Validation of Malnutrition Clinical Characteristics in Critically Ill Patients

Abstract

Background: This study aimed to validate the Malnutrition Clinical Characteristics (MCC) compared with the Subjective Global Assessment (SGA), considering anthropometric measures, comorbidities, and mortality in critically ill patients.

Methods: This longitudinal observational study included patients admitted to the general intensive care unit (ICU) of a public hospital. SGA was used as the reference standard for diagnosing malnutrition. The inclusion criteria were patients receiving nutrition support therapy and age >18 years. The nutrition therapy was optimized as close as possible to 100% of the patients' energy and protein needs regardless of the access route. Hospital length of stay (LOS), comorbidities on admission, and death were documented during the entire hospitalization of each patient. Body mass index (BMI), midarm circumference (MAC), and calf circumference (CC) were considered anthropometric measures.

Results: The convenience sample comprised 102 ICU patients. Comparing the original malnutrition classifications of SGA with MCC, the specificity was 87.5%, sensitivity was 100%, accuracy was 93.3%, positive predictive value was 87.5%, and negative predictive value was 100%. When classified in 2 groups, namely "well-nourished" and "malnourished," specificity and sensitivity were 100% between both groups. Malnourished patients had significantly higher mortality rates (P = .006) and longer LOSs (P <.001). As expected, BMI, MAC, and CC results were similar for SGA and MCC.

Conclusions: MCC was a valid tool for classifying malnutrition in ICU patients. Because the evaluation is fast and does not require expensive equipment that is difficult to handle, it is believed to be practical, low-cost, and easy to use.

Keywords: critical illness; length of stay; malnutrition; mortality; nutrition assessment; patient outcomes.