methods: Elderly patients from the Medical Information Mart for Intensive Care IV (MIMIC-IV) and the eICU Collaborative Research Database (eICU-CRD) were included in this study, totaling 11,310 and 5,627, respectively. All subjects were stratified according to their GNRI score: those with no nutritional risk (GNRI > 98), mild nutritional risk (92 ≤ GNRI ≤ 98), moderate nutritional risk (82 ≤ GNRI ≤ 92), and severe nutritional risk (GNRI < 82). The primary outcome was delirium. Results: The GNRI in MIMIC-IV and eICU-CRD were 93.82 (84.88, 101.26) and 83.39 (75.95, 89.35), respectively. Compared to the cohort without nutritional risk, there was an increased risk of delirium in patients with mild nutritional risk (OR: 1.39, 95 % CI: 1.23-1.58), moderate nutritional risk (OR: 1.72, 95 % CI: 1.52-1.95), and severe nutritional risk (OR: 2.07, 95 % CI: 1.79-2.39). Inclusion of the GNRI score in the multivariate models improved the prediction of delirium. Similar results were found in the eICU-CRD database. Conclusions: Nutritional assessment using the GNRI may be an independent predictor of the development of delirium in elderly patients in the ICU. The addition of the GNRI score to the delirium prediction model improves its predictive accuracy.
What problem does this paper attempt to address?