Comprehensive Nutrition Assessment before Left Ventricular Assist Device Implantation in Chronically Ill Hospitalized Patients

Purpose Malnutrition is associated with mortality in hospitalized patients with heart failure. The Academy of Nutrition and Dietetics and the American Society for Parenteral and Enteral Nutrition recommend a standardized set of diagnostic characteristics to identify malnutrition in clinical practice. This study sought to determine the impact of comprehensive nutrition assessment on outcomes in patients undergoing left ventricular assist device implantation (LVAD), focusing on chronically ill hospitalized patients. Methods From February 2009 to August 2018, 151 patients who received a nutritional assessment prior to LVAD implantation were retrospectively reviewed. The malnutrition score (MS) was determined from 0-3 based on weight loss, decreased energy intake, body fat depletion, and muscle mass depletion prior to surgery. Serum albumin level was also assessed as a nutrition marker. Patients were categorized into 3 groups based on MS score and albumin level: a MS of 0 and albumin = 3.5 g/dL (Group 1, n = 68), a MS of 1-3 and albumin >= 3.5 g/dL (Group 2, n = 43), a MS of 1-3 and albumin Results All patients were chronically ill and deemed at risk of developing malnutrition. Median hospital stay before LVAD implant was 10 days (IQR 6 - 17). Median age of cohort was 61 years (IQR 52 - 69) and 129 (85%) were male. Ten (8.9%) were INTERMACS class 1, 63 (42%) required mechanical circulatory support prior to LVAD, and 87 (58%) were for destination therapy. Median preoperative albumin level was 3.7 g/ dL (IQR 3.3 - 3.95) and median MS was 1 (IQR 0 - 2). Overall in-hospital mortality was 10.7%. The preoperative albumin level ( = 3.5 g/ dL) did not affect the in-hospital mortality in the entire cohort (p = 0.77). Patients who had a positive MS tended to have higher mortality compared to patients with a MS of 0 (14.5% vs. 5.9%, p = 0.09). The in-hospital mortality was 5.9% in Group 1, 16.8% in Group 2, and 12.8% in Group 3 (p = 0.2). Median length of ICU stay was 8 days (IQR 6 - 11) in Group 1, 7 (IQR 5.75 - 11) in Group 2, and 7 (IQR 6 - 12) in Group 3 (p = 0.82). Conclusion Preoperative albumin level alone was not a useful marker to predict outcome in our cohort of chronically ill hospitalized patients. Routine comprehensive malnutrition assessment may help to identify high-risk cohort.