Association Between AND‐ASPEN Malnutrition Criteria and Hospital Mortality in Critically Ill Trauma Patients: A Prospective Cohort Study

Abstract

Background Malnutrition is prevalent in trauma victims because of intense muscle wasting triggered by traumatic events and is a mortality risk. The Academy of Nutrition and Dietetics–American Society for Parenteral and Enteral Nutrition (AND‐ASPEN) tool has the potential to diagnose malnutrition in this population. This study aims to evaluate this criterion as a malnutrition diagnostic tool and as a hospital mortality predictor in critically ill trauma patients. Methods We performed an observational prospective cohort study from April 2015 to February 2017 in a Brazilian hospital. Patients were >18 years old, remained in the intensive care unit (ICU) >48 hours, and had completed data regarding the recommended clinical characteristics (RCCs) for malnutrition. The main exposure of interest was malnutrition assessed by a specialist dietitian using AND‐ASPEN. The primary outcome was all causes of mortality during hospital stay. Secondary outcome was the RCCs assessment, to predict malnutrition and hospital mortality. Results Included were 414 trauma patients. Malnutrition prevalence was 26.8% at ICU admission. The main analysis showed that malnutrition increases mortality chances 1.96 times (95% CI, 1.13–3.30; P = .015). Secondary analysis showed that all 5 RCCs were significantly associated with malnutrition. The only RCC associated with mortality was energy intake (odds ratio 1.5; 95% CI, 1.1–2.0; P = .008). Conclusions AND‐ASPEN criterion is a feasible and accurate method to recognize malnutrition and predict hospital mortality when applied by trained dietitians for the critical care trauma population. Further studies are needed considering patient subgroups such as obese and elderly.

Concurrent and Predictive Validity of AND‐ASPEN Malnutrition Consensus Is Satisfactory in Hospitalized Patients: A Longitudinal Study

“…Abahuje et al (2020) demonstrated an increase in the risk of hospital deaths (hazard ratio, 27.25; 95% CI, 2.34–316.94) in a sample of 279 surgical patients in acute care 17 . Although the critical patients are a specific population and different from the sample of the current study, other authors have been demonstrating that malnutrition evaluated by the AND‐ASPEN Consensus is also predictor of hospital mortality in ICU 16,24,26 . All studies that aimed to evaluate the predictive validity of malnutrition defined by AND‐ASPEN indicated a significative association between this condition and the outcomes analyzed (ie, prolonged LOS, mortality in hospitals and after discharge, hospital readmission), similar to the findings of the present study.…”

“…A retrospective cohort conducted with 409 hospitalized patients demonstrated an accuracy of 83.6% (95% CI, 0.79–0.88), κ of 0.710, sensitivity of 90.8%, and specificity of 75.5% 15 for AND‐ASPEN Consensus. In a recent Brazilian prospective study involving 414 critically ill trauma patients, Ceniccola et al (2020) also demonstrated satisfactory accuracy (area under the curve [AUC] = 0.78; 95% CI, 0.72–0.83), with moderate sensitivity (89.2%), low specificity (66.4%), and a κ coefficient equal to 0.570 16 . Likewise, recently, Abahuje et al (2020) showed a satisfactory accuracy of AND‐ASPEN to identify malnutrition (AUC = 0.75; CI not reported), but also a moderate κ coefficient value (0.529) 17 .…”

“…The large CI for in‐hospital mortality and for 6‐month mortality suggests an imprecise result, which possibly may be explained by the low incidence of death in the sample, in both cases. About 10 studies have been published regarding the evaluation of the predictive validity of the AND‐ASPEN Consensus 16‐19,21‐26,40 . A prospective longitudinal study carried out with 490 oncological and surgical patients from a US hospital described that malnutrition defined by AND‐ASPEN was a predictor of prolonged LOS (odds ratio, 1.67; without description of 95% CI) 21 .…”

“…In a literature search, 4 studies were identified that aimed to evaluate the concurrent validity of the AND‐ASPEN Consensus in hospitalized patients, considering the SGA as the reference method, 15‐18 and all of them demonstrated satisfactory accuracy (>75%) and moderate to substantial agreement between the tools. In addition, according to the available literature about AND‐ASPEN, this tool is also able to predict satisfactory clinical outcomes 16,17,19‐26 . Recent studies have also explored the concurrent validity of the GLIM criteria vs SGA, 18,27 as well as its predictive validity, 28‐32 with satisfactory results being demonstrated.…”

Consensus on the standard terminology used in the nutrition care of adult patients with chronic kidney disease

“…Studies reported satisfactory levels of accuracy and moderate agreement for the MCC compared to the SGA in adult hospitalized patients 94 , individuals with severe conditions in general, trauma 95 , and surgery patients 96 . In regard to outcomes, the MCC predicted longer hospitalization times 97 and higher care costs 98 .…”