BackgroundAssessment of nutrition risk in the intensive care unit (ICU) is limited by characteristics of critically ill patients, and new methods have been investigated for their applicability and predictive validity. The aim of the present study was to evaluate the validity of bioelectrical impedance analysis (BIA) parameters as predictors of nutrition risk and clinical outcomes in critically ill patients.MethodsThis was a prospective cohort study of patients admitted to an ICU. The modified Nutrition Risk in the Critically Ill score was used for assessment of nutrition risk, and BIA was performed in the first 72 hours of admission. Phase angle (PA) measurements were obtained, and bioelectrical impedance vector analysis (BIVA) was used to classify patients by hydration status (BIVA >70%). Patients were followed until hospital discharge and evaluated for hospital mortality, ICU length of stay, length of hospitalization, and duration of mechanical ventilation.ResultsEighty‐nine patients were included (62.5 ± 14.1 years, 50.6% female). A PA <5.5o showed an accuracy of 79% (95% CI 0.59‐0.83) in identifying patients at high nutrition risk and was associated with nearly 2 times greater risk for an ICU length of stay longer than 5 days (relative risk = 2.18 [95% CI 1.39‐3.40]). Hyperhydration was a significant predictor of mortality (hazard ratio = 2.24 [95% CI 1.07‐4.68]). Higher resistance and reactance values, adjusted for height, were found in survivors compared with nonsurvivors.ConclusionThe predictive validity of BIA was satisfactory for the assessment of nutrition risk, ICU length of stay, and mortality in critically ill patients.
BackgroundThis study aimed to evaluate the feasibility and validity of the Global Leadership Initiative on Malnutrition (GLIM) criteria in the intensive care unit (ICU).MethodsThis was a cohort study involving critically ill patients. Diagnoses of malnutrition by the Subjective Global Assessment (SGA) and GLIM criteria within 24 h after ICU admission were prospectively performed. Patients were followed up until hospital discharge to assess the hospital/ICU length of stay (LOS), mechanical ventilation duration, ICU readmission, and hospital/ICU mortality. Three months after discharge, the patients were contacted to record outcomes (readmission and death). Agreement and accuracy tests and regression analyses were performed.ResultsGLIM criteria could be applied to 377 (83.7%) of 450 patients (64 [54–71] years old, 52.2% men). Malnutrition prevalence was 47.8% (n = 180) by SGA and 65.5% (n = 247) by GLIM criteria, presenting an area under the curve equal to 0.835 (95% confidence interval [CI], 0.790–0.880), sensitivity of 96.6%, and specificity of 70.3%. Malnutrition by GLIM criteria increased the odds of prolonged ICU LOS by 1.75 times (95% CI, 1.08–2.82) and ICU readmission by 2.66 times (95% CI, 1.15–6.14). Malnutrition by SGA also increased the odds of ICU readmission and the risk of ICU and hospital death more than twice.ConclusionThe GLIM criteria were highly feasible and presented high sensitivity, moderate specificity, and substantial agreement with the SGA in critically ill patients. It was an independent predictor of prolonged ICU LOS and ICU readmission, but it was not associated with death such as malnutrition diagnosed by SGA.
The American Society of Parenteral and Enteral Nutrition recommends nutritional risk (NR) screening in critically ill patients with NRS-2002 ≥ 3 as NR and ≥ 5 as high NR. The present study aimed to evaluate the predictive validity of different cut-off points of the NRS-2002 in the intensive care unit (ICU). A prospective cohort study was conducted with critically ill adult patients who were screened using the NRS-2002. Hospital and ICU length of stay (LOS), hospital and ICU mortality, and ICU readmission were evaluated as outcomes. Logistic and Cox regression analyses were performed to evaluate the prognostic value of NRS-2002, and an ROC curve was constructed to determine the best cut-off point for NRS-2002. A total of 374 patients (61.9±14.3 years, 51.1% males) were included in the study. Of these, 13.1% were classified as without NR, whereas 48.9% and 38.0% were classified as NR and high NR, respectively. An NRS-2002 score of ≥ 5 was associated with prolonged hospital LOS. The best cut-off point for NRS-2002 was a score ≥ 4, which was associated with prolonged hospital LOS (OR=2.13; 95% CI: 1.39–3.28), ICU readmission (OR=2.44; 95% CI: 1.14–5.22), ICU admission (HR=2.91; 95% CI: 1.47–5.78), and hospital mortality (HR=2.01; 95% CI: 1.24–3.25), but not with ICU prolonged LOS (p=0.688). Therefore, NRS-2002 ≥ 4 presented the most satisfactory predictive validity and should be considered for NR screening in the ICU setting. Future studies should confirm the cut-off point and its validity in predicting nutrition therapy interaction with outcomes.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.